TWI649725B - Smart home monitoring system and smart home monitoring method - Google Patents

Abstract

A smart home monitoring system includes a sensing module and a behavior judgment module. The sensing module includes a plurality of door frame sensors, and each of the door frame sensors is disposed on a door frame of a corresponding one of the plurality of home appliances, and is configured to generate an opening signal correspondingly when the door frame of the corresponding home device is in an open state. . The behavior judgment module includes a learning unit and a comparison unit. The learning unit is configured to receive the open signal generated in the first time segment to record and calculate an estimated behavior mode according to the time information corresponding to the open signal and the type of the home appliance. The comparing unit is configured to compare the actual behavior record of the second time zone with the estimated behavior mode, and selectively send the first prompt message according to the comparison result.

Description

Smart home monitoring system and smart home monitoring method

The invention relates to a smart home monitoring system and a smart home monitoring method, in particular to a smart home monitoring system and a smart home monitoring method using a plurality of sensors.

As the social demographic structure tends to age, the proportion of the elderly population continues to rise, so the system of intelligent care for the elderly is getting more and more attention. Such intelligent care systems are mainly used to detect the behavior of elderly people, in order to determine whether the elderly are at risk or need assistance in order to achieve the purpose of care. Some smart care systems require the use of surveillance system technology, but once the surveillance system is used, it may cause privacy violations for older people.

Other smart care systems need to be paired with wearable tags, but the use of wearable tags can cause inconvenience in life and cause physical discomfort, so most elderly people wear wearable tags. The will is not high. Therefore, how to develop a smart care system that does not require the use of surveillance technology or the need to wear wearable labels to detect the behavior of elderly people is an important topic in the field.

The invention provides a smart home monitoring system and a smart home monitoring method, which can calculate a preset behavior pattern by using information collected by one or more perceptrons to compare the actual behavior of the user, thereby determining whether the user behavior is Abnormal, and achieve the purpose of care.

According to an embodiment of the invention, a smart home monitoring system is disclosed, which comprises a sensing module and a behavior determining module. The sensing module includes a plurality of door frame sensors, and each of the door frame sensors is disposed on a door frame of a corresponding one of the plurality of home appliances, and is configured to generate an opening signal correspondingly when the door frame of the corresponding home device is in an open state. . The behavior judgment module is connected to the sensing module. The behavior judgment module includes a learning unit and a comparison unit. The learning unit is configured to receive the open signal generated in the first time segment to record and calculate an estimated behavior mode according to the time information corresponding to the open signal and the type of the home appliance. The comparing unit is configured to compare the actual behavior record in the second time segment with the estimated behavior mode, and selectively send the first prompt message according to the comparison result.

In an embodiment, the sensing module further includes a plurality of dynamic sensors, and each dynamic sensor is disposed in a corresponding one of the plurality of regions, and is configured to detect a human activity state of the corresponding region. Generate an activity status signal. The learning unit further calculates an estimated behavior pattern associated with the user according to the time information and the region corresponding to the activity status signals generated in the first time segment, wherein the actual behavior record is associated with the second time segment The time information and area corresponding to the generated activity status signal.

In one embodiment, the sensing module further includes a plurality of pressure sensors, and each pressure sensor is disposed under the floor of a corresponding one of the plurality of regions to detect the walking frequency and the trajectory of the user. The learning unit further calculates the estimated behavior pattern associated with the user according to the walking frequency and the trajectory of the user generated in the first time segment, wherein the actual behavior record is associated with the second time segment The user's walking frequency and trajectory.

According to an embodiment of the present invention, a smart home monitoring method includes the following steps: receiving a plurality of open signals respectively generated by a plurality of door frame sensors in a first time zone; recording and corresponding to each open signal The time information and the type of the home appliance, the operation is associated with the estimated behavior pattern of the user; comparing the actual behavior record in the second time period with the estimated behavior pattern, and selectively transmitting the first prompt message according to the comparison result .

In an embodiment, the smart home monitoring method further includes: detecting, by the plurality of dynamic sensors, the activity status of the individual corresponding regions in the first time segment, thereby generating a plurality of active state signals; The time information and area corresponding to the active status signals generated in the first time zone and the time information corresponding to each of the open signals and the type of the home appliance, and the estimated behavior mode associated with the user The actual behavior record is associated with time information and an area corresponding to the activity status signals generated in the second time zone.

