DE69220142T2 - Allocation of elevator calls - Google Patents

Description

TECHNICAL AREA

The invention relates to software for elevator control.

STATE OF THE ART

It is desirable to allocate cars to serve hall calls in such a way that the performance of the elevator system is maximized. This requires the use of a number of algorithms that determine the number of passengers in the cars, determine the operating mode of the elevator system, estimate the number of passengers in the hall at each stop, and calculate the trade-offs between various performance parameters. These algorithms can be implemented as several fixed rules.

However, difficulties arise at the boundary conditions for the fixed rules. For example, a rule that underlies the system's operating mode determination, in part, depending on the time of day may be "if it is between 7:00 a.m. and 9:00 a.m. and if < other conditions> exist, then set the system operating mode to UP-PEAK." The difficulty with such a rule is that at 6:59 a.m., all of the other conditions that cause the system operating mode to be UP-PEAK may already be present, but because of the fixed rule, the system cannot be considered to be in the UP-PEAK operating mode. The system's operation may change abruptly depending on the operating mode despite the fact that the input conditions, the prevailing operating patterns, are likely to change gradually between 6:59 a.m. and 7:00 a.m.

Similar problems exist for the other algorithms related to car allocation. In general, the input conditions change gradually and mostly continuously, while the response to these changes, i.e. the system's reactions (and ultimately the allocation of a car to a hall call) change abruptly and discontinuously as soon as inputs to the system exceed boundary conditions.

DESCRIPTION OF THE INVENTION

Objects of the invention include a method for determining the number of cabin passengers according to claim 1 and an apparatus according to claim 6 for carrying out the method according to claim 1, the method comprising the following steps:

Detecting a weight signal as an indicator of the passenger weight;

providing a plurality of fuzzy sets of observed weight, each of which corresponds to a specific number of passengers and wherein for each of the fuzzy sets of observed weight, the membership value of each term corresponds to the frequency of occurrence of a specific value for the magnitude of the weight signal; and

Forming a fuzzy set of the number of passengers as an indicator of the number of cabin passengers, where for each term the membership value is equal to the membership value of a term of a specific one of the fuzzy sets of the observed weight with a basis element corresponding to the size of the weight signal, and where the basis element for each term of the fuzzy set of the number of passengers is an indicator of a number of passengers represented by the one specific fuzzy set of the observed weight.

According to the present invention, a signal which is an indicator of the passenger weight in a cabin is used to calculate terms from selecting a plurality of observed weight fuzzy sets to form a fuzzy set as an indicator of the number of passengers in the cabin, wherein (a) each of the observed weight fuzzy sets corresponds to a specific passenger number, and (b) terms of each of the observed weight fuzzy sets have: (i) basis elements corresponding to the passenger weights, and (ii) membership values corresponding to the frequency of observation of the weight represented by the corresponding basis element and the number of passengers represented by the set.

Furthermore, according to the present invention, the assignment of a car to a hall call is performed using a fuzzy set as an indicator of the convenience of the car assignment, wherein the set comprises (a) basic elements corresponding to each car of an elevator system and (b) membership values corresponding to the convenience of the assignment of the car represented by the corresponding basic element to the hall call.

Further in accordance with the present invention, the utility of assigning each car to serve a hall call is determined by: (a) estimating the performance of each car for each of a plurality of performance criteria, (b) weighing the estimated performances by values indicative of the assigned importance of each of the performance criteria, and (c) setting the utility equal to the maximum value of the weighed performance values.

Further, according to the present invention, the usefulness of assigning each car to a hall call is determined by (a) estimating the performance of each car for each of the plurality of performance criteria, (b) weighing the estimated performances by values indicative of the assigned importance of each of the performance criteria, and (c) selecting the minimum value from the weighted performance values.

Further in accordance with the present invention, (a) an instantaneous passenger rate is calculated for an elevator system whenever a hall call button is pressed or whenever an elevator serves a stop, (b) the instantaneous passenger rates are then averaged with respect to one or more of the following: (i) an up rate magnitude, (ii) a down rate magnitude, and (iii) a rest rate magnitude according to the mode of the elevator system, and (d) the number of people waiting at a stop is determined by multiplying the time since the stop was last served by one or more of the up, down, or rest rates according to the mode of the elevator system.

Furthermore, according to the present invention, rules for the start of up-peak, rules for the end of up-peak, rules for the start of down-peak and rules for the end of down-peak are separately evaluated and combined to form a fuzzy logic set as an indicator of the elevator operating mode, the set having a term corresponding to up-peak, a term corresponding to down-peak and a term corresponding to no-peak, the membership values of the terms corresponding to the values with which the elevator system exhibits characteristics of up-peak operating mode, down-peak operating mode and no-peak operating mode, respectively.

The foregoing and other objects, features and advantages of the present invention will become more apparent in light of the following detailed description of exemplary embodiments thereof, as shown in the accompanying drawings.

BRIEF DESCRIPTION OF THE DRAWINGS

Fig. 1 is a perspective, partially broken away view of an elevator system.

Fig. 2 is a data flow diagram showing the operation of the elevator control software.

Fig. 3 is a graph showing empirically observed elevator weight loading data.

Fig. 4 is a flow chart showing the operation of a weight interpretation software module.

Fig. 5 is a flow chart 120 showing the off-line construction of a weight interpretation table.

Fig. 6 is a data flow diagram showing the operation of an operation module.

Fig. 7 is a flowchart showing the operation of an upward calculation module.

Fig. 8 is a diagram showing a fuzzy logic set <SHORT TIME>.

Fig. 9 is a diagram showing a fuzzy logic set <VARIOUS CABINS>.

Fig. 10 is a diagram showing a fuzzy logic set <HEAVY LOADED>.

Fig. 11 is a data flow diagram showing the operation of a number estimation module.

Fig. 12 is a bar chart showing the fuzzy set of an average waiting time performance.

Fig. 13 is a flowchart showing the operation of a power estimation module.

Fig. 14 is a bar chart showing a fuzzy set of customer preferences.

Fig. 15 is a flowchart showing the operation of an allocation utility calculation module.

BEST MODE FOR CARRYING OUT THE INVENTION

Referring to Fig. 1, an elevator system 20 includes a first car 22, a second car 23, a first motor 24 with a pulley 26 attached thereto, a second motor 25 with a pulley 27 attached thereto, and counterweights 28, 29. A first cable 30 is passed through the pulley 26 and attached at one end to the first car 22 and at the other end to the counterweight 28. A second cable 31 is passed through the pulley 27 and attached at one end to the second car 23 and at the other end to the counterweight 29.

The first motor 24 moves the first car 22 between a plurality of building floors 32 to 34 in response to activation of one or more car call buttons 36 by a car passenger (ie, a passenger riding in car 22). A second motor 25 moves the second car 23 between building floors 32 to 34 in response to activation of one or more car call buttons 37. by a car passenger. The motors 24, 25 also move the cars 22, 23 in response to one or more hall calls between floors. A hall call occurs when one or more hall call buttons 38 are pressed by a hall passenger. A hall passenger is a prospective user of the elevator system 20 who is waiting in a floor corridor on one of the floors 32 to 34 to be served by one of the cars 22, 23.

An electronic elevator controller 40 for the elevator system 20 receives electronic input signals from the car call buttons 36, 37, the hall call buttons 38, from a first weight sensor 42 located in the floor of the first car 22, and from a second weight sensor 43 located in the floor of the second car 23. The weight sensors 42, 43 each provide an electronic signal that varies according to the weight of the passengers in the cars 22, 23, respectively. The controller 40 provides an output signal to the first motor 24 for moving the first car 22 between the various floors 30 to 32. The controller 40 also provides an output signal to the second motor 25 for moving the second car 23 between the various floors 30 to 32.

