JP6866275B2 - External system cooperation Vehicle dispatch system and method - Google Patents

Description

The present invention relates to an operation management of an elevator device, and particularly relates to an external system-linked vehicle allocation system and method for appropriately performing car allocation when an elevator platform is congested in cooperation with an external system.

Many proposals have been made for the operation management of elevator devices. Among these, there is a proposal for operation management when elevator device users are crowded at the landing on each floor.

For example, Patent Document 1 aims to obtain an elevator control device capable of allocating an appropriate number of cars to a congested floor for an elevator system in which a plurality of elevators are grouped together. A landing camera 1B and a car in-car camera 1A that detect the congestion state inside, and a landing area detecting means 3A and a car inner area that measure the passenger area and the occupied area in the current elevator landing based on the output of these cameras. Based on the detection means 3B, the additional allocation determination means 3C that predicts the future congestion degree of both the elevator landing and the car based on the output of these detection means and the past learning result, the traffic information learning means 3G, and the prediction result. It is equipped with an allocation calculation means 3D and an operation control means 3E for allocating a plurality of cars to the landing. "

JP-A-2002-302348

According to the device described in Patent Document 1, the congestion state of the elevator landing is detected, the future congestion degree of the elevator landing is predicted based on the detection output and the past learning result, and a plurality of cars are dispatched to the landing. Therefore, it is possible to properly dispatch the car when the elevator platform is crowded.

According to the method of Patent Document 1, since the future congestion degree of the elevator landing is predicted based on the past learning results, it is possible to deal with the congestion events that regularly occur in the building where the elevator is installed. Is. For example, going to and from work in the morning and evening, congestion at lunch, etc. are events that occur routinely except on weekends, so it is effective to properly allocate cars when the elevator platform is congested. ..

However, it is insufficient as a response to sudden events other than routine. The past learning results in the method of Patent Document 1 do not include sudden events other than routines, and it is impossible to predict sudden events.

Further, in Patent Document 1, users gather at the landing and respond after a call is made, so there is a high possibility that congestion has occurred, and even if it is effective for allocating an appropriate number of vehicles. , It is insufficient as a measure to prevent congestion.

From this, it is an object of the present invention to provide an external system-linked vehicle allocation system and method capable of proactively dispatching a vehicle, including a response in the event of a sudden event.

From the above, in the present invention, "an external system linked vehicle allocation system that manages the operation of a plurality of elevator devices installed in a facility such as a building, and the operation results of the plurality of elevator devices and event information from the external system are obtained. The number of users is predicted for each floor of the landing using the input receiving unit and the learning unit that stores the information of the driving performance obtained from the receiving unit as past experience data and learns, and the stored information and event information of the learning unit. An external system-linked vehicle allocation system characterized in that it is equipped with a prediction unit for predicting congestion and a landing call registration determination unit for pre-allocating an elevator car to a predicted congestion floor where congestion is predicted based on the time information of event information. ".

Further, in the present invention, "an elevator operation management method for managing the operation of a plurality of elevator devices installed in a facility such as a building, and using the operation results of the plurality of elevator devices and event information from an external system for landing. The number of users is predicted for each floor, and based on the time information of the event information, the elevator car is predicted to occur on the predicted congestion floor. It is a thing.

According to the present invention, since it is possible to pre-allocate an appropriate number of vehicles, it is possible to deal with sudden congestion as well as routine congestion.

The figure which shows the schematic structure of the elevator operation management system provided with the car dispatch processing function at the time of the predicted congestion which concerns on this invention. The figure which showed the landing environment example suitable for this invention. The figure which shows the example of the storage format of the past experience data learned by the learning part 31. The figure which shows an example of the predicted number of passengers table TB1 predicted by the number of people prediction unit 32 by floor. The figure which shows an example of the predicted number of people getting off table TB2 predicted by the number of people prediction unit 32 by floor. The flow chart which concretely illustrated the processing content in the person number prediction unit 32 by floor of FIG. The figure which shows an example of the predicted passenger number table TB1 including the actual data added in the accuracy verification. The figure which shows an example of the predicted number of people table TB2 including the actual data added in the accuracy verification. The figure which shows the processing flow of the car dispatch processing function at the time of the predicted congestion which concerns on Example 1 of this invention. The figure which shows the capacity per elevator car. FIG. 7 is a diagram showing a state of the congested floor (the state of the congested floor) when the preceding car dispatch process is performed at the time of predicted congestion in FIG. The figure which shows the maximum number of passengers (capacity * passenger rate) by unit, by floor, and maximum boarding capacity. It is a figure which shows the processing flow of the car dispatch processing function at the time of the predicted congestion which concerns on Example 2 of this invention. The figure which shows the processing flow of the car dispatch processing function at the time of the predicted congestion which concerns on Example 3 of this invention. FIG. 12 is a diagram showing a state of a congested floor (predicted congestion floor: for example, the first floor) when the preceding car dispatch process at the time of predicted congestion is performed.

