HK1133239B - System and control method for group management of elevator - Google Patents

Description

Elevator group management system and elevator group management control method

Technical Field

The present invention relates to an elevator group management system, and more particularly, to an allocation control of elevators which evaluates an elevator most appropriate for a newly generated hall call of an elevator door, taking into account the quality of service of future hall calls which have not yet occurred at the present point in time.

Background

An elevator group management system manages a plurality of elevator cars (hereinafter, simply referred to as "cars") as one elevator group, thereby providing efficient elevator operation services to users. Specifically, a plurality of elevator cars (typically 3 to 8) are managed as 1 elevator group, and when an elevator hall call occurs at a floor, an optimum car is selected from the elevator group, and the car is assigned to the elevator hall call.

In the known elevator group management system, basically, an evaluation function is calculated based on the predicted waiting time, and then allocation control is performed based on the evaluation function. In the above-described elevator group management system, when a new hall call occurs, the predicted waiting time of the hall call that has been accepted by each car is calculated, and the hall call is assigned to the car whose predicted waiting time is shortest, or the car whose maximum waiting time is shortest, or the car whose average waiting time is shortest. Here, the accepted elevator hall call refers to a newly generated elevator hall call that is temporarily assigned and an already assigned elevator hall call. The above-described method of performing assignment control based on the predicted waiting time is a basic method employed by each elevator manufacturer in elevator group management control, and has a problem that although optimal elevator car assignment is performed for an elevator hall call that has already occurred, sufficient consideration is not given to the influence of an elevator hall call that may occur in the future.

In order to solve the problems of the above-described elevator allocation methods for elevator hall calls that have occurred, various control schemes have been proposed in which the factor of elevator hall calls that will occur in the future is added to the evaluation target, and the main idea can be summarized as: 1) assuming a floor where a hall call will occur in the future, control is performed to shorten the predicted waiting time for the call, and 2) the elevator cars are arranged at equal intervals in time by the control. The former is a method of assuming and directly evaluating a floor on which a call will occur in the future. In the latter control, the waiting time of the future call is indirectly evaluated according to the time interval, and the effect is obtained that if the cars are controlled to be arranged at equal time intervals, when the future call occurs in the interval, the car located near the call floor serves the future call, and thus the maximum waiting time does not exceed the time interval.

The following exemplifies known control means taking into account the above-mentioned future calls in the control.

1) Evaluating the waiting time of future calls based on the estimated passenger occurrence rate (patent document 1)

Patent document 1 discloses a control scheme for estimating the waiting time of an elevator hall call to be generated in the future, in which the estimated occurrence rate of passengers (waiting passengers in the elevator hall) and the predicted waiting time values of the passengers are calculated to estimate a car most suitable for assignment to the newly generated elevator hall call. Patent document 1 describes that a car assigned to a future call is assumed to be the first car or the second car to arrive with probability p, although it is currently unknown which car will be used for assignment.

2) Method for performing allocation evaluation based on distribution of minimum predicted arrival time of each floor (patent document 2)

Patent document 2 discloses a control scheme for determining a distribution of service evaluation values to be provided to each floor as a service distribution evaluation index, and controlling an elevator based on the distribution of the service evaluation values. Patent document 2 describes, as a service distribution evaluation index, a distribution of minimum predicted arrival times, which are minimum values of predicted arrival times at which cars arrive at a certain floor, among the floors.

3) Evaluation of allocation to high-probability floors (patent document 3)

Patent document 3 discloses a scheme of selecting at least one high-probability floor having a high probability of long-time waiting from floors at which an elevator hall call has not occurred, and assuming that an elevator hall call has occurred on the high-probability floor. Then, an elevator to be served for the newly generated elevator hall call is determined based on the result of the evaluation of the assignment when an elevator to be served is assigned to each of the newly generated elevator hall call and the call of the high-probability floor.

4) Time interval evaluation for predicting car position (patent document 4)

Patent document 4 discloses a technique of predicting and calculating a time interval between elevators based on a predicted arrangement of the elevators at a certain time when an elevator allocated to an elevator hall call is selected, and determining a weighting coefficient of a call allocation evaluation function for each elevator based on a result of the prediction and calculation. Thus, the evaluation function of each elevator is calculated.

5) Evaluation of time interval for predicting car position (patent document 5)

Patent document 5 discloses a technique of predicting and calculating a car position and a car direction after a predetermined time has elapsed from a current time, predicting and calculating a time interval or a space interval of each car after the predetermined time, and calculating an assignment restriction evaluation value based on the predicted car interval.

Patent document 1 Japanese patent application laid-open No. 2006-298577

Patent document 2 Japanese patent application laid-open No. Hei 10-245163

Japanese patent laid-open No. 2006-213445 of patent document 3

Patent document 4 Japanese patent publication No. Hei 7-12890

Japanese patent publication No. Hei 7-72059 of patent document 5

In the above-described conventional technology, when evaluating a future call, it is assumed that the assumed future call is served by the elevator car having the shortest arrival time. However, in practice, the car to be serviced is determined based on the assignment evaluation function. In the assignment evaluation function, there are other evaluation items in addition to the evaluation item of the waiting time of the elevator hall call, and the elevator with the shortest arrival time is not always served. On the other hand, the number of elevator hall calls that have occurred is limited, while calls that may occur in the future are considered to be infinite, and if the evaluation function is calculated for each possibility, the amount of computation becomes enormous, which is not practical. However, if the number of assumed future calls is reduced in order to reduce the amount of computation, the evaluation of the future calls may become incorrect due to the reduction of the number of assumptions.

In summary, since the number of calls to be evaluated in the future is large, a simplex process is performed in the prior art, and it is assumed that the elevator with the shortest arrival time provides service. However, this method is inconsistent with the actual assignment evaluation function, so that accurate evaluation may not be obtained by using the method. Furthermore, if the evaluation is performed with an allocation evaluation function on the basis of a reduced number of future calls to be evaluated, elevator hall calls which occur substantially randomly must be predicted, so that there is still the possibility that an accurate evaluation cannot be obtained.

The following is a detailed description of the problems of the prior art.

In the method disclosed in patent document 1, it is assumed that the car assigned to a future call is not the first car to arrive, but the second car to arrive, based on the probability p, but it is difficult to determine the probability value because it is an assignment evaluation function rather than a probability that determines the actual assignment. Therefore, as long as the probability value is inaccurate, the evaluation value of the future call may also be incorrect.

In the method disclosed in patent document 2, the service quality is evaluated based on the minimum predicted arrival time of each car at the assumed floor of the future call, and the car with the minimum predicted arrival time is assumed as the assigned car for the future call. However, this method is contradictory to a method of determining actual allocation based on an allocation evaluation function, and thus accurate evaluation may not be obtained by using these methods.

In the method disclosed in patent document 3, a high-probability floor is selected and the evaluation value of each elevator is calculated for the high-probability floor, and when there are a plurality of high-probability floors, the evaluation value of each elevator must be obtained for each high-probability floor (including the determination of the ascending direction or the descending direction), which results in an extremely large amount of computation. Here, since future calls may occur in each direction of each floor in the time range from the current time, the number of objects to be considered is very large. Therefore, it is necessary to significantly reduce the number of floor calls with a high probability, but this increases the error between the predicted elevator hall call and the elevator hall call actually occurring thereafter. That is, there is difficulty in the selection of a high probability floor call, and therefore, a correct evaluation may not be obtained with this method.

In the methods disclosed in patent document 4 and patent document 5, it is assumed that a future call occurring in a floor zone in which the time interval between cars is calculated is serviced by a car behind the zone. This assumption is the same as the assumption that the car with the smallest predicted arrival time is serving, so there is a conflict between this method and the actual assignment method, with which a correct evaluation may not be obtained.

Disclosure of Invention

The present invention has been made in view of the above-described problems occurring in the prior art, and an object of the present invention is to enable a situation at a future time point to be reflected in evaluation of an elevator hall call that is likely to occur in the future in an assignment evaluation function of a newly registered elevator hall call, and to enable more accurate evaluation to be achieved by a simple method.