The electronic hardware for the controller 40 is a conventional microprocessor system, the implementation of which is known to those skilled in the art, and which includes a microprocessor (not shown), one or more ROM memories (not shown) for storing the elevator control software, one or more RAM memories (not shown), means for providing output signals (not shown) to the motors 24, 25, and means for receiving the input signals (not shown) from the car call buttons 36, 37, the hall call buttons 38, and from the weight sensors 42, 43.

Referring to Fig. 2, a data flow diagram 50 shows the operation of the elevator control software stored in the ROM memories stored and executed by the microprocessor. The software causes the controller 40 to provide the output signals to direct the operation of the motors 24, 25 in response to the electronic input signals to the controller 40. The boxes in the diagram 50 indicate program modules (parts of the elevator control software) while cylinders indicate data elements (parts of the elevator control data). Arrows between the boxes and the cylinders indicate the direction of data flow. Unlike a flow chart, no part of the data flow diagram 50 indicates any temporal relationship between the various modules. The input signals shown in the diagram 50 are summarized in the following table:

A weight interpretation module 52 is provided with a WEIGHT signal from each weight sensor 42, 43 and input data from an observed data element 53. The weight interpretation module 52 uses the WEIGHT signals and the observed weight data element 53 to estimate the number of cabin passengers. The passenger estimate is provided by the weight interpretation module 52 to the cabin passenger data element 54. The use of the data element 53 the observed weight and the WEIGHT signals from the weight sensors 42, 43 to estimate the number of cabin passengers is described in detail below.

The car passenger data element 54 is provided as an input to an operation module 56 which is also provided with an input from a DEPARTURE TIME input signal, a DEPARTURES input signal and an ARRIVALS input signal. The DEPARTURE TIME signal indicates the elapsed time (i.e., a counter that increases at a fixed rate). The DEPARTURES signal indicates the departure of the cars 22, 23 from the various floors 32 to 34. The ARRIVALS signal indicates the arrivals of the cars 22, 23 at each floor 32 to 34.

There are three operating mode variables used by the elevator control software: UP-PEAK, DOWN-PEAK, and NO-PEAK. The magnitude of UP-PEAK indicates the extent to which passengers travel from the lobby to higher floors. The magnitude of DOWN-PEAK indicates the extent to which passengers travel from higher floors to the lobby. The NO-PEAK magnitude indicates the extent to which passengers travel between floors not including the lobby. The values of UP-PEAK, DOWN-PEAK, and NO-PEAK are stored by the operating module 56 in the operating mode data element 58. The operation of the operating module 56 is described in detail below.

The DEPARTURE TIME, DEPARTURES and ARRIVALS SIGNALS are provided as inputs to a number estimator module 60. A HALL CALLS signal indicating that one of the hall call buttons 38 has been pressed is also provided as input to the number estimator module 60. The number estimator module 60 also receives inputs from the car passenger data element 54 and the operating mode data element 58. The number estimator module 60 estimates the number of Hall passengers waiting to be served by a car at a specific stop. For each floor except the top and bottom floors, there are two stops, an up stop and a down stop. For the top floor, there is only a single down stop, and for the bottom floor, there is only a single up stop. The number estimator module 60 outputs the estimate to a hall passenger data element 62. The operation of the number estimator module 60 is described in detail below.

A performance estimator module 64 receives inputs from the HALL CALLS signal and from the operating mode data element and hall passenger data element 58, 62. The performance estimator module 64 predicts the performance of each car 22, 23 in response to a particular hall call. The output of the performance estimator module 64 is provided to a performance data element 66. The operation of the performance estimator module 64 is described in detail below.

The performance data element 66 and a customer preference data element 68 are provided as inputs to an assignment convenience calculation module 70, which determines the assignment convenience of each car 22, 23 for responding to a particular hall call by predicting the performance of each car 22, 23 and weighing the predicted performance based on the customer preferences. The output of the assignment convenience calculation module 70 is communicated to an assignment convenience data element 72. The operation of the assignment convenience calculation module 70 is described in detail below.

The assignment convenience data element and the passenger preference data element 68, 72 are provided as inputs to an uncertainty filter module 74 that assigns a particular car to serve a particular hall call using the provisional assignment data element 72. The assignment is made only if the uncertainty associated with the assignment of one of the cars is below a predetermined threshold specified by data stored in customer preference data element 68. In situations where it is acceptable to wait until the last possible moment to assign a car to a hall call, uncertainty filter module 74 does not provide data to assignment data element 76 until the uncertainty associated with assignment convenience data element 72 is quite low. In situations where a car assignment must be made relatively quickly after a hall call, uncertainty filter module 74 provides data to assignment data element 76 even though the uncertainty associated with the assignment is relatively high. The operation of uncertainty filter module 74 is described in detail below.

The assignment data element 76 is provided as an input to the motion control system module 78 which provides signals to the motors 24, 25 to move the cars 22, 23. The motion control system module 78 uses the information of the assignment data element 76 to cause the assigned car 22, 23 to stop at a particular floor 32-34 in response to a hall call. There are many types of typical elevator software motion control systems known to those skilled in the art and currently in use. Many of the typical motion control systems would be suitable for implementing the motion control system module 78.

The weight interpretation module 52 converts the WEIGHT signal from each weight sensor 42, 43, one by one, into an estimate of the number of cabin passengers using fuzzy logic, which is a branch of mathematics closely related to the principles of theory and logic. Fuzzy logic involves using sets of basis elements that are only partially contained in the sets. For example, while a traditional set C may be defined as {X, Y, Z}, a fuzzy set F may be defined as {0.3 X, 0.7 Y, 0.1 Z}, where the numbers preceding the vertical bars indicate the membership value of the basic elements X, Y, and Z. The quantity 0.3 X is called a term of the fuzzy set. The basic elements X, Y, Z may represent numeric and non-numeric quantities. In cases where the basic elements X, Y, and Z represent numbers, the value of a basic element or the value of a term is simply the numeric quantity represented by X, Y, or Z. A sharp value is any value or value system that does not use fuzzy logic. A detailed description of fuzzy logic can be found in Schmucker, KJ, Fuzzy Sets, Natural Language Computations and Risk Analysis, Computer Science Press, Rockville, Maryland, 1984.

Although the description below explains the implementation of details of the fuzzy system's operation, much of the implementation can be automated by tools that translate high-level fuzzy logic statements into compilable computer language. One such development tool is the Togai Fuzzy C Development System, manufactured by Togai InfraLogic Inc., Irvine, California, which translates fuzzy logic statements into compilable C language.

The observed weight data item 53 shown in Fig. 2 is constructed by having an observer tabulate the number of passengers in the cars 22, 23 against the magnitude of the WEIGHT signals provided by the weight sensors 42, 43. Ideally, this tabulation is performed in the building in which the elevator system 20 is installed in order to have the passenger load and passenger weight distribution data set for the particular building. However, it is also possible to construct the observed weight data item 53 using typical tables showing the probabilities and distributions of the weight of people The tabulated data is used to construct several fuzzy sets stored in the observed weight data element 53. Each fuzzy set corresponds to a particular ridership. For each set, the membership value of each term corresponds to the number of times a particular magnitude of the WEIGHT signal was observed, and the basis elements correspond to the particular weights. Each set can be represented as FO(N), where N is a particular ridership, and each element can be represented as FO(N,W), where W is a particular weight.