Hereinafter, examples of the present invention will be described with reference to the drawings.

FIG. 1 shows a schematic configuration of an elevator operation management system provided with a preceding car dispatch processing function at the time of predicted congestion according to the present invention. When the function of the present invention is expressed differently from Patent Document 1, it is appropriate to use the preceding car dispatch processing function at the time of predicted congestion. The function of Patent Document 1 should be called a car dispatch processing function at the time of congestion.

FIG. 1 describes facilities and systems in a facility such as a building 1 and an external system 2. Of these, the equipment and systems in the facility such as the building 1 are the elevator operation management system 3, the landing elevator service request device 4 on each floor, the surveillance camera 5 on each floor, the building management system 6, and the like, and communication means among them. Data communication is carried out with each other via 8. In addition, a plurality of elevator control systems 7a ... 7n are installed and are controlled by the elevator operation management system 3.

Further, FIG. 1 illustrates a public institution management system as an example of the external system 2. Here, the elevator operation management system 3 is a system linked to the building management system 6 and the public institution management system 2 as an external system other than the elevator operation management system 3, and constitutes an external system-linked vehicle allocation system.

The elevator operation management system 3 according to the present invention obtains many inputs and settings and gives an output. Of these inputs and outputs, between the elevator control system 7a ... 7n and the elevator operation management system 3, the elevator control system 7a ... 7n reports the operation status information S71 to the elevator operation management system 3, and the elevator control system In 7a ... 7n, the elevators of each unit are controlled by the control command signal S72 from the elevator operation management system 3. What is characteristic here is that the elevator operation management system 3 operates and manages all the units in the building 1, and other matters are the same as normal elevator control, so the explanation here Is omitted.

In the present invention, as still other inputs, the service request signal S4 from the landing elevator service request device 4 on each floor, the video signal S5 from the surveillance camera 5 on each floor, the building management information S6 from the building management system 6, and the public institution management system 2 Obtain public institution management information S2 and the like.

FIG. 2 is a diagram showing an example of a landing environment suitable for the present invention. At the elevator platform on each floor, there are surveillance cameras 5 (5-1, 5-2, 5-3, 5-4) that monitor and photograph the space including the elevator door, and vertical buttons 4 as the platform elevator service request device 4. (4-1, 4-2, 4-3, 4-4) are installed. In addition, lanterns 20 (20-1, 20-2, 20-3) that predict the arrival of the elevator car or guide the direction are arranged. In FIG. 2, an in-car camera 21 and a load sensor 22 are further provided in the car 24 of the elevator.

As will be described below, the position of the service request signal S4 in the present invention is for confirming the service direction of going up and down from the landing to the elevator. Although the vertical button 4 is illustrated in FIG. 2, this may be a destination floor registration device for registering the destination floor from the landing.

Further, the positioning of the video signal S5 in the present invention is for measuring the number of users, and can be replaced by other means as long as the number of users can be confirmed directly or indirectly. In the example of FIG. 2, information on the number of users can be obtained from the in-car camera 21 provided in the elevator car 24 and the load sensor 22 provided in the lower part of the elevator car 24.

In this way, the up and down directions of the elevator can be confirmed from the service request signal S4, the number of users can be confirmed from the video signal S5, and the meeting in the facility can be confirmed from the building management information S6. , Events and other action schedules can be confirmed, and public institution operation information (for example, train delay) on the day can be grasped from public institution management information S2. It should be noted that some of this information includes information that is input in each existing system and used for some purpose, but the new point in the present invention is that it is used for estimating the predicted number of passengers getting on and off.

The receiving unit 36 in the elevator operation management system 3 of FIG. 1 obtains a service request signal S4, a video signal S5, a building management information S6, etc. via the communication means 8, and is in an operating state from the elevator control systems 7a ... 7n. Information S71 is input.

