JP2008280093A - Group supervisory operation system of elevator - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a group supervisory operation system of an elevator capable of controlling both of passing suppression to each landing call and waiting time reduction to a landing call of the whole building, while favorably considering balance. <P>SOLUTION: The group supervisory operation system estimates the generation of a car (except for a case of full) passing through a floor on which the landing call is generated, when the car assigned to a newly generated landing call, and decides the car assigned to the landing call according to the estimation. Therefore, the combination of the assigned car and the passing car can be considered, and passing suppression suitable for a situation of passing is enabled. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to a group management system that collectively manages the operation of a plurality of elevators, and relates to an elevator group management system that assigns an optimal elevator to registered landing calls in consideration of serviceability to users. is there.

  The elevator group management system provides a more efficient operation service to the user by handling a plurality of elevator cars as one group. Specifically, multiple elevator cars (usually 3 to 8) are managed as a group, and when a landing call is generated on a certain floor, one optimal car is selected from this group. Then, control is performed to allocate the previous landing call to the car.

  The current group management system is based on allocation control using an evaluation function based on the predicted waiting time. This is because when a new landing call is generated, the estimated waiting time of the landing call (new landing call and holding landing call) that each car takes is calculated, and the car with the smallest waiting time or the maximum The landing call is assigned to the car with the smallest waiting time or the car with the smallest average waiting time.

  However, the allocation control based on the predicted waiting time does not give priority only to a specific landing call, but controls the waiting time as long as possible (to equalize) with respect to the landing call of the entire building. . For this reason, in order to avoid a longer waiting time increase of other calls for a certain landing call, there is a case where a pass occurs even though the car is not full.

  This passage creates a situation in which passengers waiting for the elevator at the landing may be uncomfortable.

  Therefore, there are conventional techniques disclosed in Japanese Patent Publication No. 63-23108, Japanese Patent Publication No. 2-333629, and Japanese Patent Publication No. 62-45151 in view of measures for this passage.

  Japanese Examined Patent Publication No. 63-23108 discloses that, when a car other than an assigned car passes at a landing where a certain landing call is generated, the passing condition of the car for the landing call is tightened thereafter.

  Japanese Patent Publication No. 2-333629 discloses a method for evaluating the selection of a service elevator for a new landing call with the aim of avoiding a car call arrival. The method shown here adds the car call first arrival time (the difference between the car call arrival time of the other cars and the predicted waiting time of the temporarily assigned car) to the predicted waiting time for the new landing call. Thus, the estimated waiting time is evaluated by adding a correction to the first call arrival. Thus, it is described that it is possible to prevent as much as possible the occurrence of a car call arrival.

  Also, in Japanese Examined Patent Publication No. 2-38150, the registration time for landing call (an indicator similar to the waiting time) is corrected when a full passage or a car call arrival occurs for a full pass or a car call arrival. An example of group management control in which a relatively preferential service is given by evaluating a call with a long apparent registration time is disclosed.

Japanese Patent Publication No. 63-23108 JP-B-2-33629 Japanese Patent Publication No. 2-38150

  In the conventional technologies listed above, it is difficult to control in a well-balanced manner both the passage suppression for each hall call and the waiting time reduction for the hall call of the entire building.

  First, the method disclosed in Japanese Examined Patent Publication No. 63-23108 discloses that when a passage restriction command is issued when the passage is judged, the car is stopped when the passage restriction command is generated. However, it is difficult to consider the waiting time for the entire hall call.

  In Japanese Examined Patent Publication No. 2-33362 and Japanese Examined Patent Publication No. 2-38150, the case where the car is not full is not considered.

  Therefore, an object of the present invention is to provide an elevator group management system capable of controlling in a well-balanced manner both the passage suppression for each landing call and the waiting time reduction for the entire building call.

  In order to achieve the above object, the present invention predicts the occurrence of a car that passes through the floor where the landing call is generated (excluding full passengers) when determining the car assigned to the newly generated landing call. In accordance with this prediction, the car assigned to the landing call is determined.

  Thereby, it is possible to consider the combination of the allocation case and the passage, and it is possible to more appropriately suppress passage according to the situation at the time of passage.

  According to the present invention, the elevator passing through the floor of the landing place where the new hall call is generated is predicted at the time of car assignment, and the car assigned to the landing call is determined based on this prediction. It is possible to control in a well-balanced manner both the passage suppression and the waiting time reduction for the hall call of the entire building.

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

FIG. 1 shows an example of the configuration of an elevator group management system according to the present invention. Here, an example of group management consisting of three elevator cars is shown. The main components are three elevator cars 31, 32, 33, control of direction, speed, stop position, etc. based on assigned hall call and car call for each elevator, registered car call information, It is installed in the control devices 21, 22, 23, and the floors 41, 44 that collect car load information for estimating the number of passengers, car status information such as car position / direction / speed, etc., and transmit it to the group management side. The landing call registration devices 42 and 45 for registering the direction (up or down) that the user wants to go to. When a landing call is registered on a certain floor, the best service is provided from three elevators for that call. Elevator group management device 1 that evaluates and assigns elevators, and a landing lantern (guide light) 43 that guides the user of that floor by lighting which elevator is assigned.
46, car position display indicators 51 and 52 for displaying the position of each car at the landing. There are two types of landing lanterns (guide lights) that guide the assigned elevators: when the elevator guides just before the elevator arrives at the landing, and when the landing call is instantly announced at the time of registration, the former is a guidance system on arrival The latter is called an immediate reservation guidance system. In the case where the car position display indicators (51, 52) as shown in FIG.

  The following is a brief description of the circumstances underlying the present invention. The car position display indicators 51 and 52 at the landing are not installed in normal group management, but may be strongly desired by some customers. In that case, install them after understanding the adverse effects described later. There is a case. In the group management, in order to assign an appropriate elevator to the landing call registered on each floor in consideration of the situation of the entire landing call, an elevator that is not the assigned elevator is given first depending on the landing call. May pass through. For example, when there is a call that has been waiting for about 40 seconds on the floor ahead of the call that passes, it is desirable to pass in order to suppress long wait. Thus, in group management, since the waiting time of the entire hall call is emphasized, the occurrence of passing is unavoidable, and for this purpose, the movement of each car is not shown. However, when the car position display indicator is installed according to the customer's request as described above, since the passing of the car is known, the user on the passing side is frustrated and frustrated.

  In recent high-rise buildings and large-scale buildings, there is an increasing number of cases in which the hoistway with excellent design features a glass-fitted see-through type elevator. In this case, even if there is no car position display indicator, the passing of each car can be seen from the landing, so the users on the passing side are frustrated and frustrated. In the case of a see-through type elevator, since there is no car position display indicator, an immediate reservation guidance system (a system for guiding which elevator is assigned at the time of landing call registration) is often used. The user waits in front of the elevator that is informed about the reservation, and sees the other elevators passing first, so dissatisfaction and annoyance are raised.

  In this way, for elevator group management where the position of each car can be known at the landing, for example, when a new landing call is registered, control for assigning the elevator that can arrive the earliest to that call is, for example, It has been implemented. In this case, the possibility that other elevators will pass first for that landing call will certainly decrease, but the waiting time for other landing calls will not be taken into account, so for example, a longer waiting call will be longer A case arises.

  As described above, in an elevator in which the position of each car is known at the landing, it is strongly desired to suppress the passage while considering the reduction of the waiting time of the entire landing call, and the present invention solves this problem. It is intended.

The elevator group management apparatus 1 is characterized by the car arrangement data storage unit 113, the weight data storage unit 114 based on the car position relationship, the passage occurrence prediction unit 109, the passage prediction data storage unit 112, and the prediction based on the passage situation. It becomes the waiting time adjusting unit 116. The solution principle of these elements is briefly described below. At the stage of selecting an assigned car, the occurrence of passage for each landing call is predicted, the estimated waiting time until the assigned car serves the landing call, and the positional relationship for each passing car (car arrangement data storage) Section 113, weight data storage section 114 according to car position relationship, operation of passage prediction data storage section 112), number of passages (operation of passage occurrence prediction section 109, passage prediction data storage section 112), etc. The waiting time of the entire hall call is adjusted by adjusting the value of the predicted waiting time according to the situation (operation of the forecasting waiting time adjusting unit 116) and reflecting the influence of passage in the waiting time according to the situation. It is possible to suppress the passage while considering the above.

  Details of the elevator group management control device 1 will be described below. The landing call data storage unit 101 stores data such as floors, directions, and assigned cars for unanswered calls among the landing calls registered by the landing call registration devices 42 and 45 on each floor. A specific example is the hall call table shown in FIG. 10A (details of FIG. 10 will be described later). The car call data storage unit 102 sends the car call data registered in each car (31, 32, 33) via the control devices (21, 22, 23) of each car, and stores the floor, Data such as a car is stored.