Fig. 3 is a plot 90 showing a hypothetical set of fuzzy sets constructed by tabulating passenger load against WEIGHT signal magnitude. Plot 90 includes a plurality of curves 92 through 103, with curve 92 corresponding to the fuzzy set describing the various values of the WEIGHT signal for one passenger, i.e., FO(1), curve 93 corresponding to the fuzzy set describing the various values of the WEIGHT signal for two passengers, FO(2), etc. The relative magnitudes of curves 92-103 indicate the frequency with which a particular WEIGHT signal magnitude is observed and therefore indicate the membership value of the term of the fuzzy set.

Fig. 4 is a flow chart 110 showing the operation of the weight interpretation module 52. The program flow begins at a first step 111 in which a fuzzy set FW(N) (N represents a specific number of passengers) is initialized to contain no terms. Step 111 is followed by a step 112 in which a variable representing the hypothetical number of passengers, PC, is initialized to 1. Step 112 is followed by a test 113 in which the value of the variable PC is compared to PCMAX, a predetermined constant value equal to the maximum number of possible car passengers.

If PC is not greater than PCMAX, control passes from test 113 to a step 114 where a term consisting of the term in the Data element 53 of the observed weight stored fuzzy set FO(PC) is taken and added to the fuzzy set FW. The added tenn corresponds to a passenger number equal to PC and a weight equal to the magnitude of the WEIGHT signal, ie the value of the FO(PC, WEIGHT) term. After step 114 follows a step 115 in which the variable PC is incremented.

Steps 113 through 115 are repeatedly executed until PC exceeds PCMAX on test 113, after which control passes from test 113 to a step 116 in which the fuzzy set FW is normalized. Normalization of a fuzzy set involves dividing all term membership values by a constant value to make the sum of all membership values equal to 1.

After step 116 comes a step 117 in which the fuzzy set FW is defuzzified to produce a value equal to the ridership. Defuzzification is a step by which a fuzzy logic set is reduced to a sharp, i.e., single value. FW can be defuzzified by using the value of the term with the highest membership value. FW can also be defuzzified by adding the products of the membership value of each term and the ridership represented by the term. After step 117 comes a step 118 in which the calculated value of the ridership is stored in a ridership data element 54 shown in Fig. 2. The ridership data element 54 is used by the operations module 56 and the ridership estimation module 60 in a manner described in detail below. If the operation module 56 and the number estimation module 60 are able to use the elevator passenger count expressed as a fuzzy logic set and not as a sharp value, the weight interpretation module 52 does not defuzzify the set FW, but follows a logic path 119 around the defuzzification step 117 and stores the fuzzy set FW instead of a sharp value at step 118.

The policy shown by flow chart 110 of Figure 4 is intended to run on the microprocessor of elevator controller 40 in real time. However, if only a sharp value of passenger count is desired, the policy shown by flow chart 110 can be executed off-line to generate a weight interpretation table, which is a table of values of the magnitude of the WEIGHT signal against passenger count that can be indexed by weight interpretation module 52 at run time.

Fig. 5 is a flow chart 120 showing the off-line construction of a weight interpretation table that can be loaded into the ROM memories of the elevator controller 40 and accessed by the weight interpretation module 52 to provide a sharp car passenger count based on the magnitude of the WEIGHT signal. The WEIGHT signal becomes the index number of the table, and the passenger count is the entry to that index number. In a first step 122, two variables OLDPC and NEWPC are initialized to zero. The first step 122 is followed by a step 124 in which a third variable W is initialized to zero.

After step 124 follows a step 126 in which NEWPC is set equal to the passenger count calculated by the method shown by flow chart 110 of Fig. 4, except that the value of W is used instead of the magnitude of the WEIGHT signal. After step 126 follows a test 128 in which the variable NEWPC is compared with the variable OLDPC. If the variables are equal, control passes from step 128 to a step 130 in which the values of OLDPC and W are entered into the weight interpretation table to be constructed. After step 130 follows a step 132 in which the variable OLDPC is set equal to the variable NEWPC. The table thus created has index numbers defining the values of W (analogous to the magnitude of the WEIGHT signal) which cause a change in the entries into the table of the number of passengers. At run time, the weight interpretation module 52 searches the table to find two adjacent index numbers, one of which is greater than the magnitude of the WEIGHT signal and the other of which is less than or equal to the magnitude of the WEIGHT signal. The entry at the higher of the two index numbers in the table is the passenger number, which is written into the passenger number data element 54.

Step 132 (or test 128 if OLDPC is equal to NEWPC) is followed by a step 134 in which the variable W is incremented by an amount WINC, which is a predetermined constant corresponding to the "granularity" of the weight sensors 42, 43 (i.e., the minimum difference between two weight measurements). After step 134, there is a test 136 in which W is compared to WMAX, a predetermined constant equal to the largest possible magnitude of the WEIGHT signal. If W is not greater than WMAX, control passes from step 136 back to step 126. Otherwise, control passes to a step 137 in which OLDPC and W are added as the last entries to the table. After step 137, the program flow is complete.

The weight interpretation module 52 provides data to the car passenger data element 54 indicative of the number of passengers in the cars 22, 23. The car passenger data element 54, the DEPARTURE TIME signal, the DEPARTURES signal and the ARRIVALS signal are provided as inputs to the operation module 56 which determines the predominant usage pattern of the elevator system 20 and provides the results to the operation mode data element 58. The operation mode of the elevator system 20 may be described as up peak, indicating that most of the operation is from the lobby of the building to higher floors, as down peak, indicating that most of the operation is from higher floors to the lobby, or as without peak, indicating that there is no distinguishable major operating pattern.

Referring to Fig. 6, a data flow diagram 150 shows the function of the operations module 56. The DEPARTURE TIME and DEPARTURE signals and the data from the car passenger data element 54 are provided as inputs to an up-calculation module 52 which uses the UP-PEAK START rules (described in detail below) to calculate a value for the UP-PEAK variable which is stored in an up-PEAK data element 154. The upward calculation module 152 also uses the UPWARD PEAK TERMINATION rules (described in detail below) to determine a value for an UPWARD PEAK variable that is stored in an upward PEAK data element 156.

The DOWNTIME and ARRIVALS signals and data from the car passenger data element 54 are provided to a down calculation module 162 which uses the DOWNPEAK BEGINNING rules (described in detail below) to calculate a value for the DOWNPEAK variable which is stored in a downpeak data element 164. The down calculation module 162 also uses the DOWNPEAK END rules (described in detail below) to determine a value for a DOWNNOPEAK variable which is stored in a downnopeak data element 166. The up-no-peak data element 156 and the down-no-peak data element 166 are provided as inputs to a remainder calculation module 170, which combines the data of elements 156, 166 (in a manner described in detail below) to produce a value for the NO PEAK variable stored in the No Peak data element 172.

The UP-PEAK variable stored in the up-peak data element 154, the DOWN-PEAK variable stored in the down-peak data element 164, and the NO-PEAK variable stored in the no-peak data element 172 are provided as inputs to an operating mode resolver module 174 which combines the input data to provide a result to the operating mode data element 58. The result may be either a single sharp value or a fuzzy set, depending on the nature of the subsequent steps that use the information from the operating mode data element 58. A sharp value (i.e., a single statement of the operating state) can be obtained by considering the operating state as either up-peaking, down-peaking, or peakless depending on which of the UP-PEAK, DOWN-PEAK, or NO-PEAK variables is the largest.