The input signal from the receiving unit 36 is given to and used by the landing call registration determination unit 39 and the learning unit 31, but since these are parts related to the essence of the present invention, this description will be described later and is general. The comprehensive evaluation unit 37 and the allocation command unit 38, which are the functions of the system, will be explained first.

The comprehensive evaluation unit 37 determines the user's request and the moving direction from these input signals, and the allocation command unit 38 gives a control command signal S72 to the elevator control systems 7a ... 7n of each unit. To control. This part is no different from the conventional elevator control, so further explanation is omitted.

The service request signal S4, the video signal S5, the building management information S6, and the like obtained via the communication means 8 are recorded and used by the learning unit 31. Here, the service request signal S4 and the video signal S5 are stored together with the information of the time when these signals are generated, and are used as past experience information. As a result, the behavior and mode of the user at a certain scene (day of the week, season, etc.) in the past and a certain time can be statistically grasped. For example, it is possible to grasp an outline of the movement status of people at work, lunch, night, and the like. Therefore, in similar future situations, it can be presumed that the user exhibits a behavior pattern similar to that of past experience.

Incidentally, in Patent Document 1, the information used for the "past learning result" is such information. In order to realize the car dispatch processing function at the time of congestion of Patent Document 1, the service request signal S4 and the video signal S5 may be sufficient. These pieces of information can be said to be information about the operation results of a plurality of elevator devices, and are information obtained within the range controlled and managed by the elevator operation management system.

Further, the building management information S6 from the building management system 6 and the public institution management information S2 are used in order to realize the preceding car dispatch processing function at the time of predicted congestion as in the present invention.

While the service request signal S4 and the video signal S5 are used as past experience information and operation record information, the building management information S6 from the building management system 6 is used for meetings in the facility in the near future. The action schedule (venue, attendees and their attendance) is the information registered in the building management system 6, and according to this, for example, from each floor at the time of the meeting on the 5th floor from 3 o'clock today. The movement of people is predictable.

Further, if the train delay and its degree can be grasped from the public institution management information S2 as the operation information of the public institution on the day, for example, the movement trend of the user at the time of commuting is different from the normal movement trend without delay. , It is predictable that it will change.

In the present invention, the building management information S6 and the public institution management information S2 will be referred to as event information from the external system with respect to the operation record information. There are planned event information (building management information S6) and sudden event information (public institution management information S2), both of which are provided by an external system other than the elevator operation management system.

The building management information S6 and the public institution management information S2 are obtained together with the information on the occurrence time or the end time of these events.

In this way, the learning unit 31 learns the number of people who use the normal elevator on a daily basis. Here, macro information on the number of people can be output. Further, the learning unit 31 is learning the occupancy rate for each floor.

As a method of learning the occupancy rate, the number of passengers in the elevator car is detected or calculated by the load in the elevator or the camera in the car. Further, in order to detect the number of people at the landing, the number of people at the landing is directly detected by using a landing camera, a distance sensor, or the like. Alternatively, a method of recognizing that there is a person at the landing when the landing button is pressed may be used from the button registration status of the landing.

In response to the landing button registration from each floor, when arriving and the door opens, it is judged whether or not the number of people at the landing will disappear, and if there are people at the landing, it is judged that people can board from the current number of passengers. Record the occupancy rate. Specifically, if there are 10 people on the 5th floor and 10 people are in the car, and the car arrives on the 5th floor and 2 people remain in the car, the number of people in the car will be There will be 18 people. If the number of passengers is 24, the occupancy rate in the car will be 75%. Alternatively, a method of detecting the occupancy rate of the users in the car or the vacancy rate in the car at that time instead of the number of people and recording the occupancy rate or the vacancy rate may be used.

If the platform camera is not attached, a call for the 5th floor will be created, and when the car arrives on the 5th floor,
After the door is closed, if there is a call for boarding in the same direction or a service request to the same destination floor within a certain arbitrary time, it is judged that the user could not board the boarding and re-registered, and that Record the number of passengers, occupancy rate, and vacancy rate on the relevant floor.

The recorded number of passengers, occupancy rate, and vacancy rate are used as the occupancy rate, and these are learned daily by floor.

Here, the occupancy rate for each floor is obtained by calculating the usage ratio of each floor, and may be obtained by weighting according to the number of users on each floor. Furthermore, the occupancy rate by floor should be grasped by going up and down. For example, in the case of a building on the 8th floor, it is preferable to obtain the occupancy rate in the up direction and the occupancy rate in the down direction as the occupancy rate on the 5th floor.

