JP7789229B2 - Elevator operation prediction system - Google Patents

Description

This disclosure relates to an elevator operation prediction system.

A conventional elevator congestion prediction and display system includes a collected data storage unit, an elevator operation simulation unit, and a display unit. The collected data storage unit stores past elevator operation information and past people flow information. The elevator operation simulation unit predicts the elevator congestion level when planned elevator construction is carried out based on the information stored in the collected data storage unit. The display unit displays the congestion level predicted by the elevator operation simulation unit (for example, Patent Document 1).

Japanese Patent Application Laid-Open No. 2021-89454

Conventional congestion prediction and display systems do not take into account events specific to apartment complexes, such as garbage collection days, disaster prevention drills, and seasonal events. Therefore, there is a need to provide users with more appropriate information regarding elevator operation.

This disclosure has been made to solve the above-mentioned problems and aims to provide an elevator operation prediction system that can provide users with more appropriate information regarding elevator operation.

The elevator operation prediction system of the present disclosure includes a prediction management device that makes predictions regarding elevator operations in an apartment building, and a management terminal that acquires event schedule information, which is information about events scheduled in the apartment building. The prediction management device stores operation history information, which is the history of elevator operations, user number history information, which is the history of the number of elevator users by hour, and event history information, which is the history of events that have previously taken place in the apartment building, and predicts the degree of congestion in the elevator based on the operation history information, user number history information, event history information, and the event schedule information acquired by the management terminal, and creates display data for the predicted degree of congestion.
The system is equipped with a prediction management device that makes predictions regarding elevator operation, and a display device. The prediction management device calculates the estimated time until the elevator car arrives at the landing based on current operation information, which is information on the current operating status of the elevator, and current user number information, which is information on the current number of elevator users, and creates display data for the calculated estimated time. The display device displays the estimated time based on the display data created by the prediction management device.

The elevator operation prediction system disclosed herein can provide users with more appropriate information regarding elevator operation.

1 is a block diagram showing an example of the overall configuration of an elevator system according to a first embodiment. FIG. 2 is a diagram illustrating the projection display device shown in FIG. 1. 2 is a block diagram showing the configuration of the prediction management device shown in FIG. 1. FIG. 4 is a flowchart showing a learning process of the prediction management device shown in FIG. 3 . 4 is a flowchart showing a congestion degree prediction process of the prediction management device shown in FIG. 3 . FIG. 10 is a block diagram showing the configuration of a prediction management device according to a second embodiment. 7 is a flowchart showing an estimated arrival time calculation process of the prediction management apparatus shown in FIG. 6 . FIG. 2 is a configuration diagram showing a first example of a processing circuit that realizes the prediction management device according to the first and second embodiments. FIG. 10 is a configuration diagram showing a second example of a processing circuit that realizes the prediction management device according to the first and second embodiments.

The following describes the embodiments with reference to the drawings.

Embodiment 1.
1 is a diagram showing an example of the overall configuration of an elevator system according to Embodiment 1. The elevator system 1 includes an elevator operation prediction system 100 and an elevator 200.

The elevator 200 is installed in an apartment building. The elevator 200 includes an elevator car 210, an elevator control device 220, a landing detection device 230, and an outside-car display device 240.

The elevator car 210 is equipped with an in-car detection device 211 and an in-car display device 212.

The car interior detection device 211 is a surveillance camera that captures images of the inside of the elevator car 210. The car interior detection device 211 outputs the captured images to the elevator control device 220.

The car interior display device 212 is a display device installed on an operation panel inside the elevator car 210. The car interior display device 212 provides congestion level information to users inside the elevator car 210. Here, the congestion level is an index value that represents the degree of congestion in the elevator 200.

The platform detection device 230 is a surveillance camera installed at the elevator platform on each floor of the apartment building. The platform detection device 230 outputs the captured image of the elevator platform to the elevator control device 220. The car exterior display device 240 is a display device installed, for example, at the elevator platform. The car exterior display device 240 provides congestion information to users waiting for the elevator car 210 to arrive. Here, the car interior display device 212 and the car exterior display device 240 may be equipped with speakers and may provide audio notifications to users.

The elevator control device 220 controls the operation of the elevator 200. The elevator control device 220 also controls the display of the in-car display device 212 and the outside-car display device 240. If the in-car display device 212 and the outside-car display device 240 are equipped with speakers, the elevator control device 220 also controls the audio output of the in-car display device 212 and the outside-car display device 240.