The operating mode resolver module 174 may also provide a fuzzy set to indicate the operating mode of the elevator system 20. The set would have a term corresponding to the UP-PEAK variable, a term corresponding to the DOWN-PEAK variable, and a term corresponding to the NO-PEAK variable, with the membership value of each term being proportional to the values of the UP-PEAK, DOWN-PEAK, and NO-PEAK variables, respectively.

The UP-PEAK-BEGIN rules used by the up-calculation module 152 have the form:

if (ABF-ZEIT-i is SHORT TIME)

and (i is DIFFERENT CABINS)

and (ABF-LIFT-1 is HEAVILY LOADED)

and (ABF-LIFT-2 is HEAVILY LOADED)

and...

and (ABF-AUFORT-i is HEAVILY LOADED)

then set UP-PEAK

where ELEVATOR-CAB-1 is the car that left the lobby last, ELEVATOR-CAB-2 is the car that left second to last, and in general ELEVATOR-CAB-i is the car that left i-th last. DEPARTURE-TIME-i is defined as the time that has elapsed since the i-th last car left, and i is a number from 1 to N, the rule number. The number of rules N is set equal to the number of cars. However, N could also be chosen so that it is either smaller than or larger than the number of cars.

For example, the calculation rule that corresponds to i equal to 3 would be the following:

if (ABF-TIME-3 is SHORT TIME)

and (3 DIFFERENT CABINS)

and (ABF-LIFT-1 is HEAVILY LOADED)

and (ABF-LIFT-2 is HEAVILY LOADED)

and (ABF-LIFT-3 is HEAVILY LOADED)

then set UP-PEAK

For each of the N rules, the up-calculation module 152 determines the condition part of the UP-PEAK-BEGIN rule and sets the result (the last value of the UP-PEAK variable) equal to the value of the condition. The last value of the UP-PEAK variable is equal to the maximum value resulting from evaluating each of the N UP-PEAK-BEGIN rules.

Referring to Fig. 7, a flow chart 180 for evaluating the N UP-PEAK-BEGIN rules shows a first step 182 in which i is initialized to the value of 1 and is followed by a second step 184 in which a variable OLD UP-PEAK is initialized to zero. Step 184 is followed by a test 186 in which the value of i is compared to the value of N of the number of rules. If in test 186 is greater than N, program execution is complete. Otherwise, control passes from test 186 to a step 188 in which the UP-PEAK-BEGIN rule is used to calculate a value for the UP-PEAK variable. Step 188 is followed by a step 190 in which the variable i is incremented.

Step 190 is followed by a test 192 in which the value of ALT UPPEAK is compared to UPPEAK. If ALT UPPEAK is not greater than UPPEAK, control passes from step 192 to a step 194 in which ALT UPPEAK is set equal to UPPEAK. If at step 192 ALT UPPEAK is greater than UPPEAK, control passes from step 192 to a step 196 in which UPPEAK is set equal to ALT UPPEAK. Control passes through step 194 or step 196 back to test 186. Steps 192, 194, 196 ensure that the variables UP-PEAK and ALT-UP-PEAK are always equal to the largest value for UP-PEAK calculated at step 188.

The UP-PEAK-TERMINATION rules used by the up-calculation module 152 are of the form:

if (ABF-ZEIT-i is not SHORT TIME)

and (ABF-LIFT-1 is not HEAVILY LOADED)

and (ABF-LIFT-2 is not HEAVILY LOADED)

and...

and (ABF-AUFORT-i is not HEAVILY LOADED)

then set NON-UPWARD-PEAK

The up-calculation module processes the N UP-PEAK-TERMINATION rules in a manner similar to that shown in Figure 7 for processing the UP-PEAK-BEGINNING rules, such that NON-UP-PEAK is the maximum value resulting from the UP-PEAK-TERMINATION rule. The up-calculation module 152 stores the value of the NON-UP-PEAK variable in the NON-UP-PEAK data element 156.

The DOWN-PEAK-BEGIN rules used by the down-calculation module 162 have the form:

if (ARRIVAL-TIME-i is SHORT TIME)

and (i is DIFFERENT CABINS)

and (ANK-LIFT-1 is HEAVILY LOADED)

and (ANK-LIFT-2 is HEAVILY LOADED)

and...

and (ANK-LIFT-i is HEAVILY LOADED)

then set DOWN-PEAK

whereas the DOWN PEAK TERMINATION rules used by the down calculation module 162 have the following form:

if (ARRIVAL-TIME-i is not SHORT TIME)

and (ANK-LIFT-1 is not HEAVILY LOADED)

and (ANK-LIFT-2 is not HEAVILY LOADED)

and...

and (ANK-LIFT-i is not HEAVILY LOADED)

then set NON-DOWN-PEAK

As with the up-calculation module 152, the down-calculation module 162 processes the N DOWN-PEAK-BEGIN rules and DOWN-PEAK-END rules in a manner similar to that shown in Figure 7 for processing the UP-PEAK-BEGIN rules. The resulting values for DOWN-PEAK and NON-DOWN-PEAK are the maximum values calculated for the N DOWN-PEAK-BEGIN rules and DOWN-PEAK-END rules. The down calculation module 162 stores the values of the DOWN PEAK variable and the NON-DOWN PEAK variable in the DOWN PEAK data element 164 and the NON-DOWN PEAK data element 166, respectively.

The remainder calculation module 170 sets the variable NOPEAK to the maximum value of NON-UPPEAK (from the non-up-peak data element 156) and NON-DOWNPEAK (from the non-down-peak data element 166). The variable NOPEAK is provided by the remainder calculation module 170 as an output to the non-peak data element 172.

The UP-PEAK-BEGINNING, UP-PEAK-TERMINATION, DOWN-PEAK-TERMINATION rules can generally be described as if they had the form:

if < condition> then set X-PEAK

where X-PEAK is either UP-PEAK, DOWN-PEAK, UP-NOT-PEAK, or DOWN-NOT-PEAK.

For a fuzzy logic conditional expression, the value of the result variable (the variable that follows the "then" part of the condition) is set according to the value of the condition. In the preceding equation, therefore, X-PEAK is set to a value that depends on the value of the <condition>.

The conditional part of the UP-PEAK-BEGIN statement contains several simple expressions, such as (DEPARTMENT-TIME-i is SHORT TIME) and (i is VARIOUS CABINS) connected by ANDs. The value of the condition is the minimum value of the simple expressions. Evaluating the simple expressions requires quantifying SHORT TIME, VARIOUS CABINS and HEAVY LOAD.

Reference is made to Figures 8, 9 and 10, in which a first representation 200 shows a fuzzy set for representing SHORT TIME, a second representation 202 shows a fuzzy set for representing VARIOUS CARS and a third representation 204 shows a fuzzy set for representing HEAVY LOADED. The VARIOUS CARS representation 202 is for a multiple car elevator system. A representation for the two car elevator system 20 of Figure 1 would make it more difficult to represent the VARIOUS CARS fuzzy set.

The representation 200 is overlapped by several rectangles 210-217, each of which represents a term of the SHORT TIME fuzzy set. Similarly, the representation 202 is overlapped by several rectangles 220-227 to represent the terms of the VARIOUS CABINS fuzzy set, and the In Figure 204, several terms 230-242 are superimposed to represent the elements of the HEAVY LOADED fuzzy set.