  When a new landing call is registered, the new car assignment execution determination unit 103 determines whether or not to detect the new landing call and determine a car to be assigned to the call. Then, the temporary allocation car setting unit 104 performs a process of temporarily allocating each of the cars of Nos. 1 to 3 to a new landing call. For each car, a predicted arrival time table as shown in FIG. 8A is calculated by the predicted arrival time table calculation unit 105 (details of FIG. 8 will be described later). Thus, the estimated arrival time of each car for each landing call and each floor is known. If a car at a new landing call is temporarily assigned, the estimated arrival time of the car changes. This is calculated by the temporary allocation car arrival prediction time table calculation unit 106, which calculates the predicted arrival time table at the time of temporary allocation as shown in FIG. 8B. In the temporary assignment landing call data storage unit 108, the landing call table is updated as shown in FIGS. 10A to 10B, for example, corresponding to each temporary assignment.

  The passage occurrence prediction unit 109 compares the estimated arrival times of the respective cars with respect to the landing call assigned to each car for each temporary assignment, and predicts the occurrence of passage of another car with respect to the assigned car. Here, the car to be examined for passing is detected by detecting the direction state of the car from the arrival prediction time table and the like, and the object is limited to a directional car (a car having either an upward direction or a downward direction). The reason will be explained later.

  When the passage is predicted, information on the landing call (floor, direction) that has passed, the assigned car at that time, the number of passes, and the passing car is stored in the pass prediction data storage unit 112. An example is the passage prediction data table of FIG. 14 (details of FIG. 14 will be described later).

  If a car other than the assigned car has actually passed through the unanswered hall call, it is detected by the passage occurrence detection unit 110 and the hall call (floor, direction) that has passed through, Information such as the assigned car at that time, the number of passes, and the passed car are stored in the passing occurrence storage unit 111, and further, at the time of temporary allocation evaluation, the information is sent to the passing prediction data storage unit 112 and passed by prediction. Data and the actual passing data are used for allocation evaluation in a lump. Here, the actual passing data is confirmed information, while the predicted passing data is not determined but is only a prediction, so a flag for distinguishing between them is attached and the actual passing data is important. It is more effective to use it for evaluation at a higher degree. Below, in order to demonstrate easily, it demonstrates without distinguishing both.

  The car arrangement data storage unit 113 stores arrangement data on a horizontal plane on each elevator platform. Here, the arrangement of the elevators refers to, for example, an arrangement as shown in FIG. 18A (in this example, four linear arrangements) (details of FIG. 18 will be described later). The weight data storage unit 114 based on the car position relationship stores the distance index for the position relation of each car or the data of the weight coefficient calculated by the distance index based on the car arrangement data. The distance index for the positional relationship of each car is stored in matrix format data as shown in FIG. In this matrix, the row direction represents the assigned car and the column direction represents the passing car. Accordingly, a weighting factor corresponding to a corresponding distance index and a positional relationship derived therefrom can be calculated by a combination of passing cars and assigned cars. Details of FIG. 18B will be described later.

In the predicted waiting time adjusting unit 116 according to the passing situation, the passing prediction table data in the predicted passing data storage unit 112 (FIG. 14A is an example), and the distance index between each car in the weight data storing unit 114 based on the car position relationship. The matrix of data (FIG. 18B is an example), the predicted waiting time calculated by the predicted waiting time calculation unit 115, the positional relationship of the assigned cars, the number of passing cars, and the predicted waiting time of passing platform calls. The length of the predicted waiting time is adjusted according to the length of. The specific processing method will be described later with reference to FIGS.

  Here, the estimated waiting time for each unanswered landing call is calculated based on the estimated arrival time table in the estimated arrival time table calculation unit 105 for each car and the landing call duration data in the landing call duration storage unit 107 for each car. It can be obtained by adding both. The landing call continuation time represents the elapsed time from the landing call registration to that point in time, and the estimated arrival time represents the estimated time until allocation or arrival. Therefore, the estimated waiting time from registration of the call to arrival of the assigned car is obtained by adding both. In the predicted waiting time calculation unit 115, this calculation is executed.

In the waiting time evaluation value calculating unit 117, the predicted waiting time calculated by the predicted waiting time calculating unit 115, and when the passage is predicted or has already occurred, the predicted waiting time adjusting unit 116 according to the passing situation is used. The waiting time evaluation value is calculated for all unanswered landing calls or unanswered landing calls that are affected by temporary allocation (waiting time changes etc.) according to the predicted waiting time adjusted for passing action. . For example, the evaluation value is derived from the maximum value, the sum, the sum of squares, the sum of values converted by a function, and the like. This evaluation value is calculated for each temporary allocation car. For example, in the case of FIG. 1, the evaluation value Φ (k = 1) at the time of temporary allocation of the first unit, the evaluation value Φ (k = 2) at the time of temporary allocation of the second unit, the evaluation value Φ ( k = 3) is calculated respectively. Here, Φ represents an evaluation function (for example, sum total) for calculating an evaluation value, and k represents a temporarily assigned car number. The evaluation value comparison unit 118 compares evaluation values for each temporary allocation car, for example, Φ (k = 1), Φ (k = 2), Φ (k = 3), and determines the minimum evaluation value and the temporary value at that time. Find the assigned car number by sorting. The allocation car determination unit 119 determines the allocation to the temporary allocation car that gives the minimum evaluation value obtained by the evaluation value comparison unit 118.

The above is a rough flow of processing from the control configuration of the elevator group management system according to the present invention shown in FIG. 1 to the allocation car determination when a new landing call is registered. Although the description of the adjustment function changing unit 120, the passage result diagnosis unit 121, and the communication unit 112 in FIG. 1 is omitted, the details will be described later. The outline is as follows. Adjustment function changer 120
Performs the process of appropriately modifying and changing the adjustment method of the estimated waiting time based on statistical data (histogram, average value, etc.) of waiting time (or landing call duration) when the passage actually occurs Similarly, based on the statistical data of the waiting time (or landing call duration) when the passage actually occurs, the diagnosis unit 121, if this exceeds a certain standard, via the communication unit 122, the elevator management company Process to report to.

  FIG. 2 shows a flowchart of an assigned car determination process for a new hall call in the elevator group management system according to the present invention. Hereinafter, the flow will be described.

  First, the contents of the table data for the car call and the landing call are updated based on the call information collected at that time (ST100). The details are as shown in the flowchart of FIG. Hereinafter, the table data update process for the car call and the landing call shown in FIG. 3 will be described. First, the latest data of each car, landing call, and car call state is input (ST200). Next, based on the data, arrival prediction time table data AT (i, j, k) of each car is updated for each i, j, k (ST201). Here, i represents a floor, j represents a direction (up or down), and k represents a car name. Furthermore, the landing call duration table data HT (i, j, k) of each car is updated (ST202), the landing call table data HC (i, j) is updated (ST203), and the process ends.

  Returning to the flowchart of FIG. 2, after the update of each table data is completed, it is checked whether or not a new landing call is registered (ST101). If registered, the floor where the new hall call is generated is the in floor and the direction is the jn direction, and the process proceeds to the next process. If not, return to the previous step and repeat the process.

  In the case of new landing call registration, the temporary assigned car number k is first set to k = 1 (ST102), and a process of temporarily assigning a new landing call to the kth car is executed (ST103). Since the call state changes, the contents of the table data for the car call and the landing call are updated again (ST104). This process executes the process of the flowchart of FIG.

As a specific example, consider the situation shown in FIG. This represents an example of group management with four elevators in a 12th floor building. About 4 elevators, the first car A1 is upward on the second floor, the second car A2 is on the eighth floor, the third car A3 is on the second floor, the fourth car A4 is on the fifth floor positioned. For each car, Unit 1 has a car call A6 on the 11th floor, Unit 2 has a car call A7 on the first floor, and Unit 3 has a car call A8 on the 12th floor. No. 4 is assigned a landing call A5 on the 9th floor, and 10 seconds have passed since this landing call was generated. Therefore, the landing call duration of this landing call is 10 seconds. In this state, consider a situation in which a new upward landing call A9 is registered on the seventh floor.

  The arrival prediction time table of Unit 1 before provisional allocation at this time is as shown in FIG. The car position B1 is on the second floor, and the time increases clockwise from there. Here, it is assumed that the movement of the first floor is 2 seconds, and the stop time by calling is 10 seconds. FIG. 8B shows an estimated arrival time table when a new landing call B2 in the upward direction on the seventh floor is temporarily assigned to the first car. Since it stops in the 7th floor upward direction, it becomes a table value in which 10 seconds of stop time is added thereafter.