The elevator control device 220 also performs image analysis on the images received from the car interior detection device 211 to detect the number of passengers in the elevator car 210. The elevator control device 220 also performs image analysis on the images received from the hall detection device 230 to detect the number of passengers in the elevator hall for each floor. Note that a device dedicated to image analysis may be used separately from the elevator control device 220 to detect the number of passengers in the elevator car 210 and the number of passengers at the elevator hall.

The elevator operation prediction system 100 includes a prediction management device 110, a management terminal 120, and a projection display device 130.

The management terminal 120 is a terminal device used by an administrator who manages the elevator operation prediction system 100 or an administrator who manages an apartment building. The administrator inputs information about events scheduled at the apartment building into the management terminal 120. The administrator also inputs information about events that have previously taken place at the apartment building into the management terminal 120.

Here, events include garbage collection days, disaster prevention drills, site cleanups, management association meetings, etc. Events also include seasonal events and social gatherings held in apartment complexes, such as rice cake pounding competitions, Christmas parties, and barbecues.

The administrator enters event information, such as the name of the event and the date and time of the event, into the management terminal 120.

The prediction management device 110 makes predictions regarding the operation of the elevator 200. In embodiment 1, the prediction management device 110 predicts the degree of congestion of users using the elevator 200.

The prediction management device 110 is also connected to the elevator control device 220. The prediction management device 110 obtains the number of passengers in the elevator car 210 and the number of passengers at the elevator hall from the elevator control device 220.

The prediction management device 110 also obtains operation information for the elevator 200 from the elevator control device 220. Here, the operation information includes status information on whether the elevator car 210 is moving or stopped. The operation information also includes information on the direction of travel of the elevator car 210. The operation information also includes call information. The call information includes the destination floor specified by a user in the elevator car 210 and the destination direction specified by a user at the elevator hall.

The prediction management device 110 also displays the predicted congestion level on the car in-car display device 212 via the elevator control device 220. The prediction management device 110 also displays the predicted congestion level on the projection display device 130. The prediction management device 110 also displays the predicted congestion level on the car out-car display device 240 via the elevator control device 220.

The prediction management device 110 also has the functionality of a web server. When the prediction management device 110 receives a request from the mobile terminal 300 to transmit the congestion level, it transmits the predicted congestion level of the elevator 200 to the mobile terminal 300.

Here, the mobile terminal 300 is a display device such as a smartphone or tablet carried by a resident of the apartment complex. The mobile terminal 300 receives and displays the congestion level of the elevator 200 from the prediction management device 110 via a web browser.

Communication between the prediction management device 110 and the mobile terminal 300 may be performed via the Internet network or via a local network.

In addition, the prediction management device 110 can also respond when requests are received from multiple mobile terminals 300.

The projection display device 130 is a display device installed in common spaces such as the entrance, lobby, and common corridors of an apartment building. The projection display device 130 is equipped with a projection unit 131.

Figure 2 is a diagram showing the projection display device 130 shown in Figure 1. The projection unit 131 projects an image onto the floor of the first floor entrance 400, for example. As a result, the projection display device 130 can provide users near the projection surface 132 with a visually easy-to-understand representation of the congestion level predicted by the prediction management device 110. The image displayed on the projection surface 132 may be a still image or a video. Furthermore, the image displayed on the projection surface 132 may be monochrome or color.

Furthermore, by being equipped with multiple projection display devices 130, the elevator operation prediction system 100 is able to project the congestion level predicted by the prediction management device 110 at multiple locations. In addition, by being equipped with an outside-car display device 240 at the elevator hall on each floor, the elevator operation prediction system 100 is able to display the congestion level predicted by the prediction management device 110 at multiple locations.

Figure 3 is a block diagram showing the configuration of the prediction management device 110 shown in Figure 1.

The prediction management device 110 includes a receiving unit 111, a memory unit 112, a congestion level learning unit 113, a congestion level prediction unit 114, a display data creation unit 115, and a transmission unit 116.

The receiving unit 111 receives the number of passengers in the elevator car 210, the number of passengers at the elevator hall on each floor, and operation information from the elevator control device 220. This information is transmitted sequentially from the elevator control device 220.