An elevator group management system in a preferred embodiment of the present invention includes a plurality of elevators serving a plurality of floors and an assignment control device for assigning an appropriate elevator to a newly generated elevator hall call based on an assignment evaluation index, the method is characterized in that an elevator hall call which is likely to occur in the future is extracted as a candidate future call, a plurality of elevators are classified into an elevator which can be allocated to the candidate future call and an elevator which cannot be allocated to the candidate future call, a future call allocation evaluation value when an allocable elevator which can be allocated to the candidate future call is calculated, an allocation evaluation index when allocated to a newly occurring elevator hall call is calculated on the basis of the future call allocation evaluation value, and deciding an elevator allocated to the newly occurring elevator hall call based on the allocation evaluation index.

That is, when allocating each elevator to a future call to a floor where the future call is likely to occur, the elevator is classified into an elevator that can be used for allocation and an elevator that cannot be used for allocation, and is selected only from the elevators that can be used for allocation.

For each elevator, it is predicted whether a long waiting time will occur for an assigned elevator hall call or whether the elevator car will be fully loaded, etc., and the elevator is classified in advance, and the quality of service for future call floors is evaluated for elevators that can be used for assignment with certainty.

Here, as an evaluation method of the assignment evaluation value of the future call, a method of performing evaluation based on the predicted waiting time and an evaluation method of evaluating whether or not the intervals between the plurality of elevators are appropriate are typical.

Effects of the invention

According to the preferred embodiment of the present invention, the service quality for the future call floor can be evaluated more accurately.

Further, according to the preferred embodiment of the present invention, the method of assigning elevator selection processing is simple, and the calculation amount can be reduced even when a large number of future calls are evaluated.

Other objects and features of the present invention will be clarified in the following embodiments.

Drawings

Fig. 1 is a control block diagram of the entire elevator group management system according to an embodiment of the present invention.

Fig. 2 is an explanatory diagram of an elevator hall call (future call) that is likely to occur in the future.

Fig. 3 is a flowchart of a process of calculating a future call waiting time evaluation value in the elevator group management system according to the embodiment of the present invention.

Fig. 4 is a flow chart of a process for calculating the assigned failure zone for an assigned elevator hall call that may be a long wait in accordance with an embodiment of the present invention.

Fig. 5 is a diagram illustrating calculation processing for assigning a failure region in the embodiment of fig. 4.

Fig. 6 is a flowchart illustrating a calculation process for allocating a failed region in response to the predicted congestion level exceeding a threshold (and the predicted occurrence of full load) according to an embodiment of the present invention.

Fig. 7 is a diagram illustrating calculation processing for assigning a failure region in the embodiment of fig. 6.

FIG. 8 is a flowchart illustrating a process for assigning a failed region in the event that the number of call landings exceeds a threshold value, in accordance with an embodiment of the present invention.

Fig. 9 is a diagram illustrating calculation processing for assigning a failure region in the embodiment of fig. 8.

Fig. 10 is a calculation explanatory diagram for calculating a future call waiting time evaluation value using the assignment failure area in the embodiment of fig. 1 and 3.

Fig. 11 is an explanatory diagram of an example of a calculation result of assigning a failure region shown in fig. 10.

Fig. 12 is an explanatory diagram comparing the predicted waiting times of future call candidate floors in the related art and the embodiment of the present invention.

Fig. 13 is a control block diagram of the entire elevator group management system according to another embodiment of the present invention.

Fig. 14 is a flowchart for calculating the time interval evaluation value in the elevator group management system of fig. 13.

Fig. 15 is an explanatory diagram of calculation of the time interval evaluation value using the allocation failure region calculation in the embodiment of fig. 13 and 14.

Fig. 16 is an explanatory diagram of an example of the calculation result of assigning the failure region shown in fig. 15.

Fig. 17 is a diagram illustrating changes in the charge zone of each elevator and the charge zone after the assignment failure area determination at the time of the temporary setting shown in fig. 16.

Fig. 18 is an explanatory diagram of a basic concept of time interval evaluation represented by a loop.

Fig. 19 is a control example explanatory diagram of an elevator group management system according to an embodiment of the present invention.

Description of the symbols

1 group management control part

21A-21C single elevator control device

22A-22C elevator car

Floor-to-floor registration device in 23A-23C car

24A-24C load sensor

3A, 3B elevator hall calling login device

4 elevator lobby's layer login device that goes to

5A-5C people flow sensor

100 data storage section

101 predicted arrival time table calculating part

102 predicted latency calculating section

103 predicted congestion degree calculating section in car

104 call landing number calculating part

105 assigning a dead zone calculation section

106 will call the candidate floor selection part

107 allocation of active individual elevator selection parts

108 future call waiting time evaluation value calculating section

109 assigned call and newly assigned call assignment evaluation value calculation section

110 comprehensive evaluation value calculating section

111 individual elevator selection sections are assigned.

Detailed Description

Embodiments of the present invention will be described below with reference to the drawings.

First, the operation characteristics of the elevator group management system according to the embodiment of the present invention will be described in comparison with the prior art with reference to fig. 19.

Fig. 19(a) shows an elevator group management system including 3 elevators, in which the position states of 3 elevator cars are shown in a loop form. In the loop system, it is shown that the elevator starts an ascending operation (UP) from the lowest floor to reach the highest floor, and then returns to the lowest floor by a descending operation (DN) for 1 week, and the positional relationship between the elevators managed as a group can be easily understood.

In fig. 19(a), the car No. 1M 01, the car No. 2M 02, and the car No. 3M 03 are respectively in the illustrated positions on the loop. Here, it is assumed that a new elevator hall call M04 occurs in the descending direction. At this time, the car No. 3M 03 is an elevator that can provide service for the new call M04 at the earliest, but in consideration of the quality of service for elevator hall calls that may occur in the future, it is not always correct to use the car No. 3 for assignment. When the section between the cars is assumed to be a section for a future call by the car, the section in which the car No. 3 is in charge is already long, and if the car No. 3 is assigned to a new call, the call received by the car No. 3 may be waiting for a long time. That is, if the zone M05 for the car No. 1, the zone M06 for the car No. 2, and the zone M07 for the car No. 3 are compared, it can be known that the zone for the car No. 3 is already considerably long. Therefore, if car No. 3 is assigned to a new call, the zone for which car No. 3 is responsible will become longer, and after car No. 3 is delayed, the zone between it and car No. 1 will become further longer. For this reason, a long waiting time may occur for a future call, for example, the future call M08, which occurs in the zone.

In the prior art, when the case of fig. 19(a) occurs, the car No. 2M 02 is assigned to the newly occurring call, taking into account the degree of deviation between the predicted waiting time for the newly occurring call and the time interval between 3 elevator cars. This can improve the service quality of the car 3M 03 for the future call M08.

However, although the premise of this prior art is that the car No. 3 can serve all floors in the in-charge zone M07, the actual situation does not necessarily occur as it is supposed to be. An example thereof is the case shown in fig. 19 (B). In fig. 19(B), there is an elevator hall call M10 already assigned to car No. 3 on the lowest floor, and since the number of passengers on that floor is large, it can be predicted that the elevator car will be in a full state in the floor zone M11 indicated by diagonal lines. The number of passengers can be predicted from the past usage state of the building and from a waiting passenger sensor (number of passengers detected by a camera or the like) in the elevator lobby, and if the system is a group management system that performs entry to the elevator lobby on the floor, the number of passengers can be predicted from the number of entries.

As shown in fig. 19(B), it is expected that car No. 3 will be in a congested fully loaded state in a zone M11 indicated by oblique lines, and car No. 3 cannot be assigned to a future call occurring in the zone. Therefore, the car No. 2M 02 becomes a candidate car for a future call in the diagonal section, for example, the future call M08, and the quality of service for the future call is further deteriorated when the car No. 2 is assigned to the newly generated call M04.

In the elevator group management system according to a preferred embodiment of the present invention, when the situation shown in fig. 19(B) occurs, the arrangement shown in fig. 19(C) is made for the zone in charge of each car. Fig. 19(a) differs from (C) in that the zone M31 in charge of the elevator car No. 2 is added and the zone in charge of the car No. 3 is divided into 2 zones M30 and M32, because consideration is given to the congested loading zone of the car No. 3. The service quality for future calls is evaluated based on the charge section of each car, and the car assigned to the newly generated call M04 is determined based on the service quality and the evaluation of the waiting time of the newly generated call.