  The landing call duration table is as shown in FIG. FIG. 9 (a) shows a landing call duration table before the temporary allocation of the No. 4 vehicle, and the duration of the 9th floor upward landing call already assigned is 10 seconds. FIG. 9B shows a case where a new landing call in the 7th floor is temporarily assigned to the No. 4 machine, and a zero value is newly added as landing call duration data in the 7th floor upward.

  The landing call table is as shown in FIG. Before provisional assignment to a new landing call, a landing call is assigned only to No. 4 unit, which is represented as shown in FIG. When a new landing call on the 7th floor is temporarily assigned to Unit 1, this result is added as shown in FIG.

  As described above, when the table data of the car call and the landing call, which are basic data, are updated, next, for each temporary call, prediction of occurrence of passage and calculation of predicted waiting time for each landing call Reflecting the passing action on the predicted waiting time, and calculating the waiting time evaluation value, the assignment is determined to the car having the minimum waiting time evaluation value. Hereinafter, the process flow of FIG. 2 will be described.

  First, a floor call is detected by scanning each floor and direction. First, floor i = 1 and direction j = UP are determined as initial values (ST105). Check if the hall call is registered in the i-th floor and j-direction (ST106). If the hall call is registered (including registration by provisional assignment), pass other cars to the assigned car of the hall call. Prediction processing is performed (ST107). Details of this will be described later with reference to the flowchart of FIG. If it is determined by the passage prediction process that a passage will occur (ST108), the number of passages in that case (when a plurality of cars pass) and the passing car number are stored (ST109).

  Next, the predicted waiting time for the assigned car of the detected hall call (i floor, j direction) is calculated (ST110). The calculation of the predicted waiting time is obtained by adding the data HT (i, j, k ′) in the landing call duration table and the data AT (i, j, k ′) in the predicted arrival time table. Here, k ′ represents an assigned number corresponding to the landing call.

  When passing occurs, the predicted waiting time obtained previously is adjusted according to the positional relationship between the passing car and the assigned car and the number of passing times (ST111). Details of the method of adjusting the predicted waiting time will be described later with reference to FIG.

  In this example, the estimated waiting time calculated in process ST111 (when there is no passage) is used as the waiting time evaluation for the total waiting time (including adjustment) of all landing calls for each temporary assigned car. Is added to the waiting time evaluation value of the k-th car as a waiting time for a floor call in the i-th floor and j-direction when the temporarily assigned car is the k-th car (ST112). When the maximum value of the waiting time is set as the evaluation value, the maximum value so far is compared with the waiting time for the hall call in the i-th and j-th directions, and the larger value is updated as the maximum value so far.

  The values of the floor i and the direction j for searching for the hall call are updated (ST113), and the process returns to the process ST106 and the series of processes is repeated until the scanning is completed for all the floors and directions (ST114).

  When scanning has been completed for all the floors and directions, one temporary assigned car number k is added to make the next number a temporary assigned number (ST115). Until temporary allocation is completed for all the units, the process returns to the process ST103, and the process is repeated to calculate a waiting time evaluation value for each temporary allocation unit (ST116).

  When the processing is completed for all temporary assigned cars (ST116), the car number with the smallest waiting time evaluation value is searched (ST117), and the assignment to this car number is determined (ST118).

The point in the allocation determination process of FIG. 2 is that every time a temporary allocation machine is set, all landing calls are searched, and the occurrence of each pass and its status (for example, combination of assigned cars and passed cars, etc.) ) And reflecting it in the waiting time evaluation (ST107 to ST107 in FIG. 2)
(Process up to ST112).

  The details of the passing car search process for the hall call assigned car will be described below with reference to FIG. In FIG. 4, first, an unanswered hall call is searched. Initially set to floor i = 1, direction j = UP (ST300), and referring to the landing call table HC (i, j) (example in FIG. 10), if HC (i, j) is not zero, If there is an unanswered assigned machine, the process proceeds to the next process (ST301). Otherwise, the floor i and direction j are updated to the values of the next retrieval floor according to the flowchart shown in FIG.

  Hereinafter, the flowchart of FIG. 6 will be described. First, it is determined whether or not the direction j of the search floor is UP (upward) (ST210). If it is UP, the floor value i is incremented by one (ST211). It is determined whether or not the value of i after the addition is equal to or greater than the value iMAX on the top floor. If it is greater than or equal to iMAX, the direction j is changed to DN (downward) and the value of i is set to the top floor. Return to the section (a). Returning to the process ST210, if the direction j is not UP, the direction is the downward direction and the floor value i is decremented by one (ST214). If the value of i after subtraction is less than or equal to the value 1 of the lowest floor (ST215), it means that the search of the floor for one lap from the upward direction to the downward direction has been completed, and the value of the provisionally assigned car number k is Add one and go to the next temporary allocation car. The above is the processing for updating the value of the search floor.

Returning to the process ST301 in FIG. 4, if the hall call table data HC (i, j) is not zero, there is an unanswered hall call in the i-th and j-th directions, and the assigned car name for that hall call is set. Set to kh. Since the assigned car name for the hall call in the i-th floor and j-direction is stored in HC (i, j), kh = HC (i, j) is set (ST302). For example, in the situation shown in FIG. 7, when a new hall call (7th floor UP) is temporarily assigned to Unit 1 (k = 1), the hall call table is as shown in FIG. When the loop value is i = 7, j = UP, kh = HC (i = 7, j = UP) = 1, i = 9, and j = UP, kh = HC (i = 9, j = UP ) = 4.

  In the following processes (ST303 to ST312), it is predicted whether a car other than the assigned car (unit kn) will pass for the found landing call (i floor, j direction). First, a variable kk_ps () representing a passing car name is initialized (ST303). The value of the number of passages is entered in the parentheses of kk_ps (), and the name of the machine at that time is in kk_ps (). For example, if the second pass occurs and the first pass car is No. 2 and the second pass car is No. 3, then kk_ps (1) = 2 and kk_ps (2) = 3. Also, a variable cn representing the number of passages is initialized to zero (ST304).

Next, the passing car number is searched by the kk loop for the target landing call (i floor, j direction). First, kk = 1 is initialized (ST305). Next, it is checked whether the kk machine is a directional car (a car having an upward direction or a downward direction) (ST306). If it is a directional car, the process proceeds to the next process. That is, in the case of a non-directional car, the process is skipped to process ST311. Here, a car with direction refers to a car that has a landing call or car call, or a car that is running without a call (for example, a car heading to a standby position). , Refers to a parked car waiting. The point of this processing is that the passing target is only a directional car. At that time, the non-directional car is less likely to operate immediately thereafter, and is considered to keep the stopped state as it is, and is excluded from the target of passage prediction. More precisely, for a car that is not directional at the time of the new landing call, if it becomes directional due to temporary allocation, it will be subject to passage prediction, and if it is not temporarily allocated, it will remain non-directional. Yes, exclude from the target of passing prediction. As a result of such processing, it is possible to predict the occurrence of passage more accurately, including a non-directional car. The handling of the non-directional car will be supplemented later using FIG.

  When kk No. is directional, it is checked whether or not kk No. is equal to kh No. (assignment number of target landing call) (ST307). When not equal, it progresses to the next, and when equal, since the number machine is not a passing number machine, processing is skipped to process ST311.

  If kk No. is not equal to kh No. (= if kk No. is not the assigned call for the target call), the occurrence of passage is predicted by comparing the predicted arrival times for both landing call occurrence floors (ST308). Specifically, when the following equation is satisfied, the kk machine arrives at the landing call generation floor (i direction, j floor) earlier than the kh machine, and it can be predicted that the passage will occur.