In addition, the receiving unit 111 receives information about scheduled events and information about events that have taken place in the past from the management terminal 120.

The memory unit 112 associates the operation information received by the receiving unit 111 with the date and time information at the time of reception and stores the information as operation history information 121. Here, the date and time information is information on the year, month, day, hour, minute, and second.

The memory unit 112 associates the number of passengers in the car received by the receiving unit 111 with the date and time information at the time of reception. The memory unit 112 also associates the number of passengers at each floor's landing received by the receiving unit 111 with the date and time information at the time of reception. The memory unit 112 stores this information together as passenger number history information 122.

The memory unit 112 stores a history of events that have taken place in the apartment complex in the past as event history information 123. The memory unit 112 also stores information about events scheduled for the apartment complex in the form of event schedule information 125.

The congestion level learning unit 113 performs machine learning based on operation history information 121, user number history information 122, and event history information 123. The congestion level learning unit 113 stores the results of the machine learning in the memory unit 112 as learning results 124.

The congestion level learning unit 113 classifies the data in the operation history information 121 into days on which an event occurred and days on which no event occurred, based on the event history information 123. Similarly, the congestion level learning unit 113 classifies the data in the number of users history information 122 into days on which an event occurred and days on which no event occurred, based on the event history information 123. As a result, one group is formed by combining the operation history information 121 for days on which an event occurred and the number of users history information 122 for days on which an event occurred. In addition, another group is formed by combining the operation history information 121 for days on which no event occurred and the number of users history information 122 for days on which no event occurred.

The congestion learning unit 113 creates learning results based on groups of days on which no events occurred. These learning results created based on days on which no events occurred are referred to as normal day learning results.

For groups on days when an event occurred, the congestion learning unit 113 further subdivides the groups into subgroups by event name and creates learning results for each subgroup. This makes it possible to obtain learning results for each event, such as learning results for garbage collection days, disaster prevention drill days, and site cleaning days. The learning results created in this way are referred to as event day learning results.

In this way, the congestion level learning unit 113 creates learning results 124 that include learning results for normal days and learning results for event days. Note that the data structure of the learning results 124 is merely an example and is not limited to this.

The congestion prediction unit 114 predicts the congestion level of the elevator 200 for every 10 to 15 minutes on the specified day based on the learning results 124 and the event schedule information 125.

The congestion prediction unit 114 classifies the specified day as either a scheduled event day or a normal day without an event based on the event schedule information 125. If the specified day is a normal day, the congestion prediction unit 114 predicts the congestion level of the elevator 200 using the normal day learning results.

If the specified date is a scheduled event date, the congestion prediction unit 114 obtains the event date learning results corresponding to the name of the scheduled event from the memory unit 112. The congestion prediction unit 114 then predicts the congestion level of the elevator 200 using the obtained event date learning results.

The congestion prediction unit 114 predicts the passenger density in the elevator car 210 and the landing density at each floor as the congestion degree of the elevator 200. The passenger density is calculated by dividing the number of passengers in the elevator car 210 by the maximum capacity. The landing density is calculated by dividing the number of passengers at the elevator landing by a predetermined set number of people.

As such, using boarding density and boarding area density as congestion levels is merely an example and is not limited to this.

The display data creation unit 115 creates display data for the congestion level predicted by the congestion level prediction unit 114. The display data creation unit 115 creates HTML (Hypertext Markup Language) data containing the congestion level value as display data. The display data creation unit 115 also creates image data incorporating the congestion level value as display data.

The transmitting unit 116 transmits the HTML data created as display data to the mobile terminal 300. The transmitting unit 116 also transmits the image data created as display data to the projection display device 130. The transmitting unit 116 also transmits the image data created as display data to the car interior display device 212 via the elevator control device 220. The transmitting unit 116 also transmits the image data created as display data to the car exterior display device 240 via the elevator control device 220.

Figure 4 is a flowchart showing the learning process of the predictive management device 110 shown in Figure 3. The flowchart shown in Figure 4 is performed periodically, such as every hour or every day.

In step S101, the congestion level learning unit 113 acquires operation history information 121, user number history information 122, and event history information 123 stored in the memory unit 112.

In step S102, the congestion level learning unit 113 performs machine learning based on operation history information 121, user number history information 122, and event history information 123, and creates learning results 124.