In summary, in the smart home monitoring system and the smart home monitoring method of the present invention, the plurality of related sensing information of the user can be obtained by using one or more perceptrons in a time period, and according to the related sensing The information learns a normal behavior pattern of the user. Then, the actual behavior of the user in another time period is recorded by the perceptron, and the actual behavior is compared with the normal behavior mode to determine whether the behavior of the user is abnormal, and then whether to issue a prompt message. In this way, it is possible to achieve the purpose of intelligent care without using the wearable label.

The above description of the disclosure and the following description of the embodiments of the present invention are intended to illustrate and explain the spirit and principles of the invention, and to provide further explanation of the scope of the invention.

The detailed features and advantages of the present invention are set forth in the Detailed Description of the Detailed Description of the <RTIgt; </ RTI> <RTIgt; </ RTI> </ RTI> </ RTI> <RTIgt; The objects and advantages associated with the present invention can be readily understood by those skilled in the art. The following examples are intended to describe the present invention in further detail, but are not intended to limit the scope of the invention.

Referring to FIG. 1 and FIG. 2 together, FIG. 1 is a functional block diagram of a smart home monitoring system according to an embodiment of the present invention. 2 is a schematic plan view of a home according to an embodiment of the invention. As shown in FIG. 1, the smart home monitoring system 1 includes a sensing module 10 and a behavior determining module 12 connected thereto. The sensing module 10 includes a plurality of door frame sensors DS1~DS4. The door frame sensors DS1 to DS4 are respectively disposed in the door frame of the corresponding one of the plurality of home appliances in FIG. 2 . As shown in FIG. 2, the home floor is divided into four areas AR1 to AR4, and each area is provided with a corresponding home appliance APP1~APP4. In this embodiment, for example, a washing machine, a refrigerator, a dishwasher and an oven can be used. The door frame sensors DS1 to DS4 are respectively disposed on the door frames of the home appliances APP1 to APP4. The door frame sensors DS1~DS4 are configured to generate an opening signal correspondingly when the door frame of the corresponding home appliance is turned on.

The behavior determination module 12 includes a learning unit 121 and a comparison unit 123. The learning unit 121 is configured to receive a plurality of open signals generated in the first time zone to record and calculate an estimated behavior mode associated with the user according to the time information corresponding to the open signals and the type of the home appliance. In practice, the learning unit 121 can be a machine with a learning calculus function, which can obtain a regularity by automatic data analysis, and predict unknown data through the law. The first time period described herein may be a time period of two weeks, one month, two months, etc., but the invention is not limited thereto. However, in practice, if the first time period is longer, the more learning data acquired by the learning unit 121, the more accurate the estimated behavior pattern calculated.

The comparing unit 123 is configured to compare the actual behavior record in a second time segment with the estimated behavior mode, and selectively send the first prompt message according to the comparison result. The second time period described herein may be a time period of several hours to several tens of hours, but the invention is not limited thereto. In actual operation, when there is a difference between the actual behavior record and the estimated behavior mode, the comparison unit 123 sends the first prompt message to the remote device, such as a smart phone or a secured background device.

Specifically, in the pre-operation of the system, the learning unit 121 in the behavior determining module 12 is first set to make time information and home appliances corresponding to the opening signals collected by the door frame sensors DS1 to DS4. The type of device is used to learn, and then the estimated behavior pattern of the user is estimated. The time information and the type of the home appliance may be collected by a preset learning period (for example, the aforementioned first time section), and the learning period may be, for example, a period of two weeks or one month. In practice, it is assumed that the user's normal work mode is: about 7:00 in the morning, the area AR1 will use the home appliance APP1, about 8:30 in the morning to the area AR2 to use the home appliance APP2, and about 12 noon will go to the area AR3. The home appliance APP4 is used, and the home appliance APP4 is used to the area AR4 at about 3 pm. After the time information corresponding to the plurality of open signals generated by the door frame sensor DS1~DS4 for a period of time (for example, two weeks) and the type of the home appliance, the learning unit 121 can derive the estimated behavior mode of the user, and The estimated behavior pattern of the user is transmitted to the comparison unit 123.