In the case shown in fig. 19(C), if car No. 2 is assigned to the car of the newly occurring call M04, the quality of service of, for example, M08 for future calls of the zone M31 for which car No. 2 is responsible will deteriorate. In particular, since car No. 2 is located far from the zone, such future calls may be on hold for a long time. In this regard, if car number 3 is used for the assignment, the quality of service of future call M08 will not deteriorate. In the future call of the section M32 in which the car 3 is responsible, for example, the service quality of the section M40 may deteriorate, but since the car 3 does not provide service to the section M31, the time required to reach the section M32 in which the car is responsible can be further shortened. In addition, the section M31 does not have a large deterioration due to a small number of floors or the like. As described above, in fig. 19(C), which is an operation example of an embodiment of the present invention, it is determined that car No. 3 is assigned to a newly generated call.

As described above, an embodiment of the present invention is characterized in that, when evaluating the assigned car of a newly generated call in consideration of a future call, the cars are classified into a car assignable to an assumed future call and a car not assignable to an assumed future call, and the quality of service of the assigned car for the future call is evaluated. This enables more appropriate evaluation of the quality of service for future calls. For example, in fig. 19(B) showing the prior art, since it is not considered that the car No. 3 is fully loaded at a future time point, the car No. 2 is assigned to a newly generated call, but if a call occurs in the slant line section M11, the service quality of the call is deteriorated. At this time, since the 3 rd car is in a crowded full load state, the 2 nd car is responsible for providing service. In contrast, in fig. 19(C) showing an embodiment of the present invention, the car in charge of the future call is reset in consideration of the full loading state in which the car No. 3 may be crowded, so that the service quality of the future call can be more appropriately evaluated. As a result, in this example, the car No. 3 is determined as the assigned car.

An elevator group management system according to an embodiment of the present invention will be described in detail below with reference to fig. 1 and the like.

Fig. 1 is a control block diagram of the entire elevator group management system according to an embodiment of the present invention. The operation of the K elevator cars 22A to 22C is controlled by individual elevator control devices 21A to 21C of the elevators, and the group management control section 1 performs integrated control of the individual car control devices. Further, the hall call signal and the going-to floor signal inputted from the call registering devices 3A and 3B and the going-to floor registering device 4 provided in the elevator lobbies on the respective floors of the building are transmitted to the group management control section 1. Here, the elevator hall call registering devices 3A and 3B are devices for calling an elevator by buttons in the vertical direction in the related art, and the going-to-floor registering device 4 is a device for calling an elevator by inputting a going-to-floor at an elevator hall by numeric keys or the like. In the following section, all calls are collectively referred to as elevator hall calls. The traffic sensors 5A to 5C are sensors provided on the respective floors for detecting the traffic of people, and information detected by these sensors includes, for example, information of people on and off duty, information of people away from the desk, person detection information detected by infrared rays or the like, and traffic information obtained by recognition processing of camera images. Information on the flow of people for each floor and each area detected by the sensors is also transmitted to the group management control section 1. Each of the elevator cars 21A to 21C is provided with a going floor registration device 23A to 23C in the car and a load sensor 24A to 24C in the car. Information on the floor to which the elevator is going, information on the load condition (corresponding to the number of passengers) in the car, and information on the load change (corresponding to the number of passengers getting on and off the elevator) are detected, and the detected information is transmitted to the group management control section 1 via each individual elevator control device. Information on the operating state such as the position, direction, speed, etc. of each elevator is transmitted to the group control section 1 by the individual elevator control device.

Hereinafter, the group management control section 1 will be explained. The basic operation flow of the group management control section 1 is to evaluate K elevator cars for newly generated hall calls based on the obtained pieces of information and an assignment evaluation function, select the most appropriate car, and assign the most appropriate car to the hall call. The information may be detected information of each elevator car, information of each floor elevator lobby, information indicating the flow of people in the building, that is, the flow of people between floors, information obtained by counting the past information, and the like.

First, the data storage unit 100 stores various data such as the state of each elevator, the state of each hall of each floor, the traffic flow amount or traffic flow state of people in the building, and the traffic flow state of people of each floor, which are collected from the individual elevator control devices, hall call registration devices, and the like, in a storage device such as an internal memory. The predicted arrival time table calculation section 101 calculates the predicted arrival time of each elevator at each floor and in the direction (e.g., 3-floor ascending direction). The predicted waiting time of each hall call accepted by each elevator (but not yet served) is calculated in a calculation section 102 for calculating the predicted waiting time of the assigned call based on the data of the predicted arrival schedule of each elevator. The predicted waiting time can be calculated from the sum of the elapsed time from the point in time when the call occurred or the passenger arrived and the predicted arrival time. The predicted in-car congestion degree calculation unit 103 calculates the degree of congestion in the car when the car arrives at each floor (each direction) or when the car departs from each floor (each direction) from the position and direction of the current time point of each elevator. Here, the degree of congestion of each car on each floor (each direction) is calculated from the predicted value of the current number of passengers in each car and the number of persons who get on and off the elevator in the process of sequentially serving the accepted elevator hall call or car call. In the call stop count calculation section 104 of each elevator, the number of stops for the accepted call on each floor (each direction) in the process of each elevator serving the accepted call in order from its current position and direction is calculated. Here, the number of stops for received calls means the number of stops for received assigned elevator hall calls and the number of stops for car calls, and when the floor entry device is used, means the sum of the number of stops for received calls.

The assignment failure region calculation section 105 calculates the assignment failure region of each elevator based on the predicted waiting time of the assigned call, the predicted degree of congestion in the car of each floor (each direction) of each elevator, the number of times of call stop of each elevator, and the predicted arrival time table of each elevator, which are calculated as described above. The assignment of a failure zone refers to a zone in two-dimensional coordinates with a time axis (representing the time after the current point in time) and a floor position axis as coordinate axes, to which zone the relevant elevator cannot be assigned to an elevator hall call occurring at each point in time and on each floor within the zone. For example, if it is predicted that the car 3 has received a long-waiting call within a period of 35 seconds after the current time and cannot be assigned to an elevator hall call in the ascending direction of floors 2 to 8, the area surrounded by the period of 35 seconds after the current time and the floors in the ascending direction of floors 2 to 8 becomes the assignment failure area of the car 3. Actually, the shape thereof is similar to the area B07 shown by a thick line in fig. 5(C) described later. The allocation failure area is used as a judgment criterion for the purpose of facilitating the selection operation of the next allocated active elevator (allocable elevator). The specific calculation method for assigning the failure region is described in the following section.

After the assignment failure areas of the elevators are calculated, a selection unit 106 of the group management control unit 1 for selecting a future call candidate floor selects a floor on which a future call is likely to occur (hereinafter referred to as a future call candidate floor), and an elevator assignable to the future call candidate floor is selected by an assignment-enabled individual elevator selection unit 107. Here, the future call candidate floors are represented by time (current time and future time), floor, and direction (described in detail later with reference to fig. 2). For example, if a future call occurs in the ascending direction of floor 3 after 20 seconds, floor 3 after 20 seconds (ascending direction) becomes a future call candidate floor. When an allocation-enabled elevator is selected, it is determined for each elevator whether the future call candidate floor is within the allocation-disabled area of the elevator at the floor and in time. In the above example, since the assignment failure region of car No. 2 includes a floor in the 3 rd floor ascending direction 20 seconds later, it is determined that car No. 2 does not belong to the elevator to which the future call candidate floor is assigned. The assignment of a valid elevator is an elevator that can be determined with certainty at the current time point as being assignable to an elevator hall call occurring at a certain time and a certain floor with respect to a future call candidate floor as an evaluation target. For each future call candidate floor, only the elevators available for allocation are selected and the quality of service of the elevators is evaluated, so that the evaluation can be performed more accurately. For example, in the prior art, it is assumed that a future call is serviced by an elevator whose predicted arrival time is the smallest, but if the elevator is actually not available for allocation because it has received a call waiting for a long time, there is a possibility that the evaluation value will vary greatly, and the service quality of the future call will deteriorate. In contrast, in the present embodiment, since the elevators available for assignment are selected (classified) by the assignment failure region calculation process and the assignment-enabled elevator selection process as described above, only the cars that can be assigned can be accurately selected, and an appropriate evaluation value can be calculated.