The vertical axes of plots 200, 202, 204 indicate the membership value for the terms represented by rectangles 210-217, 220-227, 230-242 and the horizontal axes of plots 200, 202, 204 represent the values of the basis elements. For example, rectangle 210 represents a term of the fuzzy set SHORT TIME with a basis element value of zero and a membership value of 1.0 and rectangle 213 represents a term of the fuzzy set SHORT TIME with a basis element value of 3 and a membership value of about 0.4.

Each simple expression for the UP-PEAK-BEGIN rule is evaluated using the SHORT TIME, VARIOUS CABINS, and HEAVY LOADED fuzzy sets. The value of (ABF-TIME-i is SHORT TIME) then becomes the membership value of the SHORT TIME fuzzy set term with a base element equal to the ABF-TIME-i. For example, if the value of ABF-TIME-i is 5 minutes, the expression (ABF-TIME-i is SHORT TIME) is equal to the membership value of the SHORT TIME fuzzy set term with the base element equal to 5 minutes, which is represented in the representation 200 by the rectangle 215.

The value of the (i DIFFERENT CABINS) expression is the membership value of the term of the fuzzy set DIFFERENT CABINS with a base element equal to i. For example, if i is equal to 3, the expression (i DIFFERENT CABINS) is equal to the membership value of the term of the fuzzy set DIFFERENT CABINS with a base element equal to 3, which is shown in the representation 202 by the rectangle 223.

The evaluation of the (ABF-LIFT-i is HEAVY LOADED) expression depends on whether the number of passengers in car i, which is an input value to the operating module 56, is a sharp value or a fuzzy set. If the number of passengers is a sharp value, the (ABF-LIFT-i is HEAVY LOADED) expression equals the membership value of a term of the fuzzy set HEAVY LOADED with a base element value equal to the sharp number of passengers.

If the number of passengers in car i is expressed as a fuzzy set, the (Elevator-i is HEAVY LOADED) expression is evaluated by taking the maximum value of the membership values of the terms of a fuzzy set formed by intersecting the number of passengers fuzzy sets and the HEAVY LOADED fuzzy set. In general, a fuzzy set formed by intersecting two fuzzy sets is a fuzzy set with terms whose membership value is equal to the minimum membership value of the corresponding terms (i.e., terms with the same basis element). For example, if F1 = {0.1 A, 0.5 B, 0.7 C} and F2 = {0.3 A, 0.2 C}, then the intersection of F1 and F2 = {0.1 A, 0.2 C}. The value of the (ABF-LIFT-i is HEAVY LOADED) expression is the maximum membership value of the terms of the fuzzy set formed by intersecting the fuzzy set of the number of passengers and the fuzzy set of HEAVY LOADED.

Evaluating the UP-PEAK-END, DOWN-PEAK-BEGIN, and DOWN-PEAK-END rules is similar to evaluating the UP-PEAK-BEGIN rules shown previously. The values of the UP-PEAK, DOWN-PEAK, and NO-PEAK variables will be between 0 and 1.

The processing shown here for the operating module 56 can be done at runtime or it can be done out of sequence. In the latter case, a table is constructed with index numbers indicating possible input values to the operating module 56 and having entries indicating possible output values of the operating module 56. The construction and use of a similar table for the weight interpretation module 52 is shown in Fig. 5 and the related discussion. One skilled in the art will be able to construct and use a similar table for the operating module 56 from the specific example of Fig. 5.

Figure 11 is a data flow diagram 260 showing the operation of the number estimator module 60 which estimates the number of concourse passengers waiting at a particular stop at a particular time. The number estimator module 60 processes the ABL TIME, DEPARTURES, ARRIVALS and CONcourse CALLS signals together with data from the passenger number data element 54 and the operation mode data element 58 and writes the output values to the concourse passengers data element 62.

Referring to Figure 11, a first rate calculation module 262 receives data from the car passenger data element 54 and the ABL TIME, DEPARTURES and ARRIVALS input signals. The first rate calculation module 262 estimates the rate at which passengers arrive at a stop to wait for a car serving the stop. The calculations by the first rate module 262 (which is described in more detail below) are based on an estimate of the number of passengers entering a car at a stop and the time that has elapsed since the stop was last served. The first rate calculation module 262 provides an estimated rate and information indicative of the particular stop to a first rate data element 264.

A second rate calculation module 266 receives inputs from the ABL TIME, DEPARTURES, ARRIVALS and HALL CALL signals. The second rate calculation module 266 also estimates the rate at which hall passengers arrive at a stop to wait for a car serving the stop. The second rate calculation module 266 provides the estimated rate along with information indicative of the particular stop to a second rate data element 268. The calculations by the second rate module 266 (described in detail below) are based on the elapsed time between a car serving a particular stop and the subsequent pressing of a hall call button for that stop by a hall passenger.

A rate averager 270 uses data from the rate data elements 264, 268 together with data from the operating mode data element 58 to calculate an up rate stored in an up rate data element 272, a down rate stored in a down rate data element 274, and a rest rate stored in a rest rate data element 276. The up rate data element 272 contains information indicative of the up passenger rate (i.e., the rate at which hall passengers arrive at the lobby to wait for a car going to other floors). The down rate data element 274 contains information indicative of the rate at which hall passengers arrive at other floors to wait for a car going to the lobby. The remaining rate data element 276 contains information indicative of the rate of concourse passengers arriving to wait for cars traveling between floors on which the concourse is not located.

When new rate values are calculated by the first and second rate calculators 262, 266 and placed in a first and a second rate data element 264, 268, respectively, the rate mediator 270 sets the up, down and rest data elements 272, 274, 276 according to the current operating mode of the elevator system 20. If the operating mode is a sharp value (i.e., a single value indicative of either up, down or rest), then the rate averager 270 carries the data from the first and second rate data elements 264, 268 only into the corresponding up, down, down or rest rate data elements 272, 274, 276, respectively.

On the other hand, if the operating state of the elevator system 20 is expressed as a fuzzy set having three terms as an indicator of the extent to which the system is in the up, down and rest operating modes, the first and second rate data items 264, 268 are entered into the up, down or rest rate data items 272, 274, 276 in proportion to the membership value of the terms of the fuzzy set of the operating mode stored in the operating mode data item 58.

The operating mode data element 58 and the up, down and remainder data elements 272, 274, 276 are provided to the rate converter module 278 which uses the input data and the ARRIVALS, DEPARTURES and ABL TIME signals to estimate the number of hall passengers at a particular stop waiting to be served by the car. The number of hall passengers at a stop is determined by multiplying the rate (of one or more of the up, down and remainder data elements 272, 274, 276, depending on the operating mode of the system) by the time elapsed since the stop was last served. The number of passengers is provided to a hall passenger data element 62 as a fuzzy set or a sharp value, depending on the requirements of the subsequent step. The function of the Rate Converter Module 278 is described in detail below.

The first rate calculation module 262 calculates an instantaneous passenger rate, INSTRATE1, whenever a car stops at a floor to answer a hall call. The assumption is made that any passenger who will exit the car will do so before any hall passengers enter the car. The number of passengers entering the car is therefore determined by subtracting the minimum passenger count from the last passenger count (i.e., the number of passengers in the car when the elevator doors closed). The rate INSTRATE1 provided to the first rate data element 264 is equal to the number of passengers entering the car divided by the time elapsed since that particular stop was last served.

If the number of passengers provided by the car passenger data element 54 is a sharp value, the subtraction and divisions described above are simple. However, if the number of car passengers is expressed as a fuzzy set, the fuzzy set describing the number of passengers when the weight is a minimum is subtracted from the fuzzy set describing the number of passengers in the car when the elevator door is closed. The subtraction is performed by subtracting each combination of basis elements and taking the minimum of the membership values of the terms so subtracted. The terms having the same basis element are combined into a single term with a membership value equal to the maximum membership value of the combined terms.