AT (i, j, kk) <AT (i, j, kh) Equation (1)
FIG. 11 and FIG. 13 show the contents of the process ST308 for the elevator and call situation shown in FIG. First, FIG. 11 will be described. FIG. 11 is a diagram showing the estimated arrival time of each car at each landing call generation floor when each car is temporarily assigned to the situation shown in FIG. FIG. 11A shows the situation when the first car is temporarily assigned, and the landing call here is 7th floor UP (first car temporary assignment: C1) and 9th floor UP (4th car is already assigned). First, for a call on the 7th floor UP, kh = 1, and the predicted arrival time of each car to this floor by the kk loop is 10 seconds (kk = 1 unit = kh), as shown in the figure.
36 seconds (kk = 2), 10 seconds (kk = 3), and 4 seconds (kk = 4) are detected, and the fourth vehicle is detected as a car that satisfies Equation (1). That is, at the time of the first allocation of the first unit, the occurrence of the passage of the fourth unit is predicted for the call of the 7th floor UP (the first allocation of the first unit: C1). This is arranged in a table format in the first column of the table of FIG. Unit 1 is provisionally allocated to the new floor call 7th floor UP, and its estimated arrival time is 10 seconds (E1 in FIG. 13 (a)). The estimated arrival time of each unit for this floor is as shown in the table of FIG. 13 (a), the estimated arrival time of Unit 4 is 4 seconds (E3 in FIG. 13 (a)), and Unit 4 passes first. Can be predicted. Returning to FIG. 11 (a), for another landing call 9th floor UP, since the assigned car is No.4, kh = 4, and the predicted arrival time of each car to this floor by the kk loop is 24 seconds ( kk = 1), 40 seconds (kk = 2), 14 seconds (kk = 3), and 8 seconds (kk = 4 = kh), and it is detected that there is no car satisfying the equation (1). That is, it is predicted that there is no passing car for the 9th floor UP call. The second column of the table of FIG. 13 (a) shows this. For the call 9 floor UP that has already occurred, the assigned car is No. 4, and the estimated arrival time of each car is shown in FIG. It becomes like the 2nd column of the table of b). Since there is no car smaller than the estimated arrival time 8 seconds (E2) of the assigned car No. 4, it can be predicted that no passage for this call will occur.

Similarly, examples in which the temporary allocation number machine is set to No. 2, No. 3, No. 4 are shown in FIGS. 11 (b) and 13 (b), FIG. 11 (c), FIG. 13 (c), and FIG. (D) and FIG. 13 (d). The outline of each will be described below. At the time of the temporary allocation of Unit 2 as shown in FIG. 11B and FIG. 13B, the 7th floor UP (the temporary allocation of Unit 2) is called (C2 in FIG. 11B, E4 in FIG. 13B) ), The estimated arrival time of Units 1, 3, and 4 is small, and it is predicted that these three units will pass. For calls on the 9th floor UP (unit 4 is already assigned), it is predicted that no passage will occur. At the time of the temporary allocation of Unit 3 as shown in FIGS. 11 (c) and 13 (c), the 7th floor UP (No. 3 temporary allocation) call (C3 in FIG. 11 (c), E5 in FIG. 13 (c)) ) Is predicted to pass No. 4, and for calls on the 9th floor UP (No. 4 already assigned), it is predicted that no pass will occur. At the time of the temporary allocation of the fourth unit, it becomes as shown in FIGS. 11 (d) and 13 (d), and the seventh floor UP (the temporary allocation of the fourth unit) is called (C4 in FIG. 11 (d), E6 in FIG. )), It is predicted that no passing will occur, and for the 9th floor UP (No. 4 already assigned) call (C5 in FIG. 11 (d), E7 in FIG. 13 (d)) And Unit 3 is expected to pass.

  As described above, for each new landing call, a temporary assigned car is set, and for each temporary assigned car case, the landing call is searched, and the car passing for each landing call is The processing shown in FIG. 11 and FIG. 13 determined by comparing the estimated arrival times is one of the features of the present invention. This process corresponds to the fact that the simulation according to the actual situation is performed based on the estimated arrival time, and the passage situation can be predicted more accurately.

  Returning to the processing of FIG. 4, processing after processing ST309 will be described. When the occurrence of passage is predicted, one variable cn representing the number of passages is added each time (ST309). For the variable kk_ps () representing the passing car number, the number of passing times and the passing car name kk are input as kk_ps (cn) = kk (ST310). Thus, the pass prediction determination process for the kk machine is completed, the kk is updated to the next car (ST311), and the process is repeated until the search is completed for all the car cars (ST312).

  For all car units, when the pass prediction determination process for the assigned car for the landing call (i floor, j direction) is completed, it is determined by cn whether or not a pass occurrence is predicted (ST313). If cn is positive, the data of the floor i, the direction j, the assigned car kh, the predicted passing number cn, and the passing car number kk_ps (cn) are written in the passing table data for the temporarily assigned car number k (ST314). .

  For example, when a new hall call is temporarily assigned to Unit 1, as described with reference to FIGS. 11 (a) and 13 (a), for the 7th floor UP call (Unit 1 is temporarily assigned), Unit 4 is used. Therefore, cn = 1 and kk_ps (1) = 4. In addition, since no passage occurs for the 9th floor UP direction call (assigned No. 4 machine), cn = 0. Accordingly, the passage prediction data table is as shown in FIG. When the first car is temporarily assigned, the first floor is assigned to the hall call in the UP direction on the 7th floor, and the number of times of passage is one, and it is predicted that the fourth car will be passed. Recorded in the table. Similarly, FIGS. 14B, 14 </ b> C, and 14 </ b> D represent passage prediction tables when the second, third, and fourth machines are temporarily allocated. For example, at the time of the temporary allocation of Unit 2, it is recorded that the second floor is assigned to the 7th floor UP direction call, and that the number of passages is three times and the passage of Units 1, 3, and 4 is predicted Yes.

  Details of the predicted waiting time adjustment process (ST111) according to the positional relationship of the passing car and the assigned car and the number of times of passing in the overall flow chart for determining the assigned car for the new landing call in FIG. 2 will be described. FIG. 5 shows the details of the predicted waiting time adjustment processing for occurrence of passage. The feature of this processing is that the action of passing is reflected in the estimated waiting time of the passing hall call in a form that depends on the situation (number of times of passing, position of the passing car and assigned car, estimated waiting time of the landing call). In the point. If a passing event is predicted, increasing the waiting time according to the psychological effect seen by the user of the passing event makes it difficult to allocate the car (the passing car) by the increased amount. Will be avoided.

Hereinafter, the flowchart of FIG. 5 will be described. First, the predicted waiting time wt of the assigned car kh for the i-th floor and j-direction landing calls is calculated (ST400). Here, the flowchart of FIG. 5 corresponds to the inside of the process ST112 of the flowchart of FIG. 2, and the values in the i-th floor and the j-direction inherit the values of the flowchart of FIG. Also, the process ST in the flowchart of FIG.
107 corresponds to the flowchart of FIG. 4, and kh is set by the process ST302 of the flowchart of FIG. The predicted waiting time wt of the kh machine is obtained by the following equation.

wt = HT (i, j, kh) + AT (i, j, kh) Equation (2)
FIG. 16 is a diagram showing the predicted waiting time for each landing call in a table format for each assigned car. For example, FIG. 16 (a) shows the predicted waiting time for each landing call at the time of temporary allocation to No. 1 unit. The predicted waiting time for a call in the 7th floor UP direction (No. 1 unit is temporarily allocated) is 10 seconds, 9th floor The estimated waiting time for UP direction calls (assigned by Unit 4) is 18 seconds.

Next, it is checked based on the number of passages cn whether or not a passage occurrence is predicted for a floor call in the i-th floor and j-direction (ST401). When cn is positive, the occurrence of passage is predicted, and calculation is performed to reflect the influence on the waiting time for each passage. Next, when variable m is a parameter for the number of passes, and cn is positive (ST401), m = 1 is initialized (ST402). From the passing data table (FIG. 14), the m-th passing car number is referred to and the variable kp is set to the number (ST403).

  In the following processing, first, the degree of influence due to passage is calculated based on the positional relationship of the car that passes through the assigned car. Specifically, if the distance between the car and the assigned car is close, the waiting is easily recognized by the waiting customer, so the impact level is set to a high value. To a low value.

The actual procedure refers to the distance index d between the kh car (assigned car) and the kp car (passing car) from the car position relationship table CD () shown in FIGS. 18 to 21 (ST404, equation (3)). .

d = CD (kh, kp) Equation (3)
Hereinafter, the car positional relationship table shown in FIGS. 18 to 21 will be described. FIG. 18 is a diagram showing an example of the car position relationship table, and FIG. 18 (a) shows the arrangement of elevators. FIG. 18 (a) shows a horizontal sectional view of the elevator hall on a certain floor, which is a cross section G1 of the hoistway of the first unit G1, a hoistway cross section G2 of the second unit, a hoistway cross section G3 of the third unit, and a hoistway of the fourth unit It has a section G4 and a landing floor G5. When the four elevators are arranged in a straight line in a line as shown in FIG. 4A, the car position relationship table is as shown in the table of FIG. 18B from the distance relationship between the elevators. In the table of FIG. 18B, the row direction represents the assigned car number, the column direction represents the passed car number, and the values in the table indicate the distance index of the passed car relative to the assigned car. Represents. Here, the distance index is defined as a distance index of one unit when adjacent to each other. For example, the distance index for the first car is 1 for the second car adjacent to the first car, and 1 for the second car and the fourth car. Therefore, when the assigned car is No. 1 and the passing car is No. 4 (corresponding to the cases of FIGS. 11 (a) and 13 (a)), the distance index d is as follows from the car position table of FIG. 18 (b). It becomes like the formula.

d = CD (kh = 1, kp = 4) = 3 Formula (4)
FIG. 19 shows another example of the car position relationship table. FIG. 19A shows the arrangement of the elevators, and the arrangement of two units is facing each other (four units in total). The distance index in this case is defined as 1 unit when adjacent and 2 units when facing each other. For example, the distance index of each unit viewed from the first unit is 1 for the second unit, 2 for the third unit facing, and 3 for the fourth unit. Accordingly, the car position relationship table is as shown in FIG.