In the future call waiting time evaluation value calculation section 108, first, for each of the selected future call candidate floors, the predicted waiting time when the corresponding allocated valid elevator serves the call is calculated, and the future call waiting time evaluation value for each floor is calculated. Then, the evaluation values of the floors are integrated to calculate the evaluation value of the whole of the future call candidate floor. This value is the future call latency evaluation value. For example, for a given future call candidate floor, the minimum value of the predicted waiting times of the respective assigned elevators may be used as the future call waiting time evaluation value of the floor, or the average value of the first 2 may be used in the order from short to long. With such an evaluation value, the waiting time service index of each allocated valid elevator can be evaluated for the entire future call candidate floor.

In a calculation section 109 for calculating the assignment evaluation values of the assigned calls and the newly assigned calls, the predicted waiting times of the assigned hall calls and the newly generated hall calls (hereinafter, collectively referred to as "actual calls") are calculated, and the assignment evaluation value of the actual call obtained by integrating the calls is calculated. The assignment evaluation value may be calculated from the maximum value, the average value, the square average value, or the like of the predicted waiting time of the actual call of each elevator.

The future call evaluation value calculated by 108 and the actual call evaluation value calculated by 109 are weighted and added by a total evaluation value calculation section 110 to calculate a total evaluation value, and the elevator car having the best total evaluation value is selected as the assigned car by an assigned individual elevator selection section 111.

Hereinafter, an elevator hall call (hereinafter, referred to as a "future call") that may occur in the future will be described in detail with reference to fig. 2. In fig. 2, the horizontal axis a2 is a time axis indicating the time after the current time point with the current time point (t0) as the origin, and the vertical axis a1 is an axis indicating the floor position. The future calls are distributed within a two-dimensional area represented by the time and floor position after the current point in time. As described above, there are many possibilities in terms of time, floor, and direction, and therefore the number of future calls to be considered is enormous. For example, in fig. 2, each of 6 floors has 9 time intervals (1 time interval is about 10 seconds) from t1 to t9, and therefore, it is necessary to consider (6-1) × 2 × 9 — 90 future calls. A building managed by an elevator group is generally a building of 10 floors or more, and a target time zone is generally 2 minutes or more, so that the number of future calls to be considered is extremely large.

In the following cases, the occurrence floor and the time of the future call can be narrowed down to the corresponding future call area. The above case refers to a case where the traffic of people is concentrated on a certain floor (future call area a5 in fig. 2), a case where future calls that occur after a predetermined time (future call area a4 in fig. 2) are considered in aggregate, and a case where it is possible to detect by a traffic sensor which floor a passenger appears after a predetermined time (future call area a6 in fig. 2). The symbols shown by the triangles in the respective future call areas in fig. 2 indicate directional future calls.

Even if the area where future calls are likely to occur is narrowed down, it is still necessary to consider a plurality of future calls, and the amount of calculation for performing assignment evaluation increases. In the assignment evaluation, it is necessary to calculate a future call evaluation value when each elevator is temporarily assigned to each newly generated call, and to repeat the calculation a number of times corresponding to the number of elevators.

When assigning an elevator to a newly-generated elevator hall call, evaluation of the quality of service of future calls is important, but as shown in fig. 2, the number of future calls to be considered is large, and the influence thereof needs to be evaluated for each elevator, so the amount of calculation increases. For this reason, in the known future call evaluation, in order to simplify the calculation, it is assumed that the call is served by the elevator whose predicted arrival time is the shortest (the elevator whose arrival time is the earliest), and the quality of service for the call is evaluated under this condition.

However, in practice, when such a future call occurs, the elevator that provides service to the call is determined based on the assignment evaluation function. Therefore, the evaluation error may be increased to deteriorate the service quality for the future call, if the evaluation value does not match the service quality evaluation value obtained by the evaluation of the future call. On the other hand, when an elevator to be allocated to a future call is determined by an evaluation function, a large number of calculations are required, and real-time processing is difficult. For example, when the number of elevators is 8 and the number of future calls to be considered is 10, the allocation evaluation value to be calculated is 640, which is the temporary allocation number (8) of newly generated calls x the temporary allocation number (10) of future calls to be considered x the temporary allocation number (8) of 1 future call, and the amount of calculation becomes very large.

The present invention is directed to solving the problems presented in the future call evaluation described above. That is, the problem that the calculation amount increases due to the consideration of a large number of future calls or the calculation amount decreases so that the high-precision and accurate evaluation cannot be performed is solved, and the assignment failure area calculation section 105 and the assignment effective individual elevator selection section 107 for selecting the future call candidate floors in each elevator in the embodiment shown in fig. 1 are essential to solve the problem. The elevator which can be appropriately allocated to each future call is selected by the processing, and the quality of service such as the waiting time of the call is evaluated for the selected elevator. When selecting an elevator that can be used for assignment, it is determined whether or not the elevator can be assigned to a future call that may occur on a floor at a certain point in time, using data such as a call waiting for a long time among the assigned calls, the degree of congestion in each elevator car, and the number of times each elevator calls are placed. As mentioned above, one of the essential points is that the elevators are first classified into elevators which can be allocated substantially exactly to future calls which may occur on a certain floor at a certain point in time and which cannot be allocated to such future calls. In the prior art, since the elevator with the smallest predicted arrival time is obtained from all elevators including the elevator which cannot be used for allocation due to occurrence of long waiting time, full load and the like, and the quality of service for future calls is evaluated, if the elevator with the smallest predicted arrival time cannot actually serve the call, the long waiting time is caused as a result. In this embodiment, the above-described result can be avoided by assigning a selection process of the elevator.

Also the idea of classifying elevators according to allocable or not to future calls illustrated in fig. 1 can be applied to any type of future call. For example, the present invention can be applied to a set of future calls that may occur on each floor after a predetermined time (future call area a4 in fig. 2), a case where future calls are likely to occur on a certain floor (future call area a5 in fig. 2), a case where a person traffic sensor detects that future calls are likely to occur on a certain floor after a predetermined time (future call area a6 in fig. 2), and the like, as shown in fig. 2. The setting of the future call candidate floors is performed by the future call selection part 106 of the candidate of fig. 1.

Fig. 3 is a flowchart of a process of calculating a future call waiting time evaluation value in the elevator group management system according to the embodiment of the present invention. The flow of this process will be described below.

First, a future call candidate floor (indicated by 3 elements of time, floor, and direction) to be evaluated is set (ST 501). Then, the variable K indicating the elevator number is set to an initial value K of 1(ST 502). Then, it is determined whether the future call candidate floor to be evaluated is in the assignment failure area of the K-th elevator (ST503), and if the answer is negative (if the floor is not in the assignment failure area of the K-th elevator), the K-th elevator is added as the assignable elevator of the future call candidate floor to be evaluated (ST 504). The above-described processing is repeatedly performed until the processing is performed for all the elevators (ST505 and ST 506). In this way, by selecting (selecting) an elevator whose assignment failure region does not include a future call candidate floor as an evaluation target, an elevator that can be used for assignment can be selected. After all the assignable elevators are selected, the future call evaluation value of the candidate floor is calculated based on the predicted arrival time of the future call candidate floor to be evaluated next, the arrival time of each assignable elevator. As described above, the minimum value of the predicted waiting times of the respective allocated elevators may be used as the future call waiting time evaluation value of the floor, or the average value of the first 2 calls may be used in the order from short to long. The process is repeated for each floor (ST508, ST509) until the process is completed for all future call candidate floors to be evaluated. A total future call evaluation value is calculated from the respective future call evaluation values of all the future call candidate floors to be evaluated (ST 510). The evaluation value may be calculated from, for example, an average value or a weighted average value of the evaluation values of the future call candidate floors.