In step S103, the memory unit 112 stores the learning result 124.

Figure 5 is a flowchart showing the congestion prediction process of the prediction management device 110 shown in Figure 3. Figure 5 shows the process of displaying the congestion degree of the elevator 200 on a specified date on the mobile terminal 300. This specified date is information about a date specified by the person operating the mobile terminal 300. This specified date is referred to as the specified date to be predicted.

In addition, the flowchart in Figure 5 is performed each time a congestion level request is received from the mobile terminal 300.

In step S201, the congestion prediction unit 114 obtains the prediction target date specified by the operator of the mobile terminal 300.

In step S202, the congestion prediction unit 114 classifies the designated date to be predicted as either a normal day or a scheduled event date based on the event schedule information 125.

In step S203, the congestion prediction unit 114 predicts the congestion level of the elevator 200 on the specified prediction date. If the specified prediction date is classified as a normal day, the congestion prediction unit 114 predicts the congestion level using the normal day learning results. On the other hand, if the specified prediction date is classified as a scheduled event date, the congestion prediction unit 114 predicts the congestion level using the event implementation date learning results.

In step S204, the display data creation unit 115 creates HTML data including the congestion degree as display data. In step S205, the transmission unit 116 transmits the created HTML data to the mobile terminal 300.

In the flowchart of Figure 5, the congestion prediction unit 114 predicts the congestion level of the elevator 200 by specifying a daily unit. In contrast, the congestion prediction unit 114 can predict the congestion level by specifying a weekly, monthly, or annual unit up to one year.

When predicting congestion levels on a weekly basis, for example, the congestion level prediction unit 114 classifies each day that makes up the week as either a normal day or a day on which an event is scheduled to take place, and then predicts the congestion level. In this case, the congestion level may be predicted every 10 to 15 minutes, as described above, or may be aggregated by hour, day, etc.

The congestion prediction unit 114 can also predict the congestion level by specifying a time or time period. In this case, the congestion prediction unit 114 predicts the congestion level at the specified time or time period on the day when the request is made.

When the congestion prediction unit 114 predicts the congestion level to be displayed by the projection display device 130 and the car exterior display device 240, it executes step S201, using the day on which the flowchart in Figure 5 is executed as the designated prediction target day. Similarly, for the car interior display device 212, it executes step S201, using the day on which the flowchart in Figure 5 is executed as the designated prediction target day.

In addition, when the congestion degree prediction unit 114 predicts the congestion degree to be displayed by the projection display device 130, the car interior display device 212, and the car exterior display device 240, it executes the flowchart in Figure 5 periodically, such as every hour.

In embodiment 1, the prediction management device 110 predicts the congestion level of the elevator 200 based on operation history information 121, user number history information 122, event history information 123, and event schedule information 125 acquired by the management terminal 120. The prediction management device 110 also creates display data for the predicted congestion level.

As a result, the elevator operation prediction system 100 in embodiment 1 can predict congestion levels taking into account events specific to apartment buildings. Therefore, the elevator operation prediction system 100 in embodiment 1 can provide users with more appropriate information regarding the operation of the elevator 200.

An apartment building is a private space for its residents. For this reason, residents desire to avoid meeting each other face-to-face. They also desire to use the elevator 200 when it is not crowded. The elevator operation prediction system 100 of embodiment 1 allows residents of the apartment building to time their use of the elevator 200 to when it is not crowded.

The elevator operation prediction system 100 also includes a projection display device 130 having a projection unit 131. The projection unit 131 projects an image created as display data by the prediction management device 110 onto the entrance 400 of the apartment building. This makes it possible to notify the residents of the apartment building of the degree of congestion of the elevator 200. The elevator operation prediction system 100 also displays an image created as display data on the outside-car display device 240. This makes it possible to further notify the residents of the apartment building of the degree of congestion of the elevator 200.

In addition, the projection unit 131 projects the image created as display data onto the floor of the entrance 400. As a result, the elevator operation prediction system 100 can display the congestion level of the elevator 200 in a location where there are no walls, such as the center of the entrance 400 where occupants often pass through. In addition, the elevator operation prediction system 100 displays the congestion level of the elevator 200 on the frequently used outside-car display device 240, making the congestion level more widely known.