For example, assume that the fuzzy set F1 is {u A, v B, w C} and that the fuzzy set F2 is {x D, y E, z F}. The fuzzy set formed by subtracting F2 from F1 would be equal to:

{min (u,x) A-D), min(u,y) (A-E), min(u,z) (A-F), min(v,x) (B-D), min(v,y) (B-E), min(v,z) (B-F), min(w,x) (C-D), min(w,y) (C-E), min(w,z) (C-F)}

All terms with the same basis elements (for example, if A-D equals C-E) are combined by taking the maximum of the membership values of these terms, e.g.:

max(min(u,x), min(w,y))

As an additional step to subtracting the ridership fuzzy sets, any term of the resulting set with a basis element less than zero is eliminated, since less than zero passengers entering the car at a stop makes no sense. The first rate calculation module 262 determines INSTRATE1, the fuzzy set representing the rate of concourse passenger arrivals, by dividing the basis elements of the fuzzy set resulting from subtracting the ridership fuzzy sets by the time that has elapsed since the particular stop was last served. The resulting ridership rate fuzzy set is provided to the first rate data element 264 along with information indicative of the particular stop.

The second rate calculation module 266 determines a hall passenger arrival rate INSTRATE2 whenever a hall call button is pressed. The elapsed time (T) between the last operation of the special stop and the pressing of the hall call button is used to build the fuzzy set INSTRATE2 having base elements with values 1/T, 2/T, 3/T, ... 10/T, where each term has a membership value defined by the following formula:

Membership value = RTe-RT

where e is the natural logarithm and R is the base element of the corresponding term of the fuzzy set.

The fuzzy set of the rate INSTRATE2 generated by the second rate calculation module 266 assumes that the arrival of hall passengers follows a Poisson distribution. The number of hall passengers increases by one at a time. INSTRATE2 is supplied to the second rate data element 268 as an output value by the second rate calculation module 266, together with information as an indicator of the special stop.

The first rate calculator 262 updates the value of INSTRATE1 stored in a first rate data element 264 in response to a car serving a hall call. The second rate calculator 266 updates INSTRATE2 stored in a second rate data element 268 in response to a hall passenger pressing a hall call button. The fuzzy sets representing INSTRATE1 and INSTRATE2 are used to update the fuzzy set stored in the up rate data element 272, the down rate data element 274, and the remainder rate data element 276.

Before INSTRATE1 and INSTRATE2 are used to update the values of the overall system rates stored in the up, down and no-rate data elements 272, 274, 276, an adjustment is made to their values to compensate for the greater probability of lower floors for up hall calls and higher floors for down hall calls. When a new INSTRATE1 or INSTRATE2 is calculated in response to a descending car or hall call, the base elements of the fuzzy sets INSTRATE1 and INSTRATB2 are divided by (i-1)/(F-1), where F is the total number of floors in the building and i is the particular floor that serves the system. Similarly, the base elements of the fuzzy sets INSTRATE1 and INSTRATE2 are divided by (Fi)/(F-1) whenever a recalculation occurs in response to serving a call with an ascending car.

The rate averager 270 updates the fuzzy set stored in the up-rate data element 272 whenever INSTRATE1 is updated in response to a car serving the lobby or whenever INSTRATE2 is updated in response to a lobby call button being pressed. The new up-rate fuzzy set is calculated by the following equation:

(0.2 x UM x INSTRATE) + (1.0 - (0.2 x UM)) x (fuzzy set of old upward rate)

INSTRATE in the preceding formula means either INSTRATE1 or INSTRATE2. UM is the membership value of the operating mode fuzzy set term (from operating mode data element 58) corresponding to the up rate of operation. UM is in the range of 0 to 1. Using UM in the preceding equation causes the up rate fuzzy set to be affected by INSTRATE only to the extent that the elevator system is currently in the UP mode. The multiplication in the above equation affects the membership values of the fuzzy set terms. The addition is performed using standard techniques known to those skilled in the art for adding fuzzy sets.

The rate averager 270 updates the down rate fuzzy set and stores the new value in the down rate data element 274 whenever INSTRATE1 is updated in response to a down car. or whenever INSTRATE2 is updated in response to a down hall call button being pressed. The fuzzy set of the new down rate is calculated by the following equation:

(0.2 x DM x INSTRATE) + (1.0 - (0.2 x DM)) x {fuzzy set of downward rate}

INSTRATE in the preceding formula means either INSTRATE1 or INSTRATE2. DM is the membership value of the term of the fuzzy set of the operation mode corresponding to the down operation rate. DM is in the range of 0-1.

The rate averager 270 updates the residual rate fuzzy set and stores the new value in the residual rate data element 276 whenever INSTRATE1 or INSTRATE2 is updated. The new value of the residual rate fuzzy set is calculated by the following equation:

(0.2 x OP x INSTRATE) + (1.0 - (0.2 x OP)) x (fuzzy set of the residual rate)

INSTRATE in the preceding formula is either INSTRATE1 or INSTRATE2. OP is the membership value of the term of the fuzzy set of the operating mode corresponding to the non-peak operating state. OP is in the range of 0 to 1.

The rate converter module 278 provides a fuzzy set indicative of the number of concourse passengers waiting for a car at a particular stop at any particular time. First, a fuzzy set of the total rate is constructed by combining the fuzzy set from the up rate data element 272, the fuzzy set from the down rate data element 274, and the fuzzy set from the rest rate data element 276. The sets are defined by weighting the membership values of each term of the sets by the relative membership value of the corresponding terms of the operation mode fuzzy set such that the membership values of the up-rate fuzzy set are weighted by UM/(UM+DM+OP), the membership values of the down-rate fuzzy set are weighted by DM/(UM+DM+OP), and the membership values of the rest-rate fuzzy set are weighted by OP/(UM+DM+OP). After weighting the membership values, the three sets are added together and then the values of the basis elements of the result are divided by three to produce the total rate fuzzy set.

The DEPARTURETIME and DEPARTURE signals are used to determine the time that has elapsed since a particular stop was last served. A fuzzy set indicative of the number of concourse passengers waiting at a particular stop is formed by multiplying the values of the base elements of the total rate fuzzy set by the elapsed time. The resulting fuzzy set is provided to the concourse passenger data element 62 by the number estimation module 60.

The processing shown here for the number estimator module 60 can be done at runtime or out of order. In the latter case, a table is constructed with index numbers as indicators of possible inputs to the number estimator module 60 and entries as indicators of possible outputs of the number estimator module 60. The construction and use of a similar table for a weight interpretation module 52 is shown in Figure 5 and the related description. Those skilled in the art can extrapolate from the specific example of Figure 5 to construct and use a similar table for the number estimator module 60.

There are many indicators to measure the performance of an elevator system, such as the average waiting time, the waiting threshold and the average service time. The average wait time is the average time between a hall call and the service of the hall call. The wait threshold is the average number of people waiting longer than a constant predetermined time. The average service time is the average time between a hall passenger pressing a hall call button and the same passenger arriving at the destination floor. The details of calculating the average wait time, the wait threshold and the average service time are known to those skilled in the art. Note that the elevator performance indicators can be calculated using either crisp values or fuzzy sets.