The embodiment of the present invention shown in fig. 1 and 3 is characterized in that the elevators allocatable to the future call candidate floors are selected by the classification processing, and a method of calculating the allocation failure area in advance is adopted in order to easily classify the allocatable elevators. A specific calculation method of assigning the failure region is described below with reference to fig. 4 to 9.

First, a method of calculating an allocation failure area of an elevator hall call which has been allocated and is to be on standby for a long time will be described with reference to fig. 4 and 5. Fig. 4 shows a flowchart of the calculation method, and the following description is made in order. First, the target elevator is set (ST 001). Then, the predicted waiting time of the accepted elevator hall call (call not yet provided with service) of the target elevator is calculated (ST 002). Then, whether or not a long waiting time is predicted is checked based on the predicted waiting time of each call (ST 003). For example, when the predicted waiting time exceeds 60 seconds, it is determined that it is a long waiting time. When it is detected that the long waiting time will occur, the predicted arrival time of each floor (each direction) from the current position to the long waiting call floor of the target elevator is calculated (ST 004). The area (the area shown in the two-dimensional coordinates formed by the floor position axis and the time axis) surrounded by each floor from the current position to the long-waiting call floor and the predicted arrival time of each floor is determined as the allocation failure area of the target elevator caused by the long-waiting call (ST 005).

Fig. 5 shows an example of calculation of allocation of a failure area caused by a long waiting call. Fig. 5a shows the car position of the target elevator, i.e., the K-th elevator, and the floor where an accepted elevator hall call (elevator hall call not providing service) occurs. The accepted elevator hall call is a call in the 5 th floor ascending direction (B01), and the call is predicted to be a long-waiting call (the predicted waiting time is 60 seconds or more, for example). The elevator number K is currently located at the position in the 2 nd floor descent direction. The predicted arrival time table of each floor (each direction) of the K-th elevator at this time is shown in fig. 5B. For example, the time when the K elevator reaches the 2 nd floor ascending direction is t5 time from the current time point. The predicted arrival time from the 1 st floor ascending direction to the 2 nd floor ascending direction is t1 to tt5, and the estimated parking time is large because the probability of parking in the 1 st floor ascending direction is large and the estimated parking time is long. Further, the predicted arrival time may be calculated by adding the moving time between floors and the estimated stop time value of each floor as a supplement.

Fig. 5(C) is an exemplary diagram of the allocation failure region obtained from the above situation according to the flowchart of fig. 4. A thick line region B07 on a two-dimensional coordinate plane (B06) composed of the time axis B04 and the floor position axis B05 in fig. 5(C) is the allocation disabled region shown in fig. 5. The above-described allocation failure region is characterized by being expressed by time (time after the current time point) and floor position. When the elevator hall call occurring in the area is served by the elevator K, the long waiting time for the assigned call in the 5 th floor ascending direction is further increased, so that the value of the assigned evaluation function of the elevator K is not good when the call actually occurs, and the possibility that the elevator K will become an assigned elevator is low. Therefore, by calculating the assignment failure region and excluding the elevator cars that are not available for assignment based on the assignment failure region (selecting only the elevators available for assignment), it is possible to evaluate the future call more accurately for the quality of service. In addition, the assignment failure area of each elevator is calculated in advance, and it is determined whether each of the future call candidate floors to be evaluated is included in the assignment failure area, whereby an elevator to be effectively assigned can be calculated easily. As shown in the flowchart of fig. 4, the calculation of the assignment failure region itself can be obtained by a simple process.

Fig. 6 and 7 show an example of calculation of assignment of a failure region that occurs because the degree of congestion is predicted to exceed a threshold value (it is predicted that a full load state will occur). Fig. 6 is a flowchart showing calculation of the assignment failure region that occurs when the degree of congestion is predicted to exceed the threshold value (when the full load is predicted to occur), and the following description will be given of the procedure. First, a target elevator is set (ST101), and the predicted intra-car congestion degree of each floor (each direction) at 1 round of travel is calculated for the target elevator (ST 102). In this calculation process, the number of passengers and the number of passengers getting on and off the elevator in the process of serving the accepted elevator hall call and car call in sequence by the target elevator are predicted, and the degree of congestion in the car is calculated. The number of passengers can be estimated from the number of waiting passengers in the elevator hall and the learning data of the number of passengers on the floor in the past, and if the elevator hall registers to the floor, the number of times of registering to the floor can be estimated. The number of persons getting off the elevator can be estimated from the number of persons sitting in the car, the car call floor registered at that time, and the learning data of the number of persons getting off the elevator in the past, and if the persons go to the elevator hall are registered, the number of times of registration of each person going to the elevator hall can be predicted. The degree of congestion in the car is defined as a value obtained by dividing the load in the car by the rated load amount, or a value obtained by dividing the number of passengers in the car by the number of rated passengers in the car. In addition, the degree of congestion in the car may be replaced with the load in the car, the number of passengers in the car, or the like. After the predicted degree of congestion in the car is calculated for each floor (each direction) for 1 week, it is checked whether or not a full load situation has occurred (for example, the predicted degree of congestion in the car is 80% or more) for each floor (each direction) (ST 103). When it is predicted that a full state will occur, the predicted arrival time of each floor (each direction) from the current position to the floor from which the full state is released is calculated for the subject elevator (ST 104). Then, an area (an area shown in a two-dimensional coordinate composed of a floor position axis and a time axis) surrounded by each floor from the current position to the floor from which the full state is released and the predicted arrival time to reach each floor is determined as an allocation failure area of the target elevator, which occurs when the congestion degree exceeds the threshold value (ST 105).

Fig. 7 shows an example of calculation of assignment of a failure region that occurs because the degree of congestion is predicted to exceed a threshold value (full load is predicted to occur). Fig. 7(a) shows the car position of the K-th elevator as the evaluation target, and the assigned hall call floor and car call floor. The assigned hall call is a call in the 5 th floor down direction (C02), and it is predicted that the car will be fully loaded after servicing the call. The car call is a call in the 2 nd floor descending direction (C03), and since a person gets off the elevator at the floor, it is predicted that the congested full load situation is released at the floor. In addition, the K elevator is currently in a 6-th floor-down direction position. The predicted degree of congestion in the car at each floor (each direction) of the K-th elevator at this time is shown in fig. 7B (the unit of the number in the figure is%). The congestion degree in the 6 th floor descending direction, which is the position at the current time point, is 50%, and since there are passengers who take the elevator car when the elevator hall call in the 5 th floor descending direction is served, the congestion degree becomes 80% (C04). Then, this situation continues until the car reaches the 2 nd floor down direction, and the passenger who has made a car call in the 2 nd floor down direction will be in the 2 nd floor down elevator, so the congestion degree is reduced to 30%. And it is predicted that all people get off the elevator to become 0% after 1 floor.

Fig. 7(C) is an exemplary view of the assignment failure region obtained according to the flowchart of fig. 6 for the above-described case. A thick line region C07 on the two-dimensional coordinate plane (C06) formed by the time axis and the floor position axis in fig. 7C becomes an allocation failure region. For example, during the time from the current time point t0 to t1, the K number elevator cannot be used for assignment for future hall calls in the ascending direction that may occur at 3-5 floors, and during the time from t1 to t5, the K number elevator cannot be used for assignment for future hall calls in the ascending direction that may occur at 3-4 floors. In such an assignment failure region, even when an elevator hall call actually occurs in the region, since the K-number elevator is in a full state, the possibility of its assignment as an assigned car is low. Therefore, by predicting the degree of congestion of the elevator cars of each elevator over time, calculating the assignment failure region, and excluding the elevator cars that are not available for assignment from the failure region (selecting only the elevators that are available for assignment), it is possible to evaluate the future call more accurately for quality of service.