  FIG. 20 also shows another example of the car position relationship table. FIG. 20A shows the arrangement of the elevators, and the facing arrangement of 3 units (6 units in total). The distance index in this case is also defined as 1 unit when adjacent and 2 units when facing each other. For example, the distance index of each unit viewed from Unit 1 is 1, Unit 2 is 2, Unit 3 is 2, 2 is facing, Unit 5 is 3, and Unit 6 is 4. Accordingly, the car position relationship table is as shown in FIG.

  Returning to the flowchart of FIG. 5, after referring to the assigned car distance index d from the car position relation table (examples in FIGS. 18 to 20) (ST 404), the car position relation coefficient α is expressed by the following equation based on d: Thus, it is calculated as the reciprocal of d (ST405).

α = 1 / d Equation (5)
FIG. 15 shows a specific example of the assigned car distance index d (F1 region in the figure) and the positional relationship coefficient α (F2 region in the figure) for the case shown in FIG. Here, it is assumed that four elevators are arranged in a straight line (FIG. 18A), and the table in FIG. 18B is applied to the car positional relationship table. FIG. 15A shows each value at the time of temporary allocation of the first unit. In this case, the predicted passage occurrence is as shown in the passage prediction data table of FIG. 14A, and since the assigned car is No. 1 and the passing car is No. 4, the car position relation table of FIG. The distance index is 3, and the positional relationship coefficient is 0.33. FIG. 15B shows each value at the time of temporary allocation of Unit 2. The predicted passage occurrence at this time is as shown in the passage prediction data table of FIG. 14B, and the assigned car is the second car and the passing cars are the first, third, and fourth cars. Accordingly, from the car position relationship table of FIG. 18B, the distance indexes are 1, 1 and 2, respectively, and the position relationship coefficient is 1, 1, 0.5. In the case of the temporary allocation of the No. 3 and No. 4 units, the distance index and the positional relationship coefficient shown in FIGS. 15C and 15D are obtained from the passage prediction data tables of FIGS. 14C and 14D, respectively.

  Returning to the flowchart of FIG. 5, when the positional relationship coefficient α is obtained, the passing situation is considered from the value of α, the number of times of passage cn, and the predicted waiting time wt of the landing call (i-th floor, j-direction landing call). Adjustment processing of the predicted waiting time is performed (ST406). This process is expressed by a recurrence formula as shown below.

wt _n = F (wt _n−1 , α) Equation (6)
1 ≦ n ≦ cn Formula (7)
wt ₀ = wt Formula (8)
Here, wt _n represents the adjusted predicted waiting time value, and the recurrence formula of Expression (6) is recursively repeated, and becomes a final value when n = cn. F () represents a function for adjusting the length of the waiting time (hereinafter referred to as a waiting time adjusting function), and specific examples thereof are shown in FIGS.

  First, referring to FIG. 25, the operation of the recurrence formula of Equation (6), specifically, the operation of recursively calculating the influence of passage will be described. In order to simplify the explanation in this figure, the recurrence formula of the formula (9) excluding the positional relation coefficient α will be used.

wt _n = F (wt _n-1 ) Formula (9)
In the graph of FIG. 25, the horizontal axis represents the prediction waiting time wt _n−1 that becomes the input of the recurrence formula of Expression (9), and the vertical axis represents the adjusted prediction that becomes the output of the recurrence expression of Expression (9). This represents the waiting time wt _n . The number of passes cn is assumed to be cn = 3 (three occurrences of passage). The function F () for adjusting the waiting time is a function of the following equation.

wt _n = F (wt _n-1 ) = wt _n-1 + (1/2) · wt _n-1 formula (10)
This is represented by the straight line H1 in the graph of FIG. The predicted waiting time wt is assumed to be wt = 30 (seconds). Accordingly, the initial value wt ₀ = wt = 30 (seconds). The point J1 in the graph of FIG. 25 represents wt ₀ , and the adjusted waiting time wt ₁ = 45 (seconds) by the first pass obtained from the equation (10) is represented by the point J2 on the graph. Is done. In the graph of FIG. 25, the dotted line H2 represents a straight line having an inclination of 1 passing through the origin (thus, wt _{n + 1} = wt _n ), and the difference P1 from H1 is the adjustment of the waiting time for the first passage. In other words, it becomes a value obtained by converting the influence of passage into a waiting time. In this case, the increase is 15 seconds. Next, the waiting time adjustment by the second passage is calculated from Equation (10) as wt ₂ = (3/2) · wt ₁ . When viewed on the graph, the result is that wt ₁ is the J3 point (45 seconds) in the graph, and wt ₂ is the J4 point (67.5 seconds) in the graph. Therefore, the waiting time adjustment P2 due to the second passage is 22.5 seconds. The reason why the second adjustment amount increased with respect to the first adjustment amount is that the increase in the psychological influence (stress) of the waiting customer is taken into consideration. In other words, the psychology on the waiting side should have a stronger psychological effect on the second pass than on the first pass, and the effect will be reflected as the second increase relative to the adjustment of the first pass. Has been. Since the assigned car is determined based on this waiting time value, a measure is taken to minimize the assignment to the car that receives a plurality of passes by the above-described mechanism. Similarly, the waiting time for the third passage is wt ₂ = 67.5 seconds and wt ₃ = 101.3 seconds from the equation (10). On the graph, they are represented by points J5 and J6, respectively, and the adjustment P3 is 33.8 seconds. Thus, the adjustments for the first, second, and third passes are increased to 15 seconds, 22.5 seconds, and 33.8 seconds, respectively, taking into account the psychological effects of multiple passes. As a result, the original predicted waiting time is 30 seconds, but the adjusted waiting time of 101.3 seconds is reflected as a result of reflecting the psychological effect of three occurrences of passage based on the recurrence formula of Equation (10). To increase. Therefore, this allocation to the car is avoided as much as possible.

The waiting time adjustment function F () is a function that reflects the waiting time psychological influence due to the occurrence of passage in the waiting time, but emphasizes the suppression of passing by its input / output characteristics, or emphasizes the actual waiting time rather than passing. It is possible to Specific examples of the waiting time adjustment function F () are shown in FIGS. 21 to 24, and each feature will be described below. First, the waiting time adjustment function in FIG. 21 is the same function as Expression (10), and is expressed as follows (here, α is omitted). A linear function allows any waiting time to be adjusted accordingly.

wt _n = wt _n-1 + (1/2) · wt _n-1 formula (11)
The waiting time adjustment function F () in FIG. 22 is expressed as in Expressions (12) and (13) (here, α is omitted).

wt _n = wt _n-1 (0 ≦ wt _n-1 <20) Formula (12)
wt _n = wt _n-1 + (1/2) · wt _n-1 -10 (20 ≦ wt _n-1 ) Equation (13)
The feature of this function is that if the waiting time is 20 seconds or less, it is considered that the influence on waiting psychology is small (passage is allowed because the assigned machine arrives immediately) and no adjustment is made. It is in. As a result, the restriction on the passage restriction is relaxed at a short waiting time, so that the specific gravity of the service to other long waiting calls is relatively increased, and the waiting time performance for the entire call is improved. The waiting time adjustment function F () in FIG. 23 is expressed as in the equations (14), (15), and (16) (α = 1 in the figure).

wt _n = wt _n-1 (0 ≦ wt _n-1 <20) Equation (14)
wt _n = wt _n-1 + (1/2) · wt _n-1 -10 (20 ≦ wt _n-1 ) Equation (15)
wt _n = wt _n−1 + wt _n−1 −40 (60 ≦ wt _n−1 ) Equation (16)
In addition to the characteristics shown in FIG. 22, the increase gain of the adjustment is increased when the waiting time is longer (60 seconds or more), and it works to strongly suppress the passage of long waiting calls. . The waiting time adjustment function F () in FIG. 24 is expressed as in Expression (17).

wt _n = (wt _n-1 ) ^1.1 formula (17)
As shown in FIG. 24, the function is characterized in that the adjustment gain increases as the waiting time becomes longer, and is due to the function of the power function. Also in this case, the longer the waiting call is, the more effectively the occurrence of passage is suppressed.