The administrator of the elevator operation management system sets the occupancy rate by using the occupancy rate setting unit 40 with reference to the past occupancy rate results accumulated in the learning unit 31. The set occupancy rate defines the degree of congestion at the limit for the user to make a decision to wait for the next vehicle allocation. For example, even in a 24-seater car, if 18 people are on board, the occupancy rate is the limit at which people waiting for the next dispatch will appear. Even if the occupancy rate is high when commuting, it may be low during normal times, so it is better to set the limit occupancy rate by floor, up / down, time zone, and so on.

The past experience data learned by the learning unit 31 is organized and stored as shown in FIG. 3, for example. FIG. 3 exemplifies, for example, a storage format of the number of passengers in the past, and stores the number of passengers on each floor and the information on the boarding rate for each up and down floor in association with each other for each past date and time. There is. A storage format for the number of people who have dropped in the past is also created in the same format. It is preferable that the storage format includes information that is stored by grasping the number of people on a daily basis, for example, every 10 minutes, and that has been accumulated over a long period of time in the past. In addition, the past experience data should include event information such as meetings and various events as incidental information. The past experience data learned by the learning unit 31 is used as the past experience in the prediction processing in the following processing.

The number of people prediction unit 32 by floor predicts the number of people by floor, for example, based on past experience and the schedule of the meeting on the day. FIG. 4a shows an example of the predicted number of passengers table TB1 predicted by the floor-based number prediction unit 32, and FIG. 4b shows an example of the predicted number of passengers table TB2 predicted by the floor-based number prediction unit 32.

The predicted number of passengers table TB1 and the predicted number of people getting off the table TB2 are sequentially time data D1, D6, floor data D2, D7, predicted number of people data D3, D8, actual number of people data D4, D9, prediction accuracy data D5, D10, etc. It is composed of occupancy rates D11 and D12. In the floor-specific number prediction unit 32, the data from the top to the third row of these tables are formed by using the past experience data of FIG. 3 and the like.

For example, with respect to the predicted number of passengers table TB1, at the time of 8 o'clock (which will be described later, for example, it represents 10 minutes from 8 o'clock), the number of passengers on each floor (here, from the 1st to the 8th floor) is 20, 9 respectively. , 7, 14, 13, 7, 8, 5 people are predicted. Also, for example, regarding the predicted number of people getting off table TB2, at the time of 8 o'clock (which will be described later, for example, it represents 10 minutes from 8 o'clock), the number of people getting off each floor (here, from the 1st to the 8th floor) is 20, respectively. It shows that it was predicted that there were 5, 9, 15, 11, 15, 18, and 11.

Regarding the method of creating the predicted number of passengers table TB1 and the predicted number of passengers getting off table TB2, it can be created in consideration of past experience and today's schedule as described above, and further in consideration of today's train delay and the like. It can be obtained by correction.

FIG. 5 is a flow chart that specifically illustrates the processing content of the floor-specific number prediction unit 32 of FIG. As this premise, the number of users in time series measured daily is grasped together with the information on the occupancy rate for each floor and up / down direction by the processing in the learning unit 31, and the past experience data of FIG. 3 is formed. It is assumed that it has been done. That is, it is assumed that the past experience data corresponding to the actual number data D4 and D9 are secured and stored for the predicted number of passengers table TB1 and the predicted number of passengers table TB2 for a considerable number of days in chronological order. In addition, it is assumed that the past days are memorized including the information of the events and meetings held on that day.

The process of FIG. 5 may be started at an appropriate timing, but if it is provided as information for one day on the next day on the previous day, it will be processed at an appropriate time on the previous day. Alternatively, if it is provided by an external request, it may be started at the time of the request. In addition, when a new state change occurs, it is advisable to review it each time even on the day of the event.

In the first processing step S100 of the number of people prediction unit 32 for each floor, past experience data and the like are taken in. This includes past actual number data D4 and D9, time data D1 and D6, occupancy rate, building management information S6, and the like. In the process step S101, setting information such as a time width and a designated time is taken in.

In the process step S102, the output date (for example, tomorrow) is determined. It is determined whether the output date is a weekday, a holiday, or a partial suspension, and only those with the corresponding conditions are extracted from the past experience data of FIG. In the process step S103, for example, if the output date is a weekday, only the past experience data of the weekday is extracted, and if the output date is a holiday, only the past experience data of the holiday is extracted. If seasonal fluctuations and day-of-week fluctuations are noticeable for users, it is advisable to take these points into consideration when extracting.