After the user's behavior pattern is acquired through the learning of the learning unit 121, the smart home monitoring system can then begin intelligent monitoring. For example, in a period of about 12 noon on a certain day (for example, a second time period), the door frame sensor DS3 in the area AR3 does not detect that the user opens the door frame of the home appliance APP3. The comparison unit 123 can know the actual behavior record of the user on the certain day by comparing the actual behavior record of the user (the door frame of the home appliance APP3 is not turned on in the period of 12 noon) with the estimated behavior pattern. The original estimated behavior patterns are different. At this time, the comparison unit 123 issues a first prompt message, so that the user's relatives or friends or related security operators can confirm the situation of the user. In one example, each of the open signals has an identification code associated with a corresponding one of the home appliances. The learning unit 121 can determine, according to the identification code, the type of the home appliance corresponding to the activation signal. For example, the activation signal generated by the home appliance APP1 is provided with an identification code X001, which is preset to be associated with the home appliance APP1. When the learning unit 121 receives and recognizes that the identification code system is X001, it can be known that the home appliance that is turned on is the home appliance APP1.

Please refer to FIG. 3 further. FIG. 3 is a functional block diagram of a smart home monitoring system according to another embodiment of the present invention. As shown in FIG. 3, the sensing module of the smart home monitoring system 1 further includes a plurality of dynamic sensors MS1 to MS4. The motion sensors MS1 to MS4 are respectively disposed in the plurality of areas AR1 to AR4. The dynamic sensors MS1~MS4 are mainly used to detect the activity status of the corresponding area, and generate an activity status signal accordingly. Specifically, the motion sensors MS1 to MS4 can be set at the door edge to detect which areas the user enters and exits at which time periods. For example, the user may be fixed at about one o'clock every afternoon to enter the area AR2 nap, until about two o'clock in the afternoon will come out from the area AR2. These activity situations are detected by the motion sensor MS2 and transmitted to the learning unit 121.

Similarly, the dynamic perceptrons MS1~MS4 collect the plurality of active status signals of the user in the first time zone (for example, two weeks), and the learning unit 121 can generate the multiple active status signals according to the time zone. The time information and area corresponding to the activity status signal are matched with the time information corresponding to the opening signal and the type of the home appliance to learn, thereby calculating the estimated behavior pattern associated with the user.

Then, in the second time zone, at least some of the dynamic sensors MS1 - MS4 detect and generate a plurality of active status signals, and the learning unit 121 records the time information and the area corresponding to the active status signals as actual behavior records. . Finally, the user's estimated behavior pattern is compared to the actual behavior record to decide whether to send the first prompt message. In other words, in addition to the application of the door frame perceptron, the present embodiment is further equipped with a dynamic sensor, thereby establishing a user according to the time information corresponding to the activation signal, the type of the home appliance, and the time information and area corresponding to the activity status signal. Estimate behavior patterns to substantially enhance the effectiveness of surveillance and care.

In an embodiment, as shown in FIG. 3, the sensing module further includes a plurality of sound sensors VS1~VS4, which are respectively disposed in the plurality of areas AR1~AR4. The sound sensors VS1~VS4 have the function of detecting sounds in the environment to generate sound information. More specifically, the sound sensors VS1~VS4 can be used to detect sounds with abnormal volume, such as short loud sounds. The behavior judgment module includes a sound database 125, which stores a plurality of context sound information. In practice, the sound database 125 stores abnormal situation sound information, such as a human body hitting the ground sound or an abnormal tapping sound.

In this embodiment, the behavior determining module 12 is configured to compare the sound information detected by the sound sensors VS1 VS VS4 with the context sound information to selectively issue a warning message. When any of the plurality of sound sensors VS1 to VS4 detects the sound information, and the voiceprint of the sound information is similar to or the same as the voiceprint of any of the plurality of context sound information, it is determined that the user may have an accident. The behavior determination module 12 issues a warning message accordingly.

In the embodiment of FIG. 1 and FIG. 2, when the estimated behavior pattern is different from the actual behavior record, a first prompt message is generated, and the first prompt message can be transmitted to an external device, so that the user can get instant Rescue. However, sometimes the actual behavior pattern of the user in a single period of time is different from the estimated behavior pattern, but it is not that the user actually has an accident, but only for other reasons and does not follow the normal behavior pattern. At this time, if the first prompt message is directly sent according to the result of the inconsistent behavior mode of the single time period, the trigger message may be triggered.