  The operation of the positional relationship coefficient α in the waiting time adjustment represented by Expression (6) will be described. Here, Equation (6) is expanded as Equation (18) so that the function of α can be understood.

wt _n = F (wt _n−1 , α) = wt _n−1 + α · P (wt _n−1 ) Equation (18)
α is the reciprocal of the distance index d as shown in Equation (5). Therefore, the longer the distance index (the farther the positional relationship is), the smaller α is. As a result, since waiting time adjustment is performed by multiplying α as shown in Expression (18), the longer the positional relationship is, the smaller the waiting time after adjustment becomes, and the influence of passage is evaluated less. If the distance between the assigned car and the car is far away and the waiting passenger does not recognize the passing, relax the restriction on the passing (proactively allow the passing), and service to other long-waiting landing calls The goal is to prioritize the call, and it is possible to improve the waiting time performance of the entire call without causing psychological influence (stress) due to passage. For example, in the example of FIG. 15B, the assigned car is No. 2 and the passing cars are Nos. 1, 3, and 4 (from the passing prediction data table of FIG. 14B). Here, the arrangement of the elevator is as shown in FIG. 18 (a). Since the first and third units are adjacent to the second unit, the passage relationship is easily recognized and the positional relationship coefficient α is 1. Since Unit 4 is far away, the passage is difficult to recognize and α is 0.5. In the waiting time adjustment by the equation (18), the waiting time is adjusted by multiplying by α, so that the waiting time adjustment of the No. 4 machine in which passage is difficult to be recognized is calculated to be shorter than the others. Therefore, in this case, adjustment is made so that the passage of the fourth car is allowed. Thus, by considering the positional relationship between the assigned car and the passing car, it is possible to select the assigned car so as to suppress the psychological influence due to the passing while considering the entire waiting time. Note that such waiting time adjustment based on the positional relationship coefficient is particularly effective in the case of an immediate reservation method in which the assigned car is guided to passengers in advance.

  FIG. 17 shows an example of the result of waiting time adjustment according to the passing situation (passing frequency, positional relationship, waiting time), the car and calling situation is based on the situation of FIG. 7, and the elevator arrangement is based on the situation of FIG. . As the waiting time adjustment function, the function of Expression (19) based on Expression (18) and Expression (11) is applied.

wt _n = F (wt _n−1 , α) = wt _n−1 + α · (1/2) · wt _n−1 formula (19)
First, a method of deriving the result when the first machine in FIG. 17A is temporarily allocated will be described. Focusing on the 7th floor UP direction call that is predicted to pass, the original predicted waiting time is 10 seconds (see FIG. 16 (a)), and the number of passes is 1 time (see FIG. 14 (a)). Since the positional relationship coefficient α is 0.33 (see FIG. 15A), the wait time after adjustment is wt ₁ = 12 seconds from the equation (19). Similarly, when the second unit in FIG. 17 (b) is temporarily allocated, the original predicted waiting time is 36 seconds (FIG. 16 (b)) and the number of passes is 3 times for the call in the 7th floor UP direction that receives the pass. (Fig. 14
(B)), the positional relationship coefficient α is 1,1,0.5 (FIG. 15B), so the waiting time after adjustment is wt ₃ = 101 seconds. At the time of temporary allocation of Unit 3 in FIG. 17 (c), the adjusted waiting time at the time of temporary allocation of Unit 4 in FIG. 17 (d) is obtained in the same manner and is shown in the result.

  If the allocation evaluation value for allocating an optimal car for a new call is calculated as the sum of waiting times of all calls (or all calls affected by the allocation) as described in FIG. The evaluation values before adjustment are as shown in FIGS. 16A to 16D for each temporary assignment, and the evaluation values after adjustment are as shown in FIGS. 17A to 17D. Specifically, the evaluation values before adjustment are 28 seconds (Unit 1), 54 seconds (Unit 2), 28 seconds (Unit 3), and 32 seconds (Unit 4) for each temporarily assigned unit. After adjustment, the time is 30 seconds (No. 1), 119 seconds (No. 2), 33 seconds (No. 3), and 53 seconds (No. 4). As a result, in the evaluation value before adjustment (evaluation based only on the prediction waiting time), what was assigned to either No. 1 or No. 3 is determined to be assigned to No. 1 after adjustment. The reason for the difference is that, in the latter case, focusing on the difference in passing situation, when the unit 1 was assigned, the remote unit 4 passed, and when the unit 3 was assigned, the adjacent unit 4 passed. Yes, effective allocation is implemented in that the allocation to Unit 1 can reduce the psychological impact on waiting customers.

26 to 28 show examples of the waiting time evaluation function. This waiting time evaluation function is a function that converts the adjusted predicted waiting time (or the predicted waiting time if there is no passage) into a waiting time evaluation value, and is used for the purpose of increasing the evaluation value as the waiting time as the waiting time increases. . FIG. 26 shows the case where the predicted waiting time value is used as it is as the waiting time evaluation value (this is used in the examples shown so far), and FIG. 27 shows a long waiting time by combining a plurality of linear functions (primary functions). When the evaluation value is increased as much as possible, FIG. 28 shows a case where the evaluation value is strongly increased as the waiting time is increased by a nonlinear function (square function in the example in the figure). In the method of the present invention, the influence of passage is adjusted as the length of the waiting time by the waiting time adjusting function according to the situation. Therefore, by appropriately using the evaluation functions shown in FIGS. 26, 27, and 28, it is possible to more accurately evaluate the influence of passage. For example, when it is desired to strongly evaluate the occurrence of passing or long waiting, the evaluation functions of FIGS. 27 and 28 may be used. It is also effective to change this evaluation function according to the traffic situation of the building at that time.

  FIG. 12 shows an example of handling of the present invention for a non-directional car. For a non-directional car, when a new call is temporarily assigned to the car, it becomes a directional car. In this case, it is handled as a directional car. On the other hand, as described in process ST306 of FIG. 4, when searching for another car passing through the assigned car, the car is excluded from the search target. As already mentioned, the reason is that the non-directional car is more likely to be waiting at the current position than moving to either one at that time. FIG. 12A shows the state of four cars, a landing call, and a car call. Among these, No. 1 machine D1 corresponds to a non-directional car. FIG. 12B shows an example of an estimated arrival time table for the first car which is a non-directional car. The vertical direction represents the floor, and the horizontal direction represents the UP and DN directions. The numerical values in the table represent the estimated arrival time for that floor and direction. Since there is no direction, it is possible to move in both the up and down (UP / DN) directions, resulting in a table of estimated arrival times as shown in FIG. Normally, the estimated arrival time table for a non-directional car is as shown in FIG. 12B. However, in the present invention, when searching for another car passing through the assigned car, the other car is a non-directional car. In this case, an arrival prediction time table as shown in FIG. This is such a predicted arrival time table because it is assumed that the non-directional car continues to wait on the floor that is currently stopped. As in process ST306 in FIG. 4, by removing the non-directional car from the target of the pass search, it becomes possible to predict the pass according to the actual situation, and the accuracy of the allocation evaluation is improved. Can be implemented in a balanced manner.

FIG. 29 shows the temporal relationship between the predicted waiting time and the passing time of another car. FIG. 29 shows a time axis, where the starting point (leftmost point) of the time axis represents the registration time of the landing call, and the time when a certain amount of time has passed (point of L4) represents the predicted time when another car passes. ing. The time length (L2) from the landing call registration time to the time when another car passes is the predicted waiting time at the time of passage. As time further elapses, an estimated time at which the assigned car arrives is reached. The time length (L3) from the time when another car passes to the time when the assigned car arrives is the estimated waiting time at the time of passing. Further, the length of time (L1) from when the landing call is registered until the assigned car arrives becomes the predicted waiting time. In the description so far of the present invention, the waiting time is adjusted for the predicted waiting time (the length of time L1 in FIG. 29). However, the predicted waiting time until the time when another car passes (see FIG. 29). (L2 time length) or a predicted waiting time (time length L3 in FIG. 29) from when another car passes until the assigned car arrives. In the former case, the psychological impact when another car passes while the waiting customer is waiting for a while will be subject to evaluation, and in the latter case, after the other car passes, The psychological impact on the waiting time can be evaluated. It is possible to evaluate the psychological effects of waiting customers more finely by evaluating each of them separately and adding them together.

  In the following, the diagnosis of the passage result shown in FIG. 1 (the operation of the passage result diagnosis unit 121 in FIG. 1) will be described in detail with reference to FIGS. Diagnosis of passing results means that when the passing of another car actually occurs, for example, the number of occurrences is recorded, and if the number is large, there is a problem (the situation is likely to be frustrated by elevator users) This is to notify the elevator maintenance company, etc. A feature of the diagnosis of the passage result according to the present invention is that the passage occurrence and the waiting time data of the call in which the passage has occurred are evaluated as a set. It is important to note that the passage is a problem when the other cars pass and the elevator car does not arrive at the landing.