The prediction management device 110 also performs machine learning based on operation history information 121, user number history information 122, and event history information 123. The prediction management device 110 also predicts the congestion level of the elevator 200 based on the machine learning learning results 124 and event schedule information 125. As a result, the elevator operation prediction system 100 can obtain more accurate congestion level prediction results.

In addition to the above, the prediction management device 110 may also predict the congestion level of the elevator 200 based on a weather forecast. In this case, the prediction management device 110 stores weather history information, which is a history of past weather conditions. The prediction management device 110 then acquires weather forecast information from the management terminal 120 or an external system. The prediction management device 110 then predicts the congestion level of the elevator 200 based on the weather history information and weather forecast information in addition to the operation history information 121, passenger count history information 122, event history information 123, and event schedule information 125. This allows the elevator operation prediction system 100 to obtain more accurate congestion prediction results.

Furthermore, while passenger count history information 122 is assumed to be information that includes both the number of passengers in elevator car 210 per hour and the number of passengers at the landings on each floor per hour, this is not limited to this. Passenger count history information 122 is determined depending on the desired congestion level. For example, if the desired congestion level is only the boarding density of elevator car 210, passenger count history information 122 will be only the number of passengers in elevator car 210 per hour. Furthermore, if the desired congestion level is only the landing density, passenger count history information 122 will be only the number of passengers at the landings on each floor per hour. Furthermore, if the desired congestion level is only the landing density on a specific floor, passenger count history information 122 will be only the number of passengers at the landings on a specific floor per hour.

The car interior detection device 211 may also be a load cell. In this case, the car interior detection device 211 measures the load weight to determine the number of passengers in the elevator car 210.

Furthermore, after an event has been held, the event schedule information 125 may be added directly to the existing event history information 123. This allows the administrator to avoid the need to enter past events into the management terminal 120.

Also, if there is luggage in the elevator car 210, the luggage may be added to the number of passengers depending on the size, weight, etc. of the luggage.

In addition, the car interior display device 212, the projection display device 130, and the car exterior display device 240 may periodically switch screens to sequentially display the congestion level for that day, the congestion level for the week, and the congestion level for the month.

Embodiment 2.
6 is a block diagram showing the configuration of a prediction control device 110 in the second embodiment. The overall configuration of the elevator system 1 is the same as that in the first embodiment.

The prediction management device 110 in embodiment 2 is equipped with an estimated arrival time calculation unit 117.

The estimated arrival time calculation unit 117 calculates the estimated time until the elevator car 210 arrives at each floor. The estimated arrival time calculation unit 117 calculates the estimated arrival time at each floor based on the current number of occupants in the car information 126, the current landing information 127, and the current operation information 128.

Here, current car occupancy information 126 is information on the number of passengers currently in the elevator car 210. Current hall information 127 is information on the number of passengers currently at the hall on each floor. Current operation information 128 is information on the current operation status of the elevator 200. These pieces of information are sequentially transmitted from the elevator control device 220 and stored in the memory unit 112. Furthermore, current passenger number information 129 is information on the current number of passengers in the elevator 200, and is information that includes current car occupancy information 126 and current hall information 127.

The calculation method of the estimated arrival time calculation unit 117 will now be explained in detail. Here, we consider the example of a 10-story apartment building with no basement, and the case of calculating the estimated arrival time to the first floor. Also, we assume that the elevator car 210 is currently located between the fifth and fourth floors and is moving downward.

If there are no passengers in the elevator car 210, the estimated arrival time to the first floor depends on whether there are passengers at the landings on the fourth, third, or second floors who have specified a downward direction. On the other hand, the status of the landings on the first floor and the fifth to tenth floors does not affect the estimated arrival time to the first floor. Therefore, the estimated arrival time calculation unit 117 excludes the status of the landings on the first floor and the fifth to tenth floors from consideration.

The estimated arrival time calculation unit 117 determines whether there are any downward calls from the fourth floor to the second floor. If there are downward calls from the third floor, the estimated arrival time calculation unit 117 calculates the dwell time on the third floor based on the number of passengers at the third floor hall. The more passengers there are at the hall, the longer it takes for passengers to board the elevator car 210. Therefore, the estimated arrival time calculation unit 117 calculates the dwell time so that the more passengers there are at the hall, the longer it will be.

In addition, if there are also calls for passengers going up at the third floor boarding area, the estimated arrival time calculation unit 117 calculates the dwell time assuming that half of the passengers are heading down.