In view of this, in an embodiment, in addition to comparing the foregoing actual behavior record to selectively send the first prompt message, the comparing unit 123 is further configured to use the estimated behavior mode and another one in the third time zone. The actual behavior records are compared, and the second prompt message is selectively sent based on the comparison result. The third time period is a time period after the second time period. The smart home monitoring system further includes a backend device 14 that connects the behavior determination module 12 and the rescue unit 2. In practice, the ambulance unit 2 can be a system of adjacent medical units where the user is located. The backend device 14 is configured to notify the rescue unit 2 to perform rescue when receiving the first prompt message and receiving the second prompt message.

In other words, in this embodiment, if the backend device 14 only receives the prompt message sent by the single time zone, the rescue request is not sent to the rescue unit 2, but is sent continuously in two periods. When the message is prompted, the rescue request will be formally presented to the rescue unit 2. In this way, it is possible to avoid wasting rescue resources based on a single prompt message, suggesting unnecessary rescue needs. In other examples, the user can set the backend device 14 to receive the prompt message sent by three or more time periods according to actual needs, and then notify the rescue unit 2 to perform the rescue. Of course, in an embodiment, if necessary, the user can also set the backend device 14 to receive the prompt message sent by the single time zone, and directly request the rescue unit 2 for the rescue.

Please refer to FIG. 3 and FIG. 4 together. FIG. 4 is a schematic plan view of a home according to another embodiment of the present invention. In this embodiment, as shown in FIG. 3 and FIG. 4, the sensing module 10 further includes a plurality of pressure sensors PS1~PS25, and the pressure sensors PS1~PS25 are disposed under the floor of the corresponding one of the plurality of areas. Used to detect the user's walking frequency and trajectory. In practice, when the user walks on the ground, a pressure sensor of a part of the pressure sensors PS1~PS25 is triggered. For example, if the user walks from the area A3 to the AR4 according to the trajectory TK according to a specific walking speed, the pressure sensors PS15, PS18, PS20, PS21, and PS22 are triggered one by one to cause the sensors to generate pressure. The user's walking frequency and trajectory are obtained by the signal, and are transmitted back to the learning unit 121.

Similarly, in this embodiment, in the initial pre-operation phase, the learning unit 121 first receives and detects the use according to at least some of the pressure sensors in the first time segment (for example, two weeks). The walking frequency and trajectory and their corresponding time information are calculated in association with the user's estimated behavior pattern. Then, in the second time period, at least a part of the pressure sensor detects the actual walking frequency and the trajectory of the user in the second time zone and the corresponding time information, and the learning unit 121 incorporates the user into the user. Actual behavior record. More specifically, it is assumed that the user usually moves from the area A3 to the AR4 following the trajectory TK according to the inertial walking speed at about a certain time. If the day is about the specific time, the pressure sensor does not measure the walking record (including the walking frequency and the trajectory), or the walking record is obviously different (for example, the walking frequency suddenly becomes slow or the trajectory stays at a certain pressure) When the sensor is on the floor, it may cause an accident on behalf of the user. At this point, the system will send a first prompt message, so that the user's relatives or friends or the security practitioner can immediately confirm the user's situation to avoid the opportunity to miss the rescue.

In short, in the foregoing embodiment, the estimated behavior mode includes time information corresponding to the plurality of open signals of the first time zone, the type of the home appliance, the user's walking frequency and the trajectory, and the corresponding time. News. The embodiment of the present invention calculates the estimated behavior pattern through the information collected by the two types of perceptrons, thereby achieving more accurate monitoring and care effects.

Considering the importance of home security, the system can be equipped with anti-theft monitoring functions in addition to the care function. In an embodiment, as shown in FIG. 2 and FIG. 3, the sensing module 10 further includes a plurality of anti-theft sensors SS1 SS SS4, which are respectively disposed in corresponding regions. The anti-theft sensors are used when the smart home monitoring system 1 is set to the maintenance monitoring mode. More specifically, when the user is at home, the smart home monitoring system 1 is in the care monitoring mode, the door frame sensor and the behavior determining module are in the running state, but the anti-theft sensor is in the disabled state. When the user is ready to go out, the smart home monitoring system 1 can be switched to the security monitoring mode to enable the anti-theft sensors to be enabled. When the enabled anti-theft sensors detect that there is a person entering or exiting or an abnormal sound, the alarm signal of the security device 30 is triggered to notify the user. In the security monitoring mode, the door frame sensor and the behavior determining module are automatically switched to the disabled state to avoid false triggering of the prompt message.