The HALL CALL signal is provided as an input to the performance estimator 64 which, in response to a particular hall call, uses the operating mode data element and the hall passenger data element 58, 62 to construct a plurality of fuzzy sets of performance. Each fuzzy set of performance corresponds to a particular elevator performance indicator. Each term of each set represents the estimated value of the particular performance indicator corresponding to serving the hall call with a particular car. The performance estimator 64 stores the fuzzy sets in the performance data element 66.

Referring to Fig. 12, a plot 290 has a plurality of bars 292-297, the height of each bar indicating the inverse of the estimated average waiting time associated with a particular car. A higher bar indicates a lower average waiting time. The plot 290 may represent a fuzzy set of performance, wherein each base element of the set corresponds to a particular car and the membership value of a base element corresponds to the height of each bar 292-297. Similar fuzzy sets may be constructed for any other elevator performance indicator that can either be directly measured or derived from direct measurements. The specific indicators selected and the method of calculation depend on a variety of functional factors known to the person skilled in the art.

Referring to Fig. 13, a flow chart 300 shows steps for constructing a plurality of performance fuzzy sets, which are represented on the flow chart 300 by the symbol P. The notation P (I, C) indicates the C-th term corresponding to the car number equal to C of the I-th performance fuzzy set.

In a first step 302, P is initialized so that it contains no terms and no fuzzy sets. The step 302 is followed by a step 304 in which an index variable I for indexing all fuzzy sets of performance to I is initialized. The step 304 is followed by a test 306 in which I is compared with IMAX, a predetermined constant equal to the number of performance indicators.

If at step 306 I is not greater than IMAX, control passes from step 306 to a step 307 in which C, an index variable for indexing over the terms of the fuzzy set of performance (and therefore corresponding to each car), is initialized to 1. Step 307 is followed by a step 308 in which C is compared to CMAX, the number of cars in the system. If at test 308 C is not greater than CMAX, control passes from step 308 to a step 309 in which it is assumed that car C is assigned to serve a particular hall call at a particular stop.

Step 309 is followed by a step 310 in which P (I, C) is determined, which corresponds to the C-th term of the I-th fuzzy set. In step 310, it is assumed that the car C is assigned to a particular hall call and the performance of the system is calculated using equations and calculation methods appropriate for the I-th performance indicator. The P (I, C) calculated in step 310 Value is the membership value of the C-th term in the I-th fuzzy set of the performance.

After step 310 comes a step 311 in which the index variable C is incremented. After step 311, control passes back to test 308. If at test 308 C is greater than CMAX, indicating that the performance of the system for the I-th performance indicator for all cars has been calculated, control passes from test 308 to a step 312 in which the index variable I is incremented as an index of the special performance criterion. Control passes from step 312 back to test 306 where I is compared with IMAX. If at step 306 I is greater than IMAX, all fuzzy sets of performance have been calculated and processing is complete.

Referring to Figure 14, a customer preference representation 320 has a plurality of bars 322-328, where each bar 322-328 corresponds to a particular elevator performance indicator and the height of each bar 322-328 indicates the importance of the performance indicator to the customer. For example, the height of bar 323 representing average service time is greater than the height of bar 322 representing average wait time, indicating that when given a choice between optimizing performance using average wait time or optimizing performance using average service time, the customer prefers to use average service time.

The representation 320 may represent a fuzzy set of customer preferences, where each base element of the set corresponds to a particular elevator performance indicator and the height of each bar 322-328, as an indicator of the relative importance to the customer of each elevator performance indicator, corresponds to the membership value of each term of the fuzzy set. The fuzzy set of customer preferences may be established by the elevator manufacturer or it may be provided by the customer entered using a variety of data entry devices apparent to those skilled in the art. The fuzzy set of customer preferences is stored in the customer preference data element 68.

The performance data element and the customer preference data element 66, 68 are provided as input values to the assignment convenience calculation module 70, which determines the convenience of assigning each car to serve a hall call and provides to the assignment convenience data element 72 a fuzzy set of assignment conveniences with base elements corresponding to each car, the membership value of each base element corresponding to the convenience of assigning the associated car to the particular hall call.

Referring to Fig. 15, a flow chart 340 shows the operation of the allocation utility calculation module 70. The symbol AU indicates the allocation utility fuzzy set and the symbol SP indicates a plurality of weighted performance fuzzy sets. The symbol CP indicates the customer preference fuzzy. In a first step 342, the AU and SP fuzzy sets are initialized to be empty. The step 342 is followed by a step 344 in which an index variable I for indexing in the performance and weighted performance fuzzy sets is initialized to 1. The step 344 is followed by a test 346 in which I is compared to IMAX, a predetermined constant equal to the number of performance fuzzy sets (i.e., the number of performance indicators).

If at step 346 I is not greater than IMAX, control passes from step 346 to a step 348 in which C, a variable for indexing across all the cars of the elevator system, is set to 1. Step 348 is followed by a test 350 in which C is compared to CMAX, a predetermined constant equal to the number of cars in the elevator system. If C is not greater than CMAX at step 350, control proceeds from step 350 to a step 352.

At step 352, the membership value of the C-th term in the I-th weighted fuzzy set of performance is set equal to the membership value of the C-th term of the I-th fuzzy set of performance times the membership value of the I-th term of the fuzzy set of customer preferences. If CP(I) is close to one, indicating that the I-th performance indicator is important to the customer, the membership value of the C-th term of the I-th weighted fuzzy set of performance will be approximately equal to the C-th term of the I-th fuzzy set of performance. On the other hand, if CP(I) is at or near zero, indicating that the I-th performance indicator is not important to the customer, the C-th term of the I-th weighted fuzzy set of performance will be equal to or near zero, regardless of the value of the C-th term of the I-th fuzzy set of performance.

After step 352 comes a step 354 in which the variable C is incremented. After step 354, control passes back to test 350 where C is compared to CMAX. If at step 350 C is greater than CMAX, indicating that all of the terms of the I-th weighted fuzzy set of performance have been calculated, control passes from step 350 to a step 356 in which I, the variable used to index over the performance indicators, is incremented. After step 356, control passes back to test 346 where I is compared to IMAX, the number of performance indicators.

If at test 346 I is greater than IMAX, control passes from test 346 to a step 358 in which the index variable I is initialized to 1. After step 358, a test 360 follows where I is compared with IMAX, the number of performance indicators. If at step 360 I is not greater than IMAX, control passes to a step 362 where the membership value of the C-th term of the fuzzy sets of the utility of the assignment is set equal to the first term of the I-th weighted fuzzy set of the performance. After the step 362 follows a step 364 in which the variable C for indexing over the terms of the fuzzy sets of the weighted performance and the utility of the assignment is set to 2.

After step 364, there follows a test 366 in which C is compared to CMAX, the number of cars in the system. If C is not greater than CMAX, control passes to step 368 where the C-th term of the allocation utility fuzzy set is set equal to the greater of the C-th term of the I-th weighted performance fuzzy set and the previous value of the C-th term of the allocation utility fuzzy set. Step 368 ensures that the C-th term of the allocation utility fuzzy set is always equal to the maximum value of the C-th term of all weighted performance fuzzy sets.

Step 368 is followed by a step 370 where the variable C is incremented. After step 370, control passes back to test 366. If at step 366, C is greater than CMAX, control passes from step 366 to a step 372 where I, the index variable for the performance indicators, is incremented. After step 372, control passes back to test 360 where, if I is greater than IMAX, processing is complete. The resulting fuzzy set of assignment utility is fed to assignment utility data element 72.