The estimated arrival time calculation unit 117 calculates the estimated arrival time to the first floor based on the rated speed of the elevator car 210 and the calculated dwell time at the third floor.

In addition to the above example, consider a case where there are multiple passengers in the elevator car 210 whose destinations are the 4th, 2nd, and 1st floors. In this case, the estimated arrival time calculation unit 117 calculates the dwell time on the 4th and 2nd floors and then calculates the estimated arrival time to the 1st floor.

The estimated arrival time calculation unit 117 calculates the dwell times on the fourth and second floors by apportioning the number of passengers in the car. Specifically, the estimated arrival time calculation unit 117 calculates the dwell times on the fourth and second floors by assuming that one-third of the passengers will get off on the fourth floor, one-third of the passengers will get off on the second floor, and one-third of the passengers will get off on the first floor.

The estimated arrival time calculation unit 117 uses this method to calculate the estimated arrival time for each floor. Note that the method for calculating the estimated arrival time is not limited to this.

Figure 7 is a flowchart showing the estimated arrival time calculation process of the prediction management device 110 shown in Figure 6. The flowchart in Figure 7 shows the process when a request for an estimated arrival time is made from the mobile terminal 300. The flowchart in Figure 7 is also performed each time a request for an estimated arrival time is made from the mobile terminal 300. It is also possible to make a request for an estimated arrival time from the car exterior display device 240. In this case, the car exterior display device 240 has a touch panel, and the request for the estimated arrival time is made via the touch panel of the car exterior display device 240.

In step S301, the receiving unit 111 receives the current number of occupants in the car information 126, the current landing information 127, and the current operation information 128 from the elevator control device 220. The memory unit 112 stores the current number of occupants in the car information 126, the current landing information 127, and the current operation information 128.

In step S302, the estimated arrival time calculation unit 117 calculates the estimated arrival time at each floor based on the current number of occupants in the car information 126, the current platform information 127, and the current operation information 128.

In step S303, the display data creation unit 115 creates HTML data as display data. This HTML data contains the estimated arrival time at each floor. In step S304, the transmission unit 116 transmits the created HTML data to the mobile terminal 300. Note that if the request for the estimated arrival time comes from the car exterior display device 240, the transmission unit 116 transmits the display data to the car exterior display device 240.

Furthermore, when the car interior display device 212 displays the estimated arrival time, it may display the estimated arrival time for all floors. Alternatively, the car interior display device 212 may display only the estimated arrival time for the destination floor specified within the elevator car 210.

Furthermore, when the projection display device 130 and the outside car display device 240 display the estimated arrival time, the estimated arrival time for all floors may be displayed. Alternatively, if the projection display device 130 is installed at the first floor entrance 400, it may only display the estimated arrival time for the first floor.

Furthermore, if a projection display device 130 and an outside car display device 240 are installed on each floor, the projection display device 130 and the outside car display device 240 may only display the estimated arrival time at the floor on which the vehicle is installed, and may not need to display the estimated arrival time at other floors.

Furthermore, when the car interior display device 212, the projection display device 130, and the car exterior display device 240 display the estimated arrival time, the flowchart in Figure 7 is repeatedly and continuously executed. This allows the car interior display device 212, the projection display device 130, and the car exterior display device 240 to display the estimated arrival time in real time.

The elevator operation prediction system 100 in embodiment 2 includes a prediction management device 110 and a display device. The prediction management device 110 calculates the estimated time until the elevator car 210 arrives at the landing based on current operation information 128 and current number of users information 129. The prediction management device 110 also creates display data for the calculated estimated time. The display device displays the estimated time based on the display data created by the prediction management device 110.

As a result, the elevator operation prediction system 100 can provide users with more appropriate information regarding the operation of the elevator 200. Furthermore, the elevator operation prediction system 100 in embodiment 2 can give users an opportunity to consider alternative methods of travel, such as stairs or an escalator. This can prevent the elevator 200 from becoming overcrowded.

The elevator operation prediction system 100 also includes a projection display device 130 installed at the building entrance 400 as a display device. The projection display device 130 has a projection unit 131 that projects an image created as display data by the prediction management device 110 onto the entrance 400. The elevator operation prediction system 100 displays the image created as display data by the prediction management device 110 on the car exterior display device 240.

This allows users to be informed of the expected arrival time of elevator 200.