Please refer to FIG. 5. FIG. 5 is a flowchart of a method for a smart home monitoring method according to an embodiment of the present invention, which is applicable to the smart home monitoring system of FIG. 1. As shown in FIG. 5, the monitoring method includes: in step S501, the learning unit 121 receives a plurality of opening signals respectively generated by the plurality of door frame sensors DS1 DSDS4 in the first time zone. Next, in step S503, the learning unit 121 records and calculates an estimated behavior pattern associated with the user according to the time information corresponding to each of the activation signals and the type of the home appliance. Next, in step S505, the comparing unit 123 compares the actual behavior record in the second time zone with the estimated behavior mode, and selectively transmits the first prompt message according to the comparison result. In an embodiment, the smart home monitoring method further includes detecting, by the plurality of dynamic perceptrons MS1 MS MS4, the activity status of the individual corresponding regions in the first time segment, thereby generating a plurality of active states. The signal is then calculated according to the time information and the area corresponding to the activity status signal generated in the first time zone, and the time information corresponding to the opening signal and the type of the home appliance, and the pre-related to the user is calculated. Estimate behavior patterns. The actual behavior record is associated with the time information and area corresponding to the activity status signal generated in the second time zone.

Please refer to FIG. 6. FIG. 6 is a flowchart of a method for a smart home monitoring method according to another embodiment of the present invention, which is applicable to the smart home monitoring system of FIG. 3. Steps S601 to S605 of the smart home monitoring method shown in FIG. 6 are similar to steps S501 to S505 of the smart home monitoring method shown in FIG. 5. FIG. 6 is different from FIG. 5 in that FIG. 6 further includes: in step S607, the comparing unit 123 further compares another actual behavior record in the third time segment with the predicted behavior mode and according to The comparison result selectively transmits a second prompt message, wherein the third time zone is a time zone subsequent to the second time zone. Next, in step S609, when the backend device 14 receives the first prompt message and receives the second prompt message, the ambulance unit 20 is notified to perform the rescue.

In summary, the smart home monitoring system and the smart home monitoring method of the present invention mainly acquire a plurality of relevant perceptual information of a user in a learning time section through a plurality of perceptrons, and according to the related perceptual information. Use learning calculus to push a normal behavior pattern of the user. Then, the actual behavior of the user in another time period is recorded by the perceptron, and the actual behavior is compared with the normal behavior mode to determine whether the behavior of the user is abnormal in the other time period, thereby determining whether to issue Prompt message. In this way, effective smart care can be achieved without using wearable tags or without applying surveillance system technology.

Although the present invention has been disclosed above in the foregoing embodiments, it is not intended to limit the invention. It is within the scope of the invention to be modified and modified without departing from the spirit and scope of the invention. Please refer to the attached patent application for the scope of protection defined by the present invention.

1‧‧‧Smart home monitoring system

10‧‧‧Sense module

12‧‧‧ Behavioral Judgment Module

121‧‧‧Learning unit

123‧‧‧Comparative unit

125‧‧‧Sound database

14‧‧‧ Back-end equipment

20‧‧‧ambulance unit

30‧‧‧Security device

AR1~AR4‧‧‧Area

APP1~APP4‧‧‧Home Appliances

DS1~DS4‧‧‧ Door Frame Sensor

MS1~MS4‧‧‧Dynamic Sensor

VS1~VS4‧‧‧Sound Sensor

PS1~PS25‧‧‧ Pressure Sensor

SS1~SS4‧‧‧Anti-theft sensor

TK‧‧ track

FIG. 1 is a functional block diagram of a smart home monitoring system according to an embodiment of the invention. 2 is a schematic plan view of a home according to an embodiment of the invention. FIG. 3 is a functional block diagram of a smart home monitoring system according to another embodiment of the present invention. 4 is a schematic plan view of a home according to another embodiment of the present invention. FIG. 5 is a flow chart of a method for a smart home monitoring method according to an embodiment of the invention. FIG. 6 is a flow chart of a method for a smart home monitoring method according to another embodiment of the present invention.