In the processing step S104, the usage average for each time is obtained for the extracted time-series usage results for a plurality of days, and the predicted number of passenger data D3 and D8 of the predicted number of passengers table TB1 and the predicted number of passengers getting off the table TB2 are used. Since the above processing is performed for the number of users for each floor, floor data D2 and D7 are also obtained.

In the processing step S105, the presence or absence of the building management information S6 is confirmed. For example, if a meeting is scheduled to be held on the 5th floor from 15:00 today, in the processing step S106, the movement of the user according to the holding scale. The predicted number of passenger data D3 and D8 of the predicted number of passengers table TB1 and the predicted number of passengers getting off the table TB2 obtained in the processing step S104 are modified to reflect how to use the elevator. If the past experience data includes the experience of having a meeting with the same purpose as this meeting in the past, the predicted number of passengers table TB1 and the predicted number of passengers will be disembarked with reference to the user information at that time. It is preferable to correct the predicted number of people data D3 and D8 in the table TB2.

In the process step S107, the presence or absence of the public institution management information S2 is confirmed. For example, if the information that the train arriving at the nearest station of the building is delayed at 8 o'clock today is obtained, the process step S108 is performed. The predicted number of passenger data D3 and D8 of the predicted number of passengers table TB1 and the predicted number of passengers getting off the train TB2 obtained in the processing steps S104 and S106 are obtained by reflecting the movement of the user and the way of using the elevator according to the degree of the delay. Fix it. The processing of the processing steps S107 and S108 will be executed based on the external information obtained on the day.

As described above, the past results are corrected based on the action schedule and the information of the public institution, and the predicted number of people data D3 and D8 of the predicted number of passengers table TB1 and the predicted number of people getting off the table TB2 are obtained. Although not specified in the flow of FIG. 5, the occupancy rate data set by the occupancy rate setting unit 40 using the past experience data of FIG. 3 is the predicted number of passengers table TB1 and the predicted number of passengers getting off the table TB2. It is reflected in the columns of the occupancy rates D11 and D12.

The floor-specific number prediction unit 32 is further equipped with an accuracy verification function, and the information of the lower two stages is further added to the data of the upper three stages created by the floor-specific number prediction unit 32 based on the actual experience of the predicted day. And add data. FIG. 6a shows an example of the predicted number of passengers table TB1 including the actual data added by the accuracy verification, and FIG. 6b shows an example of the predicted number of passengers table TB2 including the actual data added by the accuracy verification.

In this case, for example, with respect to the predicted number of passengers table TB1, at 8 o'clock, the number of passengers on each floor was predicted to be 20, 9, 7, 14, 13, 7, 8, and 5, respectively. There were 18, 13, 10, 19, 14, 14, 10, and 9, and it can be seen that the accuracy of each was 82, 38, 60, 75, 92, 88, 95, 90%.

In this case, for example, with respect to the predicted number of people getting off table TB2, at 8 o'clock, the number of people getting off each floor was predicted to be 20, 5, 9, 15, 11, 15, 18, and 11, respectively, but in reality It can be seen that there were 17, 13, 15, 12, 12, 17, 19, and 10 persons, and the accuracy of each was 89, 69, 70, 74, 93, 50, 80, and 56%.

The addition of the data D4 and D9 in the tables of FIGS. 6a and 6b means that the data D4 and D9 are added to the learning unit 31 as new past experience data.

The landing call registration determination unit 39 of FIG. 1 normally performs a landing call registration determination process based on the service request signal S4 from the landing elevator service request device 4 on each floor, and causes the comprehensive evaluation unit 37 and the allocation command unit 38. The car is dispatched to the registered floor through the vehicle, but in the present invention having the preceding car dispatch processing function at the time of predicted congestion, the car further functions as follows.

FIG. 7 shows a processing flow of the preceding car dispatch processing function at the time of predicted congestion according to the first embodiment of the present invention. In the first embodiment, the elevator car is made to stand by for opening the door on the predicted congestion floor based on the predicted congestion information.

In this case, the predicted congestion information is the building management information S6 from the building management system 6 and the public institution management information S2 from the public institution management system 2. According to this information, it can be used as predictive congestion information because it includes the concentration and time of users based on the action schedule of meetings, events, etc. in the facility in the near future, or the train delay.