The projection unit 131 also projects the image created as display data onto the floor of the entrance 400. As a result, the elevator operation prediction system 100 can display the estimated arrival time of the elevator 200 in a location where there are no walls, such as the center of the entrance 400 where residents often pass through. The elevator operation prediction system 100 also displays the estimated arrival time of the elevator 200 on the outside-car display device 240, which is frequently used by users. This allows users to be more aware of the information.

The prediction management device 110 in embodiment 2 may be provided with the receiving unit 111, memory unit 112, estimated arrival time calculation unit 117, display data creation unit 115, and transmission unit 116 shown in Figure 6. Furthermore, the memory unit 112 shown in Figure 6 may store information on the number of passengers currently in the car 126, current platform information 127, and current operation information 128.

Furthermore, the elevator 200 in embodiment 2 may be installed not only in apartment buildings, but also in buildings such as office buildings, commercial facilities, and public facilities.

Furthermore, the elevator operation prediction system 100 in embodiment 2 may be linked to a call registration system that registers calls in advance using the mobile terminal 300. In this case, the elevator operation prediction system 100 can obtain the number of passengers at the landing on each floor, call information for each floor, the number of passengers in the car, and the destination floor in the car from the call registration system.

In addition, in embodiments 1 and 2, the elevator system 1 is equipped with both the projection display device 130 and the outside-car display device 240 to transmit information outside the elevator car 210, but only one of them may be used. In addition, in embodiments 1 and 2, at least one of the outside-car display device 240, the inside-car display device 212, and the mobile terminal 300 may be included in the elevator operation prediction system 100 as a display device.

Furthermore, each function of the predictive management device 110 in embodiments 1 and 2 is realized by a processing circuit. Figure 8 is a configuration diagram showing a first example of a processing circuit that realizes each function of the predictive management device 110 in embodiments 1 and 2. The processing circuit 600 in the first example is dedicated hardware.

The processing circuit 600 may be, for example, a single circuit, a composite circuit, a programmed processor, a parallel programmed processor, an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Gate Array), or a combination of these. Furthermore, each function of the predictive management device 110 may be realized by a separate processing circuit 600. Alternatively, the functions of the predictive management device 110 may be realized collectively by the processing circuit 600.

Figure 9 is a diagram showing a second example of a processing circuit that realizes each function of the predictive management device 110 in embodiments 1 and 2. The processing circuit 700 of the second example includes a processor 710 and a memory 720.

In the processing circuit 700, each function of the predictive management device 110 is realized by software, firmware, or a combination of software and firmware. The software and firmware are written as programs. The software and firmware are then stored in memory 720. The processor 710 realizes the functions of each part by reading and executing the programs stored in memory 720.

The programs stored in memory 720 can be said to cause the computer to execute the procedures or methods of each of the above-mentioned sections. Here, memory 720 refers to non-volatile or volatile semiconductor memory such as RAM (Random Access Memory), ROM (Read Only Memory), flash memory, EPROM (Erasable Programmable Read Only Memory), and EEPROM (Electrically Erasable and Programmable Read Only Memory). Magnetic disks, flexible disks, optical disks, compact disks, minidisks, DVDs, etc. also qualify as memory 720.

Some of the functions of each of the above-mentioned parts may be realized by dedicated hardware, and some may be realized by software or firmware.

In this way, the processing circuitry can realize the functions of each of the above-mentioned parts through hardware, software, firmware, or a combination of these.

The processing circuits 600 and 700 can be applied not only to the predictive management device 110, but also to the management terminal 120 and the mobile terminal 300.

1 Elevator system, 100 Elevator operation prediction system, 110 Prediction management device, 120 Management terminal, 130 Projection display device (display device), 131 Projection unit, 200 Elevator, 210 Elevator car, 212 In-car display device (display device), 240 Out-car display device (display device), 300 Mobile terminal (display device), 400 Entrance (shared space).