The processing shown here for the performance estimator module 64 and the assignment convenience module 70 can be performed at runtime or can be performed out of sequence. In the latter case, a table is formed with index numbers as indicators of possible inputs to the performance estimator module 74 and the assignment convenience module 70 and with entries as indicators of possible output values to the performance estimator module 64 and the allocation utility module 70. The construction and use of a similar table for the weight interpretation module 52 is shown in FIG. 5 and the associated description. One skilled in the art will be able to extrapolate from the specific example of FIG. 5 to construct and use a similar table for the performance estimation module 64 and the allocation utility module 70.

The assignment convenience data element 72 is provided as input to the uncertainty filter module 74, which determines the final cabin assignment (a sharp value) by selecting the term of the assignment convenience fuzzy set with the highest membership value. The uncertainty filter module 74 will only provide an assignment to the assignment data element 76 if the assignment uncertainty is below a certain predetermined value stored in the customer preferences data element 68, which is provided as input to the uncertainty filter module 74. The uncertainty of an assignment is defined as the membership value of the term with the highest membership value divided by the sum of the membership values for all terms of the assignment convenience fuzzy set. A customer who prefers a relatively quick assignment of a car to a hall call would specify a high level of uncertainty, while a customer who does not care about a quick assignment would specify a low level of uncertainty.

Alternatively, an uncertainty filter module 74 may assign an assignment to the assignment data element 76 after a constant predetermined time stored in the customer preference data element 68. The value of the assignment will be the cabin designated by the base element of the assignment convenience fuzzy set with the highest associated membership value.

As a third alternative, the uncertainty filter module 74 may set the uncertainty threshold as a function of the time elapsed since the hall call button was pressed. As the elapsed time increases, the threshold decreases. The threshold versus time function may be linear or non-linear, depending on the requirements of the particular elevator system.

The processing shown here for the uncertainty filter module 74 can be done at runtime or can be done out of order. In the latter case, a table is formed with index numbers as indicators of possible input values to the uncertainty filter module 74 and with entries as indicators of possible output values of the uncertainty filter module 74. Construction and use of a similar table for the weight interpretation module 52 is shown in Fig. 5 and the associated description. One skilled in the art will be able to extrapolate from the specific example of Fig. 5 to construct and use a similar table for the uncertainty filter module 74.

The invention shown herein is applicable to any elevator system having any number of cars stopping at any number of floors, having any maximum capacity, maximum speed, or having any other specific set of physical characteristics. Accordingly, the invention can be practiced regardless of the physical design of the elevator system including the drives, counterweights, cable guides, door mechanisms, hall call and car call signaling devices.

Although the invention has been described in a single floor elevator system with a lobby at the lowest floor of the building, the invention can be practiced regardless of whether the elevator system has more than one lobby floor and whether or not the lobby floor is the lowest floor in the building.

Furthermore, the invention can be practiced independently of the method used to perform other elevator dispatch functions, independent of the particular electronic hardware used to practice the invention or the design of the load weighing device. Portions of the processing shown here can be implemented with electronic hardware that is simple in terms of hardware/software equivalence, which is described (in another field) in U.S. Patent No. 4,294,162 entitled "Force Feel Actuator Fault Detection with Directional Threshold" (Fowler et al.). Instead of reading and writing data to and from data elements, the hardware would communicate by receiving and sending electronic signals.

Although only the runtime operation of the operation module 56, the count estimator module 60, the performance estimator module 64, the allocation utility calculation module 70, and the uncertainty filter module 74 are shown here, the modules 56, 60, 64, 70, 74 can be run out of order to generate lookup tables containing all possible input values and the resulting output values. The out of order generation and use of a lookup table is described for the weight interpretation module 52 in Figure 5 and the related text of the application. Many of the modules that use fuzzy values for input values can be adapted to sharp input values in a manner obvious to those skilled in the art. The performance estimator module 64 and the customer preference data element 68 can be adapted to use any type of elevator performance criteria. The invention can be carried out regardless of the mechanism used to set or change the customer preferences.

Claims (12)

1. Method for determining the number of cabin passengers, comprising the following steps: Determining a weight signal as an indicator of the passenger weight; Providing a plurality of observed weight fuzzy sets, each of which corresponds to a specific number of passengers, and wherein for each of the observed weight fuzzy sets, the membership value of each term corresponds to the frequency of occurrence of a specific value for the magnitude of the weight signal; and forming a passenger count fuzzy set as an indicator of the number of cabin passengers, wherein for each term, the membership value is equal to the membership value of a term of a specific one of the observed weight fuzzy sets with a basis element corresponding to the magnitude of the weight signal, and wherein the basis element for each term of the passenger count fuzzy set is an indicator of a number of passengers represented by the one specific observed weight fuzzy set. 2. A method for determining the number of cabin passengers according to claim 1, further comprising the step of: Defuzzify the fuzzy set of ridership to produce a single sharp value for ridership. 3. A method for determining the number of cabin passengers according to claim 2, wherein the defuzzification is carried out by setting the one sharp value of the passenger number equal to the base element of the term of the fuzzy set of the passenger number with the highest membership value. 4. Method for determining the number of cabin passengers according to claims 1 to 3, further comprising the following steps: Normalize the fuzzy set of passenger numbers by dividing the membership value of each term by the sum of the membership values of all terms. 5. A method for determining the number of cabin passengers according to claim 4, further comprising the step of: Defuzzify the fuzzy set of ridership to produce a single sharp value for ridership. 6. Device for determining the number of cabin passengers, comprising: a device for detecting a weight signal as an indicator of the passenger weight; means for providing an observed weight signal as an indicator of a plurality of fuzzy sets of observed weight, each set corresponding to a specific number of passengers, and for each of the sets the membership value of each term corresponding to the frequency of occurrence of a specific value for the magnitude of the weight signal; and means, responsive to the weight signal and the observed weight signal, for determining a fuzzy passenger signal as an indicator of a fuzzy set of passenger counts, for each term the membership value being equal to the membership value of a term of a specific one of the observed weight fuzzy sets with a basis element corresponding to the magnitude of the weight signal, and wherein the basis element for each term of the passenger count fuzzy set is an indicator of a number of passengers represented by the one specific observed weight fuzzy set. 7. Apparatus for determining the number of cabin passengers according to claim 6, further comprising: means, responsive to the fuzzy passenger signal, for defuzzifying the fuzzy set of passenger counts to provide a passenger count signal indicative of a single sharp value for the number of passengers. 8. Apparatus for determining the number of cabin passengers according to claim 7, wherein the defuzzification is carried out by selecting the term of the fuzzy set of the number of passengers with the highest membership value. 9. Device for determining the number of cabin passengers according to claims 6 to 8, further comprising: means, responsive to the fuzzy passenger signal, for providing a normalized passenger count signal by dividing the membership value of each term of the fuzzy passenger count set by the sum of the membership values of all terms to produce a normalized fuzzy passenger count set. 10. Device for determining the number of cabin passengers according to claims 6 to 9, further comprising: means, responsive to the normalized passenger count signal, for defuzzifying the normalized fuzzy set of passenger counts to provide a passenger count signal indicative of a single sharp value of the number of passengers in the car. 11. Apparatus for determining the number of cabin passengers according to claim 10, wherein the device for defuzzification comprises: means for multiplying the membership value of each term of the fuzzy set of passenger numbers by the base element of each term; and a device for adding the results obtained by the multiplication step. 12. Use of a method according to claims 1 to 5 to construct a table with index numbers as an indicator for the passenger weight and entries as an indicator for the number of passengers, in which a variable is used instead of the size of the weight signal.