Fig. 8 and 9 show calculation examples of allocation failure regions that occur when the number of stops exceeds a threshold value. The process is particularly suitable for floor-to-floor logging group management in elevator lobbies. In the case of the group management of the landing-to-floor registration type, since the landing-to-floor is performed in the elevator lobby, it is possible to know at the time of assignment which floor the elevator is stopped by the passenger (known from the landing floor called by the landing-to-floor), and which floor the elevator is stopped by the passenger (known from the landing-to-floor called by the landing-to-floor). The sum of the two is the number of stops for the layer call. In the landing-ahead group management, even if a call occurs on the same floor, a classification control for allocating elevators is performed according to the landing-ahead, and in this case, elevators whose number of landing-ahead exceeds a predetermined value are controlled so as not to be allocated to a newly-occurring call, thereby achieving the purpose of reducing the number of stops of each elevator (as a result, 1-week travel time is shortened). The allocation failure region here is obtained by extracting a region that is not available for allocation that occurs because the number of stops exceeds a threshold value.

Fig. 8 is a flowchart showing calculation of the allocation failure area caused by the number of stops exceeding the threshold value, and the procedure thereof will be described below. First, the target elevator is set (ST201), and the number of stops of the target elevator due to the received call (forward floor call) on each floor (each direction) during 1 week of travel is calculated (ST 202). For the number of stops due to the received call on each floor (each direction), it is checked whether or not a case exceeding a threshold value (for example, whether or not the number of stops due to the received call is > 4) occurs (ST 203). When the threshold value is exceeded, the predicted arrival time at each floor (each direction) from the current position of the target elevator to the floor at which the state exceeding the threshold value is released is calculated (ST 204). An area (an area shown on a two-dimensional coordinate composed of a floor position axis and a time axis) surrounded by each floor from the current position to the floor from which the state exceeding the threshold is released and the predicted arrival time at the floor is determined as an allocation failure area of the target elevator (ST 205).

Fig. 9 shows an example of calculation of allocation of a dead zone that occurs when the number of stops exceeds the threshold value. Fig. 9(a) shows the car position of the K-th elevator as the evaluation target, the registration floor on which the assigned travel floor call has been made, and the travel floor. The floor on which the assigned call to the floor is registered is a call in the ascending direction of floor 1, and the floor to be registered is floor 3, floor 4 or floor 6. In addition, the K elevator is currently in a 4-th floor-down direction position. The assigned failure region of the K-number elevator at each floor (each direction) at this time is shown in fig. 9 (B). In the 4-story descending direction, which is the position at the current time point, the total number of stops (predicted value) that occur due to the call to the floor in the 1-story ascending direction is 4. When arriving at floor 1, the call stops are reduced by 1, so the number of stops remaining is 3. In addition, after arriving at floor 3, the number of call stops is reduced by 1, and the remaining number of stops becomes 2, and after arriving at floor 4, the number of call stops is reduced by 1, and the remaining number of stops becomes 1. Then, the final travel-to floor 6, known at the current point in time, is reached, after which the remaining number of stops at the floor is zero.

Fig. 9(C) is an exemplary view of the assignment failure region obtained according to the flowchart of fig. 8 for the above-described case. In fig. 9C, 2 bold line regions D03 and D04 on a two-dimensional coordinate plane (D02) including a time axis and a floor position axis are assigned failure regions. Even if an elevator hall call actually occurs in the area, the K-th elevator has a low possibility of being used as an allocated elevator because the number of stops of the K-th elevator due to the accepted call to the floor exceeds the threshold. Therefore, by predicting the number of times each elevator stops due to the received call with time, calculating the assignment-disabled area, and excluding the elevator cars that are not available for assignment from the assignment-disabled area (selecting only the elevators that are available for assignment), it is possible to evaluate the future call more accurately for quality of service.

The method of generating the assignment failure region, which occurs when 3 cases, that is, the long waiting time of the assigned elevator hall call, the predicted congestion degree in the elevator car exceeds the threshold value, and the number of call stops exceeds the threshold value, have been described above. However, since some of the allocation failure areas are generated by the simultaneous occurrence of a plurality of the above cases, the allocation failure areas may be generated for each of the 3 types, and then the areas where the areas overlap (areas obtained by OR operation) may be used as the allocation failure areas. In this case, since many factors that cannot be allocated can be considered, the evaluation accuracy of future calls can be further improved. Of course, the allocation failure region may be used alone, or 2 of them may be used in an overlapping manner.

Fig. 10 shows an example of calculation of the future call waiting time evaluation value when evaluation is performed using the assignment failure area based on the embodiment described in fig. 1 and 3. Here, the elevator group management of 3 elevators will be described as an example. Fig. 10 a shows a position state of 3 cars at the current time point (shown in a loop form (this is described later with reference to fig. 18)). In the figure, the vertical direction indicates the floor position, and the horizontal direction indicates the position of each car in the ascending direction and the descending direction, respectively. In addition, the numbers in the square indicating the car indicate the numbers of the elevators.

Fig. 10(B) shows the predicted waiting time of each future call candidate floor in the state shown in fig. 10 (a). Here, the set of future call candidate floors to be evaluated is a set of all floors and directions after a predetermined time (corresponding to the area a4 in fig. 2). In fig. 10B, the horizontal axis represents time, the vertical axis represents floor position, and the origin of the time axis represents the current time point (current time). The positions of the car No. 1 (H01), the car No. 2 (H02), and the car No. 3 (H03) are shown in the figure, respectively, and correspond to the position state of fig. 10 (a). In this case, the future call waiting time evaluation value is obtained using all the floors and directions at the time point tf (line H06) as future call candidate floors. Here, the predicted waiting time of a future call indicated by a triangle symbol (for example, one of the triangles is the triangle H07) at the time tf is obtained as an example. First, a predicted arrival time table (obtained by plotting predicted arrival times of respective floors (respective directions)) of respective elevators is calculated. The predicted arrival time data at each floor (each direction) in the predicted arrival time table is shown on the coordinates of fig. 10B, and for example, when the data of car No. 1 (H01) is shown on the coordinates, the predicted trajectory indicated by curve H04 is obtained. Likewise, when the data of car No. 2 (H02) is shown on the coordinates, the predicted trajectory represented by curve H05 is obtained. The predicted trajectories correspond to curves obtained by connecting the positions (predicted positions) where the elevators are located at respective points in time.

The assigned failure zone for each elevator is calculated as follows.

Fig. 11 shows an assigned failure region of car No. 2 of the present embodiment. The assignment failure region of car No. 2 is a region J01 enclosed by a thick line on a two-dimensional coordinate consisting of a time axis and a floor position axis. This zone is calculated, for example, from the hall call of the elevator which is assigned and predicted to be a long-waiting call received by car No. 2.

As a result of considering the assignment failure region of car No. 2 shown in fig. 11, the predicted waiting time of the future call candidate floor shown by the triangle in fig. 10(B) is as shown by the arrow line in the figure. For example, the predicted waiting time (predicted arrival time of the service car) of the future call candidate floor (H07) is expressed by the length of the segment of the arrowed line (H08). The predicted waiting time of each future call candidate floor is determined by the predicted arrival time of the car that arrives earliest, but some future call candidate floors may differ in consideration of the assignment failure area. Among the 4 future call candidate floors indicated by the reference character H10, the car that reaches the call floor first is the car No. 2, but the call floor is within the assignment failure region of the car No. 2 (the assignment failure region of the car No. 2 is a region surrounded by the broken line H09 and the line H05 of the predicted trajectory from the current position (H02) of the car No. 2). Therefore, the car that arrives first in the allocable elevator is car No. 1, and therefore the predicted waiting time is determined by car No. 1.

Fig. 12(a) and (B) are diagrams comparing predicted waiting times of future call candidate floors in the related art and the embodiment of the present invention, respectively. The main difference is the predicted wait time for the set of future call candidate floors, denoted by the symbol H10. In the conventional example of fig. 12(a), the predicted waiting time is simply determined by car No. 2 which can arrive at the call floor at the earliest. In contrast, in the embodiment of the present invention shown in fig. 12(B), since the car available for allocation is selected first and then the predicted waiting time is determined based on the allocated car selected, the predicted waiting time of the future call candidate floor indicated by reference character H10 is determined based on car No. 1.