Claims (8)

a prediction management device that makes predictions regarding elevator operations in an apartment building; and a management terminal that acquires event schedule information that is information about events scheduled in the apartment building and includes at least the names of the events ;
The prediction management device
storing operation history information, which is a history of the operation of the elevator; user number history information, which is a history of the number of users of the elevator by time; and event history information, which is a history of events that have been held in the apartment building in the past;
performing machine learning based on the operation history information, the number of users history information, and the event history information, and predicting the congestion level of the elevator based on a learning result of the machine learning and the event schedule information acquired by the management terminal;
Create data to display the predicted congestion level,
When performing the machine learning,
Based on the event history information, the data in the operation history information and the data in the number of users history information are grouped into days on which an event occurred and days on which no event occurred;
For days when an event occurred, the data is further divided into subgroups subdivided by event name, and the machine learning is performed for each of the subdivided subgroups. As a learning result of the machine learning, an event implementation date learning result is created for each subgroup.
The machine learning is performed on a group of days on which no events occurred, and a normal day learning result is created as the learning result of the machine learning.
When predicting the congestion degree of the elevator,
Based on the event schedule information, classifying the prediction target date as either a scheduled event date or a normal day on which no event is scheduled to be held;
If the target prediction date is the scheduled event date, a congestion degree is predicted using the event implementation date learning result corresponding to the name of the event included in the event schedule information .
When the prediction target day is the normal day, the elevator operation prediction system predicts the congestion level using the normal day learning result. a projection display device having a projection unit,
The elevator operation prediction system according to claim 1 , wherein the projection unit projects the image created as the display data by the prediction management device onto a common space of the apartment building. The elevator operation prediction system of claim 2, wherein the projection unit projects the image created as the display data onto the floor of the common space. The prediction management device
It stores weather history information, which is a history of past weather conditions,
Obtaining weather forecast information from the management terminal or an external system;
4. The elevator operation prediction system according to claim 1, wherein the congestion level of the elevator is predicted based on the operation history information, the number of users history information, the event history information, the event schedule information, the weather history information, and the weather forecast information. a display device;
the prediction management device calculates an estimated time until the elevator car arrives at the landing based on current operation information, which is information on the current operation status of the elevator, and current user number information, which is information on the current number of users of the elevator, and creates estimated time display data as data for displaying the calculated estimated time;
2. The elevator operation prediction system according to claim 1, wherein the display device displays the estimated time based on the estimated time display data created by the prediction management device. the display device is a projection display device installed in a common space within a building,
the projection display device has a projection unit that projects an image created by the prediction management device as the estimated time display data onto the common space,
The elevator operation prediction system according to claim 5 , wherein the projection display device displays the estimated time by projecting the image using the projection unit. The elevator operation prediction system of claim 6, wherein the projection unit projects an image created as the estimated time display data onto the floor of the common space. a prediction management device that makes predictions regarding elevator operations in an apartment building;
a management terminal that acquires event schedule information , which is information about events scheduled in the apartment complex and includes at least the names of the events; and a mobile terminal carried by a resident of the apartment complex.
Equipped with
The prediction management device
storing operation history information, which is a history of the operation of the elevator; user number history information, which is a history of the number of users of the elevator by time; and event history information, which is a history of events that have been held in the apartment building in the past;
performing machine learning based on the operation history information, the number of users history information, and the event history information, and predicting the congestion level of the elevator based on a learning result of the machine learning and the event schedule information acquired by the management terminal;
A configuration is provided in which display data of the predicted congestion level is generated,
The mobile terminal designates a time to be predicted on a day to be predicted ,
the prediction management device receives the prediction target date and the specified time from the mobile terminal, predicts the congestion level at the specified time, creates reply data that is the display data regarding the predicted congestion level , and returns the reply data to the mobile terminal;
The mobile terminal displays the reply data,
The prediction management device
When performing the machine learning,
Based on the event history information, the data in the operation history information and the data in the number of users history information are grouped into days on which an event occurred and days on which no event occurred;
For days when an event occurred, the data is further divided into subgroups subdivided by event name, and the machine learning is performed for each of the subdivided subgroups. As a learning result of the machine learning, an event implementation date learning result is created for each subgroup.
The machine learning is performed on a group of days on which no events occurred, and a normal day learning result is created as the learning result of the machine learning.
When predicting the congestion degree of the elevator,
classifying the prediction target date as a scheduled event date or a normal day on which no event is scheduled to be held based on the event schedule information;
If the target prediction date is the scheduled event date, a congestion degree is predicted using the event implementation date learning result corresponding to the name of the event included in the event schedule information.
If the prediction target day is the normal day, the normal day learning result is used to predict the congestion degree.
Elevator operation prediction system.