  FIG. 30 shows the specific evaluation method. In the graph of FIG. 30, the horizontal axis M1 represents the actual waiting time of the landing call that has passed through (the duration of the landing call), and the number of occurrences as indicated by the vertical axis M2. Accordingly, each vertical bar in the graph represents the waiting time and the number of occurrences (per week in the figure) for calls that have passed through. For example, the vertical bar M3 indicates how many times a call that has passed and whose waiting time was 20 seconds occurred in one week. On the other hand, a curve M4 indicated by a dotted line is a determination curve for determining whether or not to notify the maintenance company that there is a problem. This determination curve M4 has a characteristic that the allowable number of occurrences is reduced even when the waiting time is short, and the allowable number of occurrences is lowered as the waiting time becomes long. That is, the passage of a short waiting time is allowed, but the passage of a long waiting time is not allowed. This is a characteristic that is determined according to the psychology of the waiting customer. In the case of FIG. 30, since the number of occurrences (M5) when the waiting time is 70 seconds exceeds the threshold curve, the group management device indicates that there are many passages with a long waiting time and there is a possibility that the user is dissatisfied The side will judge and notify the maintenance company. In this way, by determining not only the number of occurrences of passing but also the waiting time of calls that have passed, it is possible to make a more accurate diagnosis according to the user's psychology, and the group management side determines Thus, prompt improvement is possible. For example, even if the total number of passages is small, if the waiting time at that time is long, the user is likely to be dissatisfied with the situation, and this situation is accurately judged by the judgment method of FIG. It becomes possible.

  FIG. 31 shows a flowchart for the passing result diagnostic method of FIG. First, it is checked whether a car other than the assigned car has passed first in response to the landing call (ST501). If there is a car that has passed first, the landing call continuation time (corresponding to the actual waiting time of the call) when the assigned car arrives is recorded for the passed hall call (ST502). It is checked whether or not the passing occurrence frequency distribution is counted (ST503). For example, this is done every other week. In the case of counting, a histogram of the number of times of passage with respect to the landing call duration time (corresponding to the graph of FIG. 30) is created (ST504). It is checked whether or not there is a location that exceeds the determination threshold (each value of the determination threshold curve in FIG. 30) with respect to the result of the histogram (ST505). If not, the elevator maintenance management company is notified that there is no problem with the passage occurrence (ST506). If there is a part that exceeds the threshold, the elevator maintenance management company is notified that the passing diagnosis result exceeds the judgment threshold (ST507). Further, when the waiting time adjustment function is automatically adjusted on the group management side, the waiting time adjustment function is adjusted so as to increase the inclination of the waiting time adjustment function, that is, in the direction of suppressing passage (ST508). This adjustment is executed by the adjustment function changing unit 120 in FIG. In this way, by evaluating the combination of the number of occurrences of passage and the waiting time at that time, it is possible to make a more accurate diagnosis according to the user's psychology, and to automatically adjust the adjustment function based on the diagnosis result Thus, it is possible to improve to quick passage suppression (suppression of a case where the vehicle passes even though it waits for a long time).

In FIG. 32, the influence of the predicted passage occurrence is corrected to the predicted waiting time by the waiting time adjustment function, and when the adjusted predicted waiting time exceeds the determination condition (determination threshold), the allocation is reviewed. The control structure of the Example which aims at is represented. The aim is not only to wait for a long time, but also to review the allocation car sequentially, including the effects of passing (from the time of allocation to the time of arrival). It will try to be balanced. By using the waiting time adjustment function, the number of passages, the waiting time, and the passing situation of the positional relationship can be reflected in the adjusted waiting time value, which is possible. Hereinafter, FIG. 32 will be described. 32, the elements different from those in FIG. 1 are the landing call determination unit 140 for the allocation review based on the passing situation and the predicted waiting time and the waiting time evaluation value calculation unit 141 for the allocation car before the review (for the same elements as in FIG. 1). The same sign is attached). The allocation review based on the passage status and the predicted waiting time is performed by using the waiting time adjustment function to adjust the influence of the occurrence of passage to the length of the waiting time. In this case, however, the predicted waiting time is used instead of the actual waiting time. Regularly (for example, every second), search for passing by a car other than the assigned car for each landing call, and if the occurrence of passing is predicted, adjust the predicted waiting time with the wait time adjustment function. Thus, the determination is made based on the allocation review criteria (predicted waiting time value criteria). As a result of this process, the situation changes after the assignment, and a review of the assignment is performed for calls that are expected to be passed multiple times or for calls that have been waiting for a long time. A more appropriate allocation with less can be performed. In addition, the waiting time evaluation value calculation unit 141 for the assigned car before the review compares the waiting time evaluation value after the review (adjustment by passing as an evaluation value) with the evaluation value before the review, and the former evaluation value Only when it is good, the actual review is carried out. This makes it possible to avoid unnecessary review.

  FIG. 33 and FIG. 34 show an example in which the user is notified when the traffic demand is congested and the control for relaxing the passage relaxation condition (allowing passage more actively than usual) is performed. Yes. Since there are many elevator users at the time of congestion, the waiting time of the entire call cannot be taken into account if the passage is suppressed, and long waiting may occur frequently. In order to avoid this, the aim is to notify the user in advance that "passage will occur for the entire service during congestion" and to control the waiting time of the entire call rather than suppressing the passage. It becomes.

33 differs from FIG. 1 in that a traffic flow determination unit 150, a passage relaxation determination unit 151 corresponding to the traffic flow, an information guide device (for example, a liquid crystal display) 53 installed at a landing on each floor,
54. The traffic flow determination unit 150 recognizes the traffic flow of the time zone or the predicted traffic flow of the previous time and determines whether the traffic flow is congested. The determination result of the traffic flow or the result of the detected traffic flow type is sent to the passage relaxation determination unit 151. The passage relaxation determination unit 151 relaxes the action of the passage suppression control depending on whether or not the traffic is in a congested state or the congestion state (the traffic volume such as the number of users). More specifically, the action of the waiting time adjustment function (equation (6) or equation (18)) that reflects the influence on the length of the waiting time depending on the passing situation is relaxed. For example, in the case of Expression (18), the value of α and the gain value (slope) of the function of P (wt) are reduced according to the congestion situation. In addition, at the same time as reducing the action of the passage suppression control, or in advance, a message “currently congested and passage occurs depending on the situation” is sent to the information guide devices 53 and 54 at the landings. Notify for. Based on these measures, we understand that it is inevitable that the waiting customer will pass through depending on the congestion situation, and relax the passage suppression control to give priority to the waiting time of the entire call. It is possible to take control. As a result, it is possible to reduce the waiting time of the entire call, especially the occurrence of a long waiting time, while alleviating dissatisfaction with the elevator users due to the occurrence of passage during congestion demand.

FIG. 34 shows a flowchart for the control of FIG. This processing is performed in the traffic flow determination unit 150 and the passage relaxation determination unit 151 corresponding to the traffic flow in FIG. First, it is determined whether or not the number of passengers per unit time of the elevator is greater than a predetermined value to determine whether it is a crowded demand (ST601). Similarly, it is determined whether the current time is within the recognition range in a busy time zone (learned from past data) (ST602). When it is determined that the time is congested (ST603), the information guidance device of each landing is displayed and guided to indicate that it is congested (ST604). Then, the slope (gain) of the waiting time adjustment function is lowered (this corresponds to reducing the slope of α and P (wt) in the equation (18)) to approach the straight line characteristic of the slope 1 (ST
605). This means that it is close to the state without adjustment. In this way, the control is continued in a state where the adjustment due to the passage is weakened or eliminated, and this control is continued until the congestion condition is removed (ST606). By such processing, it is possible to reduce the waiting time of the entire call, especially the occurrence of long waiting time, while alleviating the dissatisfaction with the elevator user due to the occurrence of passage during the congestion demand.

  As described above, the elevator group management system according to the present embodiment is used for elevator group management in such a way that the movement of each elevator car can be seen at the landing, such as a see-through type elevator or an elevator equipped with a car position indicator at the landing. On the other hand, according to the situation where the passage occurs, the assignment evaluation is performed so as to suppress the passage while considering the serviceability (mainly waiting time) of the entire hall call, and the balance between the two is more appropriate. An assigned elevator can be selected. As a result, it is possible to achieve an appropriate balance between suppressing passage and reducing waiting time for the entire call.