As a result, by comparing the future call candidate floors indicated by reference character H10, it is known that, in the prior art, the estimated values of the predicted waiting times of the future call candidate floors are estimated to be excessively short as indicated by arrow H12 in fig. 12 (a). On the other hand, in the embodiment of the present invention, as shown by arrow H11 in fig. 12(B), the predicted waiting time can be estimated more accurately from car No. 1 that can be actually used for assignment. For example, in the evaluation of the conventional technique, the estimated waiting time of the future call candidate floor evaluated is shorter than the actual value, and when a future call actually occurs at the location indicated by reference character H10, car No. 2 cannot be assigned to the future call, and therefore, there is a possibility that the call will wait for a long time. In contrast, in the embodiment of the present invention, since the long wait at the call floor can be accurately evaluated, the above-described situation can be avoided when selecting an elevator assigned to a newly-occurring call, and the possibility of occurrence of a long wait can be reduced.

Fig. 13 is a control block diagram of the entire elevator group management system according to another embodiment of the present invention. Fig. 13 shows a case in which the method for evaluating future calls based on the time interval between the cars is applied. As is well known, in order to suppress the occurrence of long waiting time, it is preferable to bring a plurality of elevators into a state of being close to equal intervals.

The same reference numerals are given to the same components as those in fig. 1 in fig. 13, and the description thereof will be omitted. In fig. 13, the sections different from fig. 1 are a service elevator selection section 120, a service floor zone selection section 121, and a time interval evaluation value calculation section 122. The above-described part is used when applying the concept of the invention for classifying (selecting) allocated elevators to the time interval evaluation method of the prior art.

Here, a basic assumption of the time interval evaluation will be described with reference to fig. 18. Fig. 18(a) shows the position and the traveling state of each car in the group management of 3 elevators. The cars No. 1K 01 to No. 3K 03 are in the states shown in FIG. 18 (A). The arrows on the cars indicate the traveling directions of the cars. For example, car No. 1 is in the ascending direction. When the 3-station case is expressed in a loop form, it is as shown in fig. 12 (B). The expression "loop" means that each elevator car is shown on a 1-cycle loop from ascending to descending. Fig. 12(B) shows a state in which the cars No. 1K 01 to No. 3K 03 are in the loop. When the time interval evaluation is performed, the time interval indicating the section between 2 elevators at the front and rear positions among the elevators on the loop is evaluated. If the zone is large, when an elevator hall call occurs in the zone, it may take a certain time until a subsequent elevator arrives at the call floor, so that a long wait is liable to occur. Therefore, in order to suppress the occurrence of long waiting time, it is preferable that the time intervals of the respective sections are set to be equal intervals, and when the time interval evaluation is performed, the evaluation is performed based on an index used for the evaluation, for example, a maximum value of the interval, a variance of the interval, a square sum of the interval, or the like. That is, a set of consecutive floors served by the same elevator is set in a section of the elevator which can serve an elevator hall call, and interval evaluation is performed based on the time required to pass the service-enabled section. This is the interval evaluation based on the time length, but may be simplified to the evaluation based on only the length of the physical interval.

The point of this interval evaluation is that future calls that may occur in the individual car sectors shown on the loop are taken into account and evaluated on the assumption that they are served by the elevators located at the rear of the individual sectors. In this regard, the same concept is used for providing service by an elevator whose predicted arrival time is the smallest.

The following description will proceed with an embodiment of the present invention, with reference back to fig. 13. In the service elevator selection unit 120, for each floor (future call candidate floor) in the section of each elevator shown on the loop, an elevator having the smallest arrival time is selected as a service elevator from among the allocated elevators for each floor. In the service floor zone selection part 121, a service floor zone is selected again on the basis of each selected service elevator. Here, the initial service floor zone is a zone of each elevator shown on the loop, but since the service elevator of each floor changes, it is necessary to newly select a service floor zone according to this processing. In the time interval evaluation value calculation 122, the time interval is calculated for the service floor section after the reselection, and the time interval evaluation value is calculated from the maximum value of the interval, the variance of the interval, the sum of squares of the interval, and the like.

Fig. 14 is a flowchart showing a time interval evaluation value calculation process that is an outline of the configuration example of the elevator group management system described in fig. 13. Hereinafter, a method of predicting the position and direction of each elevator car after a predetermined time to evaluate the time interval of the positional relationship will be discussed. The processing steps will be described below. First, the predicted position and direction of each elevator are calculated from the predicted arrival time table (ST 301). Thereafter, the position order shown on the loop is calculated from the predicted position and direction of each elevator after a predetermined time (ST 302). The interval between 2 elevators at the front and rear positions is determined according to the position order of the elevators (ST 303). For each elevator, a zone located in front of the elevator (in front in the 1-circumference direction shown as a loop) is assumed to be a zone in charge of the elevator (ST 304). Then, it is checked whether or not the floor (direction) and time in the responsible zone are included in the allocation failure area (ST305) for the responsible zone of each elevator. When the floor and direction in the responsible zone are included in the allocation failure area, the floor and direction are added to the responsible zone of the subsequent elevator serving the original elevator by changing (ST 306). After the floors and directions of the responsible zones of all the elevators are checked through the above-described processing (ST307), a time interval evaluation value is calculated from the time intervals of the responsible zones of the elevators that have been changed (ST 308). In step ST306, the floor and direction of the service elevator changed are checked to see whether the floor and direction of the newly set service elevator belong to the allocation failure area, and the process is repeated until the allocated service elevator becomes a service elevator.

Fig. 15, 16, and 17 show an example of calculation of the time interval evaluation value using the allocation failure region according to the embodiment described in fig. 13 and 14. Here, the elevator group management of 3 elevators will be described as an example. Fig. 15 a shows the position state (shown in a loop) of 3 cars at the current time.

Fig. 15(B) shows an example of calculation of a service floor zone of car No. 1E 08 shown in fig. 15 (a). Here, the time interval of each car after the predetermined time tf (time indicated by a line E07) is evaluated. In fig. 15(B), the horizontal axis represents time, the vertical axis represents floor position, and the origin of the time axis represents the current time point (current time). The positions of the cars E01-E03 shown in the figure correspond to the position of FIG. 15 (A). In this case, the time interval evaluation value for the predicted position of each car at the time point tf (line E07) is obtained.

First, a predicted arrival time table (obtained by plotting predicted arrival times of respective floors (respective directions)) of the respective elevators is calculated, and predicted trajectories of the respective elevators are generated on the basis of the calculated predicted arrival time table. For example, when car No. 1E 01 is shown on the coordinates, the predicted trajectory indicated by curve E04 can be obtained. Similarly, when the car No. 2E 02 is shown on the coordinates, the predicted trajectory shown by the curve E05 can be obtained, and when the car No. 3E 03 is shown on the coordinates, the predicted trajectory shown by the curve E06 can be obtained. The predicted trajectories correspond to curves obtained by connecting the positions (predicted positions) where the elevators are located at respective points in time.

Based on the predicted trajectory, the predicted position of each elevator at time tf is determined. The predicted position and direction of the car No. 1 at the time tf are the position and direction indicated by the reference character E08, and the predicted position and direction of the car No. 2 are the position and direction indicated by the reference character E09.

Then, an assignment failure region is calculated for each elevator.

Fig. 16 shows an assigned failure region of car No. 1 in the present embodiment. The assigned failure region of car No. 1 is a region F01 enclosed by a thick line on a two-dimensional coordinate having time and floor position as coordinate axes. In the allocation failure region F01, the section after the time tf is the section E11. Therefore, since car No. 1 is predicted to be in a state of long waiting time, full load, or the like after time tf, this section E11 is an assignment failure region (unavailable for assignment) of car No. 1.

As a result of considering the assignment failure region of the car No. 1 shown in fig. 16, the car No. 1 is changed to the zone E12 as the zone E10 up to the car No. 2 as the service floor zone of the service elevator (at the time of temporary setting). The zone E11 in the assigned failure zone for car No. 1 is excluded from the service floor zones for car No. 1. The subsequent car No. 3 becomes a service elevator in this zone. Thereafter, the time interval is calculated for the section to which the service elevator can be allocated excluding the allocated defective floor, and the time interval evaluation value is calculated based on the time interval, whereby accurate time interval evaluation with higher accuracy can be performed.