In the processing flow of the vehicle dispatching processing function for predicting congestion in FIG. 7, first, in the processing step S110, "is there delay information / event information of a public institution?" Is confirmed. The delay information of the public institution is determined by the public institution management information S2 from the public institution management system 2, and the event information is determined by the building management information S6 from the building management system 6. If there is no delay information / event information in particular (S110, NO), the process proceeds to process step S115, and the pre-vehicle allocation process due to the door opening standby is canceled.

When congestion is predicted from the public institution management information S2 or the building management information S6 (S110, YES), in the processing step S111, it is confirmed that "does the time of the information match the current time?". Since the public institution management information S2 and the building management information S6 include delay time and time information regarding the holding and ending of the conference, the relationship between the event occurrence time and the current time is determined. For example, when the current time reaches one minute before the event occurrence time, it is determined that the information and the current time match (S111, YES), and the process of the process step S112 is started. When the information does not match the current time (S111, NO), the process proceeds to the processing step S115, and the pre-vehicle allocation process due to the door opening standby is canceled.

In the processing step S112, the predicted number of users and the number of occupants per elevator car are compared. As for the predicted number of users, as shown in FIGS. 4a and 4b, the number of users by floor is obtained by time, floor, and getting on and off, so refer to this. In addition to the number of users based on past experience data, the number of users is based on public institution management information S2 and building management information S6, and is considered to include the number of people who will increase transiently due to events in the near future.

Here, the number of occupants is the number of passengers or the number of passengers multiplied by the above-mentioned occupancy rate. Here, the number of passengers will be described first as the number of passengers. As shown in FIG. 8, the number of passengers (capacity) per elevator car is known as a specification for each elevator machine, and is, for example, 24 people / car and 1560 kg / car. In the processing step S112, the predicted number of users and the number of occupants per elevator car (capacity) are compared, and if the predicted number of users> the number of occupants per elevator car, the number of passengers is determined. In the processing step S114, a plurality of elevators are dispatched in advance, and these are set to wait for the opening of a plurality of elevators. When the predicted number of users ≤ the number of occupants per elevator car, one elevator is pre-distributed in the processing step S113, and a plurality of elevators are placed on standby for opening.

FIG. 9 is a diagram showing a state of a congested floor (congested expected floor: for example, the first floor) when the preceding car dispatch process at the time of predicted congestion of FIG. 7 is performed. For example, one of the three elevator cars (24a) or a plurality of elevator cars (24a, 24c) are pre-distributed, and when the elevator car 24 arrives at a congested floor, each door 25 (25a, 25b, 25c) is opened and is in an open state. Wait as it is (wait for door opening). At this time, if it is clear that most of the users move in the same ascending and descending directions, the landing elevator service requesting device 4 on the predicted congested floor does not display the destination direction for both ascending and descending. It may be a thing. Users who appear at the landing at the scheduled congestion time will board the elevators waiting for the door to open.

In the processing step S102 of FIG. 7, when comparing the predicted number of users with the number of occupants per elevator car, the number of occupants (capacity) * occupancy rate is set as the number of occupants. Here, as for the occupancy rate, the occupancy rate setting unit 40 of the floor-specific number prediction unit 32 sets the number of people by time, floor, and boarding / alighting as shown in FIGS. 4a and 4b. To do. Here, the reason why the number of passengers is taken into consideration in consideration of the occupancy rate is based on this user psychology because the user's psychology tends not to board (wait for the next) when the car is congested above a certain level. Therefore, as a more practical operation, a large number of cars are dispatched.

FIG. 10 shows the maximum number of passengers (capacity * occupancy rate) obtained for each unit, floor, and floor. In the processing step S102 of FIG. 7, the predicted number of users and the number of passengers in FIG. 10 are compared. Then, the number of elevators to be dispatched is determined.

The above processing enables smooth movement of users in a predicted congestion state where a large number of users are expected to gather on the same floor due to train delays or the holding of a large conference. In particular, it can be said that the effect of the present invention is great when the entrance floor of a building directly connected to a train or subway station is provided.

FIG. 11 shows a processing flow of the preceding car dispatch processing function at the time of predicted congestion according to the second embodiment. In the second embodiment, the call button in the down direction is automatically registered by going up to the landing elevator service requesting device 4 on the predicted congestion floor based on the predicted congestion information.

In this case, the predicted congestion information is the building management information S6 from the building management system 6 and the public institution management information S2 from the public institution management system 2. Based on this information, it is possible to clarify the concentration and time of users based on action schedules such as meetings and events in the facility in the near future, or train delays, and it can be used as predicted congestion information.