Claims (8)

A smart home monitoring system includes: a sensing module, comprising a plurality of door frame sensors, each of the door frame sensors being disposed on a door frame of a corresponding one of the plurality of home appliances, for corresponding to the home appliance When the door frame is in an open state, an opening signal is generated correspondingly; a behavior determining module is connected to the sensing module, and the behavior determining module includes: a learning unit, configured to receive the generated in a first time segment Turning on the signals to record and calculate an estimated behavior pattern associated with a user according to the time information and the type of the home appliance corresponding to each of the open signals; and a comparing unit for using a second time zone An actual behavior record in the segment is compared with the estimated behavior pattern, and a first prompt message is selectively sent according to the comparison result, and the comparing unit is further configured to use the estimated behavior pattern and a third time segment. Comparing another actual behavior record, and selectively transmitting a second prompt message according to the comparison result; and a backend device connecting the behavior judgment And the ambulance unit, the backend device is configured to notify the ambulance unit to perform rescue when receiving the first prompt message and receiving the second prompt message; wherein the third time zone is the second time The time period after the section. The smart home monitoring system of claim 1, wherein the sensing module further comprises: a plurality of dynamic sensors, each of the dynamic sensors being disposed in a corresponding one of the plurality of regions for detecting the corresponding The state of personnel activity in the area, according to which an activity status message is generated The learning unit further calculates the estimated behavior pattern associated with the user according to the time information and the region corresponding to the activity status signals generated in the first time segment, wherein the actual behavior record is associated with The time information and area corresponding to the active status signals generated in the second time zone. The smart home monitoring system of claim 1, wherein the sensing module further comprises a plurality of sound sensors, each of the sound sensors being disposed in a corresponding one of the plurality of regions for detecting sound in the environment To generate sound information, the behavior determination module includes a sound database, and stores a plurality of context sound information, and the behavior determination module is configured to respectively use the sound information detected by the sound sensors to the context sounds. The information is compared to selectively issue a warning message. The smart home monitoring system of claim 1, wherein each of the open signals has an identification code associated with a corresponding one of the home appliances, and the learning unit determines, according to the identification code, the home appliance corresponding to the open signal The type of equipment. The smart home monitoring system of claim 1, wherein the sensing module further comprises a plurality of pressure sensors, each of the pressure sensors being disposed under a floor of a corresponding one of the plurality of areas for detecting the The learning frequency and the trajectory of the user, the learning unit further calculates the estimated behavior pattern associated with the user according to the walking frequency and the trajectory of the user generated in the first time segment, wherein the actual behavior record Corresponding to the walking frequency and trajectory of the user generated in the second time zone. The smart home monitoring system of claim 1, wherein the sensing module further comprises a plurality of anti-theft sensors for enabling the smart home monitoring system when the security monitoring mode is set to a security monitoring mode. When the anti-theft sensor detects that a person enters or exits or an abnormal sound, The alarm signal of the security device is triggered to notify a user terminal, wherein the door frame sensor and the behavior determination module are disabled in the security monitoring mode. A smart home monitoring method includes: receiving a plurality of open signals respectively generated by a plurality of door frame sensors in a first time zone; recording and according to time information corresponding to each of the open signals and a type of home appliances, The operation is associated with a predicted behavior pattern of a user; comparing an actual behavior record in a second time period with the estimated behavior pattern, and selectively transmitting a first prompt message according to the comparison result A further actual behavior record in the third time zone is compared with the predicted behavior mode and a second prompt message is selectively sent according to the comparison result, wherein the third time zone is the second time zone a subsequent time segment; and when the first prompt message is received and the second prompt message is received, an ambulance unit is notified to perform the rescue. The smart home monitoring method of claim 7, further comprising: detecting, by the plurality of dynamic sensors, the activity status of the individual corresponding one area in the first time zone, thereby generating a plurality of activity status signals; And calculating the pre-related to the user according to the time information and the area corresponding to the activity status signals generated in the first time period and the time information and the type of the home appliance corresponding to each of the activation signals The behavioral mode is estimated; wherein the actual behavior record is associated with time information and an area corresponding to the activity status signals generated in the second time zone.