In fig. 17, regarding the example of fig. 16, fig. 17(a) shows a zone in charge of each elevator at the time of temporary setting, and fig. 17(B) shows a zone in charge after the allocation failure region determination. Here, the section in charge of each elevator at the time of temporary setting shown in fig. 17(a) also corresponds to the time interval evaluation in the related art. In addition, in the present embodiment, in terms of assigning the failure region, it is assumed that only the car No. 1 has a failure region as shown in fig. 16.

The zone in charge of each elevator at the time of temporary setting shown in fig. 17(a) is a zone G04 for the 1 st car G01, a zone G05 for the 2 nd car G02, and a zone G06 for the 3 rd car G03. In contrast, the responsible zone after the assignment failure region determination shown in fig. 17(B) is a zone G04A for the car No. 1G 01, a zone G05 for the car No. 2G 02, and a zone G06A for the car No. 3G 03. In the car No. 1, since the section E11 in fig. 15, which is an allocation failure area, exists in the zone in charge at the time of temporary setting, the length of the section is shortened as a result of taking this into consideration. On the other hand, since the section in which the car No. 1 cannot be assigned is assigned to the car No. 3, the length of the assigned section of the car No. 3 increases. As a result, in the temporary setting, as shown in fig. 17(a), the intervals of the elevators are relatively close to the equal intervals, but after it is determined whether each elevator can actually provide service to the floors in the respective zones, as shown in fig. 17(B), it is known that a large deviation occurs. The interval in charge when temporarily setting is the same as the interval considered in the related art, and it is known that the time interval cannot be accurately evaluated by the method in the related art. In contrast, in the method provided in the embodiment of the present invention, the floor zones in charge of the assignable elevators are set and the time intervals are obtained, so that accurate time interval evaluation with higher accuracy can be performed.

Finally, as a supplementary explanation, the assumption of the allocation of the valid elevator classification (selection) described above will be further explained. The basic assignment evaluation formula of general group management is as follows.

Assignment evaluation value of waiting time for newly generated call + waiting time for assigned call + degree of congestion in car … … (1)

Each time a new hall call occurs, the evaluation value of each elevator is calculated based on the assignment evaluation formula. Therefore, this formula should be used for the future call candidate floors, but as shown in fig. 2, the future call candidate floors are distributed in a two-dimensional area constituted by a time axis and a floor position axis, and therefore the number thereof is large, and if each elevator is temporarily assigned and evaluated for each future call candidate floor, the amount of computation becomes enormous. Therefore, in the embodiment of the present invention, the waiting time of the assigned call and the degree of congestion in the car are regarded as important points, and it is predicted that the former will have a long waiting time, and when the latter is in a full load state, the elevator is judged to be unallocated and excluded from the objects of service evaluation. That is, in other words, the elevators are classified (selected) into allocable elevators and unallocated elevators. That is, by considering the fact that the elevator cannot be used for allocation based on the assumption of the allocation evaluation value of the formula (1), the amount of calculation required is reduced, and the elevator that can be actually used for allocation is predicted more accurately. In addition, in the conventional technology, in order to reduce the amount of computation, a future call is evaluated by the following evaluation formula, and there is a possibility that accurate evaluation cannot be performed because there is a deviation between an actual value and an evaluation value.

Waiting time for newly generated call … … (2) is assigned rating

Claims (6)

1. An elevator group management system comprising a plurality of elevators for providing services to a plurality of floors and an assignment control device for assigning an appropriate elevator to a newly generated elevator hall call based on an assignment evaluation index, the elevator group management system further comprising:

an extraction unit that extracts an elevator hall call that is likely to occur in the future as a candidate future call;

a classification unit that classifies the plurality of elevators into elevators assignable to the candidate future calls and elevators not assignable to the candidate future calls;

an assignment evaluation value calculation unit that calculates a future call assignment evaluation value when the assignable elevator of a future call assignable to the candidate is assigned to the future call of the candidate;

an allocation evaluation index calculation unit that calculates an allocation evaluation index when allocated to a newly occurring elevator hall call on the basis of referring to the future call allocation evaluation value; and

an assigned elevator decision means for deciding an elevator assigned to the newly generated elevator hall call based on the assignment evaluation index,

the classification means includes means for classifying an elevator whose predicted waiting time for serving the candidate future call is within a predetermined value as an allocable elevator.

2. An elevator group management system having a plurality of elevators for providing services to a plurality of floors and a group management control device for allocating an appropriate elevator to a newly generated elevator hall call based on an allocation evaluation index, the elevator group management system further comprising:

an extraction unit that extracts an elevator hall call that is likely to occur in the future as a candidate future call;

a classification unit that classifies the plurality of elevators into elevators assignable to the candidate future calls and elevators not assignable to the candidate future calls;

an interval evaluation value calculation means for correcting the scheduled service floor areas for the candidate future calls of the elevators determined based on the positional relationship between the elevators based on the result of the classification, and calculating the interval evaluation values of the plurality of elevators based on the corrected scheduled service floor areas;

an allocation evaluation index calculation unit that calculates an allocation evaluation index when allocated to a newly occurring elevator hall call on the basis of the interval evaluation value; and

an assigned elevator decision means for deciding an elevator assigned to the newly generated elevator hall call based on the assignment evaluation index,

the classification means includes means for classifying an elevator whose predicted waiting time for serving the candidate future call is within a predetermined value as an allocable elevator.

3. The elevator group management system according to claim 2, wherein the interval evaluation value calculation unit has a setting unit that sets a set of consecutive floors served by the same elevator as an elevator hall call-serviceable section;

the plurality of elevators have a calculation means for calculating the interval evaluation value based on the time length or distance length of each section in which a service can be provided to an elevator hall call.

4. An allocation control method for an elevator group management system having a plurality of elevators for providing services to a plurality of floors and an allocation control device for allocating an appropriate elevator to a newly generated hall call based on an allocation evaluation index, the allocation control method comprising:

an extraction step of extracting an elevator hall call that is likely to occur in the future as a candidate future call;

a classification step of classifying the plurality of elevators into an elevator that can be allocated to the candidate future call and an elevator that cannot be allocated to the candidate future call;

an assignment evaluation value calculation step of calculating a future call assignment evaluation value when the assignable elevator of the future call assignable to the candidate is assigned to the future call of the candidate;

an assignment evaluation index calculation step of calculating an assignment evaluation index at the time of assignment to a newly generated elevator hall call on the basis of the future call assignment evaluation value; and

an allocated elevator determining step of determining an elevator allocated to the newly generated elevator hall call based on the allocation evaluation index,

the classifying step includes a step of classifying an elevator having a predicted waiting time within a predetermined value when the candidate future call is served as an allocable elevator.

5. An allocation control method for an elevator group management system having a plurality of elevators for providing services to a plurality of floors and a group management control device for allocating an appropriate elevator to a newly generated elevator hall call based on an allocation evaluation index, the allocation control method comprising:

an extraction step of extracting an elevator hall call that is likely to occur in the future as a candidate future call;

a classification step of classifying the plurality of elevators into an elevator that can be allocated to the candidate future call and an elevator that cannot be allocated to the candidate future call;

an interval evaluation value calculation step of correcting the scheduled service floor areas of the elevators determined based on the positional relationship between the elevators for the candidate future calls based on the result of the classification, and calculating the interval evaluation values of the plurality of elevators based on the corrected scheduled service floor areas;

an allocation evaluation index calculation step of calculating an allocation evaluation index when allocated to a newly generated elevator hall call on the basis of the interval evaluation value; and

an allocated elevator determining step of determining an elevator allocated to the newly generated elevator hall call based on the allocation evaluation index,

the classifying step includes a step of classifying an elevator having a predicted waiting time within a predetermined value when the candidate future call is served as an allocable elevator.

6. The allocation control method of an elevator group management system according to claim 5, wherein the interval evaluation value calculation step has a setting step of setting a set of consecutive floors served by the same elevator as a zone of the elevator which can be served by an elevator hall call;

the plurality of elevators have a calculation means for calculating the interval evaluation value based on the time length or distance length of each section in which a service can be provided to an elevator hall call.