In the processing flow of the car dispatch function at the time of congestion in FIG. 11, first, in the processing step S110, "is there delay information / event information of a public institution?" Is confirmed. The delay information of the public institution is determined by the public institution management information S2 from the public institution management system 2, and the event information is determined by the building management information S6 from the building management system 6. If there is no delay information / event information in particular (S110, NO), the process proceeds to process step S204.

When congestion is predicted from the public institution management information S2 or the building management information S6 (S110, YES), in the processing step S111, it is confirmed that "does the time of the information match the current time?". Since the public institution management information S2 and the building management information S6 include delay time and time information regarding the holding and ending of the conference, the relationship between the event occurrence time and the current time is determined. For example, when the current time reaches one minute before the event occurrence time, it is determined that the information and the current time match (S111, YES), and the process of the process step S200 is started. When the information does not match the current time (S111, NO), the process proceeds to step S105: 204.

In the processing step S200, it is determined "whether the floor is expected to flow in from a public institution or the general enrolled floor?". Since the floor where the inflow is expected from public institutions is generally located at the bottom of the building and most of the users from the outside are looking to go up, the call button in the up direction is automatically registered in the process step S201, and the inflow is expected. In the case of a general enrollment floor, in the process step S202, the downbound call button is automatically registered as the direction in which there are many destination floors on the floor. As a result, the landing elevator service requesting device 4 on these floors is displayed in the up or down direction. By the automatic registration of the call button, the car is dispatched via the comprehensive evaluation unit 37 and the allocation command unit 38 in FIG.

In the processing step S200, the inflow from a public institution is assumed, but it is possible to perform automatic call registration corresponding to an event such as a meeting in the same manner. For example, at the end of a meeting held on the middle floor, it is possible to respond by automatically calling and registering up and down.

After the call button is automatically registered, the timer is started in the process step S203. This timer, for example, sets the time until the congestion is resolved as appropriate. During this period, the automatic registration state of the call button is basically continued, and the call display in the up or down direction is continuously displayed. To do.

In the process step S204, it is confirmed "Is a button in another direction registered by the time the timer time expires?". If NO, you do not have to do anything. When a button in a different direction is registered, various measures can be taken after this. In the process step S205, an example of performing a process of canceling the automatic registration of the call button is shown on the premise that the congestion has been eliminated. Here, confirmation that the congestion has been eliminated may be determined by using the image of the surveillance camera. Alternatively, as the process of the process step S205, a method of continuously displaying the call display until the set timer elapses and lighting the button for the button in the other direction is also conceivable.

It is also possible to set the time until the congestion by the timer is eliminated in two stages from the estimated congestion start time. In the first stage, which is the closest to the estimated congestion start time, the vehicle will be redistributed, but in the second stage after that, the vehicle allocation will be adjusted while observing the congestion, or the car allocation process will be interrupted during congestion. It is better to do it.

According to the second embodiment, since the call button is automatically registered as the preceding car dispatch process at the time of predicted congestion, the user can confirm the vehicle dispatch status.

FIG. 12 shows a processing flow of the preceding car dispatch processing at the time of predicted congestion according to the third embodiment. In the third embodiment, the elevator car is pre-distributed to the predicted congestion floor based on the predicted congestion information.

In this case, the predicted congestion information is the building management information S6 from the building management system 6 and the public institution management information S2 from the public institution management system 2. Based on this information, it is possible to clarify the concentration and time of users based on action schedules such as meetings and events in the facility in the near future, or train delays, and it can be used as predicted congestion information.

In the processing flow of the vehicle dispatching processing function for predicting congestion in FIG. 12, first, in the processing step S110, "is there delay information / event information of a public institution?" Is confirmed. The delay information of the public institution is determined by the public institution management information S2 from the public institution management system 2, and the event information is determined by the building management information S6 from the building management system 6. If there is no delay information / event information in particular (S110, NO), the process proceeds to process step S305, and advance vehicle allocation is not performed.

When congestion is predicted from the public institution management information S2 or the building management information S6 (S110, YES), in the processing step S111, it is confirmed that "does the time of the information match the current time?". Since the public institution management information S2 and the building management information S6 include delay time and time information regarding the holding and ending of the conference, the relationship between the event occurrence time and the current time is determined. For example, when the current time reaches one minute before the event occurrence time, it is determined that the information and the current time match (S111, YES), and the process of the process step S200 is started. When the information does not match the current time (S111, NO), the process proceeds to the processing step S305, and the vehicle is not dispatched in advance.

Since the processing of the processing step S112 is basically the same as the processing of the processing step S112 of FIG. 7, the explanation is omitted, but in short, the predicted number of users and the number of occupants per elevator car are compared. When the predicted number of users> the number of occupants per elevator car, a plurality of elevators are pre-distributed in the processing step S304. When the predicted number of users ≤ the number of passengers per elevator car, one elevator is pre-distributed in the processing step S303. It is better to consider the occupancy rate in this judgment.

FIG. 13 is a diagram showing a state of a congested floor (congested expected floor: for example, the first floor) when the preceding car dispatch process at the time of predicted congestion of FIG. 12 is performed. For example, one of three elevator cars (24a) or a plurality of elevator cars (24a, 24c) are pre-distributed, and when arriving at a congested floor, each door 25 (25a, 25b, 25c) remains closed. stand by. At this time, if it is clear that most of the users move in the same direction, the landing elevator service requesting device 4 on the congested floor may not display the destination direction both up and down. Good.

Users who appear at the boarding area at the scheduled congestion time can promptly board in sequence because the waiting elevator doors are opened by pressing the call button of the boarding area elevator service request device 4.

1: Facilities such as buildings 2: External system (public institution management system) 3: Elevator operation management system 4: Elevator service request device on each floor 5: Surveillance camera on each floor, 6: Building management system, 7a・ ・ ・ 7n Elevator control system, 8: Communication means, 31: Learning unit, 32: Number of people prediction unit by floor, 36: Reception unit, 37: Comprehensive evaluation unit, 38: Allocation command unit, 39: Landing call registration judgment Department, 40: Boarding rate setting unit, S2: Public institution management information, S4: Service request signal, S5: Video signal, S6: Building management information, S72: Control command signal

Claims (4)

It is an external system linked vehicle dispatch system that manages the operation of multiple elevator devices installed in facilities such as buildings.
A receiver that inputs the operation results of a plurality of elevator devices and event information from an external system, a learning unit that stores and learns information on the operation results obtained from the receiver as past experience data, and a memory of the learning unit. In the prediction unit that predicts the number of users for each floor of the landing using the information and the event information, and in the prediction congestion floor where congestion is predicted when the time information including the delay of the event information and the current time match. Equipped with a landing call registration judgment unit that dispatches elevator cars in advance
The landing call registration determination unit uses the occupancy rate of the user in the elevator device to determine the number of elevator cars to be pre-distributed to the predicted congested floor where congestion is expected to occur, and the inflow based on the event information is predicted. Call button automatic registration installed at the landing on the predicted congestion floor is performed in the direction from the predicted congestion floor to the general enrollment floor, and if a call in another direction occurs on the predicted congestion floor within a predetermined time, the call button is automatically registered. Cancel registration and
The elevator car pre-distributed to the predicted congestion floor is an external system-linked vehicle dispatch system characterized by waiting for the door to open on the predicted congestion floor. The external system linked vehicle dispatch system according to claim 1.
An external system-linked vehicle allocation system characterized in that the operation results of a plurality of elevator devices are information on the number of users on each floor and the direction of movement. It is an external system linked vehicle allocation method that manages the operation of multiple elevator devices installed in facilities such as buildings.
The operation results of multiple elevator devices and the information of the operation results obtained from the event information from the external system are stored and learned as past experience data, and the learned memory information and the event information are used by the user for each floor of the landing. The number is predicted, and when the time information including the delay of the event information and the current time match, congestion is predicted to occur.
The number of elevator cars to be pre-distributed to the predicted congested floor where congestion is expected to occur is determined using the user's occupancy rate in the elevator device, and the elevator car is installed at the landing on the predicted congested floor where inflow is predicted based on the event information. Call button automatic registration is performed in the direction from the predicted congestion floor to the general enrollment floor, and if a call in another direction occurs on the predicted congestion floor within a predetermined time, the call button automatic registration is canceled.
An elevator car that has been pre-distributed to a predicted congested floor is an external system-linked vehicle dispatch method that is waiting for the door to open on the predicted congested floor. The external system-linked vehicle allocation method according to claim 3.
An external system-linked vehicle allocation method, characterized in that the operation record of a plurality of elevator devices is information on the number of users on each floor and the moving direction.

Images