  Furthermore, according to the present embodiment, paying attention to the combination of whether the allocation is passed or not, the influence of the passing action on the waiting time is adjusted by the length of the waiting time according to the combination. Therefore, if it is difficult for the waiting passengers at the landing to recognize the passage according to the combination of the assigned car and the passing car, the control for mitigating the passing action is performed, and the serviceability for the remaining landing calls is improved. Can be improved. In other words, in the prior art, the control that waits uniformly for the passage is performed, but the control for waiting for the entire landing call is performed by performing the control that waits according to the combination of whether the passage is allotted or not. Serviceability of waiting time can be improved.

It is a figure which shows an example of a structure of the elevator group management system by this invention. It is a figure which shows the flowchart of the allocation car determination process with respect to the new hall call of the elevator group management system by this invention. It is a flowchart figure of the car and platform call table update in a present Example. It is a passing car search flowchart figure with respect to the hall call allocation car in a present Example. It is a flowchart figure of the prediction waiting time adjustment by a prediction waiting time and a passage effect in a present Example. It is a flowchart figure of i and j update and loop completion determination in a present Example. It is a figure which shows the condition of the car before allocation in a present Example, a landing call, and a car call. It is a figure which shows the example of prediction time table AT (i, j, k) in a present Example. It is a figure which shows the example of the hall call continuation time table HT (i, j, k) in a present Example. It is a figure which shows the example of the hall call table HC (i, j) in a present Example. It is a figure which shows the condition of the car at the time of each car temporary allocation in a present Example, a landing call, and a car call. It is a figure which occupies the example of the arrival prediction time table with respect to the non-directional car in a present Example. It is a figure which shows the process of the passage generation | occurrence | production prediction by the comparison of the arrival prediction time in a present Example. It is a figure which shows the example of the passage prediction data table in a present Example. It is a figure which shows the example of evaluation of the positional relationship for every allocation case and passage in a present Example. It is a figure which shows the example of prediction waiting time calculation with respect to each boarding call in a present Example. It is a figure which shows the example of prediction waiting time adjustment by the passage condition (positional relationship, frequency | count) in a present Example. It is a figure which shows the example of the cage positional relationship table with respect to 4 units | sets linear arrangement | positioning in a present Example. It is a figure which shows the example of the cage positional relationship table with respect to 4 units | sets facing arrangement | positioning in this Example (entrance / exit 2 places). It is a figure which shows the example of the cage positional relationship table with respect to 6 units | sets facing arrangement | positioning in a present Example. It is a figure which shows the example 1 of the waiting time adjustment function by the passage effect | action in a present Example. It is a figure which shows the example 2 of the waiting time adjustment function by a passing action in a present Example. It is a figure which shows the example 3 of the waiting time adjustment function by a passing action in a present Example. It is a figure which shows the example 4 of the waiting time adjustment function by a passage effect | action in a present Example. It is a figure which shows the example of a calculation (recursive calculation) at the time of multiple times generation | occurrence | production in a present Example. It is a figure which shows the example 1 of waiting time evaluation functions in a present Example. It is a figure which shows the example 2 of waiting time evaluation functions in a present Example. It is a figure which shows the waiting time evaluation function example 3 in a present Example. It is a figure which shows the relationship between the prediction waiting time in a present Example, and the platform continuation time at the time of passage. It is explanatory drawing of the passage diagnosis determination in a present Example. It is a figure which shows the flowchart of the passage result diagnostic part in a present Example. It is a figure which shows an example of a structure of the elevator group management system in the other Example by this invention. It is a figure which shows an example of a structure of the elevator group management system in the other Example by this invention. It is a figure which shows the flowchart in the elevator group management system of FIG.

Claims (11)

In the elevator group management system that decides the cars to be assigned to the multiple cars that serve multiple floors and landing calls,
Means for predicting the presence or absence of a car passing through the floor where the landing call has occurred when determining the car assigned to the landing call;
An elevator group management system comprising means for determining a car assigned to the newly generated landing call based on the result of the predicting means. In the elevator group management system that decides the cars to be assigned to the multiple cars that serve multiple floors and landing calls,
Means for tentatively assigning each elevator to a landing call, and means for predicting the occurrence of a car passing through the floor where the landing call has occurred, if temporarily assigned;
An elevator comprising: means for calculating an evaluation value for the temporary assignment based on the result of the prediction means; and means for determining a car assigned to the newly generated landing call based on the evaluation value Swarm management system. In a group management system for elevators that determines a car to be allocated in consideration of a predicted waiting time to each landing floor for a plurality of cars that serve a plurality of floors and a landing call,
Means for predicting whether or not a car passing through the floor where the landing call has occurred when determining a car to be assigned to the landing call;
Means for adjusting the prediction waiting time based on a result of the prediction means;
An elevator group management system comprising means for determining a car to be assigned to the newly generated landing call based on the adjusted predicted waiting time. In the elevator group management system that determines the cars to be allocated in consideration of the estimated waiting time for the landing floor for each of the multiple cars serving multiple floors and the landing call,
Means for predicting whether or not a car passing through the floor where the landing call has occurred when determining a car to be assigned to the landing call;
If there is an occurrence of a car passing by the means for predicting, means for increasing and adjusting the prediction waiting time;
An elevator group management system comprising means for determining a car to be assigned to the hall call based on the adjusted predicted waiting time. It has a means for displaying a plurality of cars serving a plurality of floors and the position of the car on the floor, and for a newly registered landing call, the waiting time for the landing floor is estimated. In the elevator group management system that determines the car to be assigned in consideration,
Means for temporarily allocating each elevator to a newly generated landing call; and means for predicting the occurrence of a car passing through the floor where the landing call is generated, if temporarily allocated;
If there is an occurrence of a car passing by the means, means for increasing the predicted waiting time; and
Based on the adjusted predicted waiting time, means for calculating an evaluation value in the case of temporary allocation;
An elevator group management system comprising means for determining a car assigned to the hall call based on the evaluation value. For each newly-registered hall call that has multiple cars that serve multiple floors, calculate the evaluation value when each car is temporarily assigned to the hall call. In the elevator group management system for determining a car assigned to the landing call by comparing the evaluation values of
Means for predicting the waiting time of each landing call for a newly registered landing call to which each car is temporarily assigned and a landing call already assigned;
For each car, means for predicting that other cars will pass first for the temporarily assigned landing call that the car is responsible for;
When predicting the occurrence of passage, means for adjusting the length of the predicted waiting time according to the occurrence and / or the number of passages;
Means for calculating an evaluation value including a predicted waiting time for the adjusted hall call;
An elevator group management system, wherein an assignment is determined to a car having the best evaluation value. For each of a plurality of cars serving a plurality of floors, a means for displaying the position of the car on the floor, and a newly registered landing call, each car was temporarily assigned to the landing call. A group management system for elevators comprising means for calculating an evaluation value for each case and determining a car assigned to the landing call by comparing the evaluation values of each car, and a warning light for notifying the assigned car In
For each car, a means for predicting the waiting time of each landing call for a newly assigned landing call that is assigned to that car and for an unserviced landing call that is already assigned;
A means for predicting that each car will pass first for a landing call assigned to each car;
When predicting the occurrence of passage, according to the relationship of the position at the platform of the car that is temporarily assigned and the car that has been predicted to pass, means for adjusting the predicted waiting time of the landing call,
Means for calculating an evaluation value based on the adjusted predicted waiting time,
An elevator group management system, wherein an assignment is determined to an elevator having the best evaluation value. Considering the number of cars that serve multiple floors, the means for displaying the position of the car on the floor, and the estimated waiting time at each car for newly registered landing calls In the elevator group management system comprising means for determining the assigned car and a warning light for notifying the assigned car,
Means for predicting the presence or absence of a car passing through the floor where the landing call has occurred when determining the car assigned to the landing call;
If there is an occurrence of a car passing by the means, means for increasing the predicted waiting time; and
Means for adjusting the incrementally adjusted predicted waiting time based on a positional relationship of the car assigned to the pass-predicted car and the car at the landing;
An elevator group management system comprising means for determining a car assigned to the landing call based on the adjusted predicted time. In claim 7 or 8,
When the platform of the car is arranged on both sides facing each other, the means for adjusting the predicted waiting time of the platform call is the case where the positional relationship on the same side is compared with the case where the positional relationship at the platform is the facing relationship In addition, the elevator group management system is characterized in that the adjustment amount of the predicted waiting time is increased. In claim 3-9,
An elevator group management system comprising means for determining a review of a car allocation for a hall call based on the adjusted predicted waiting time. In an elevator group management system that manages multiple elevators that service multiple floors,
For each car, means for detecting that other elevators have passed first for an unserviced landing call assigned to that car;
When a passage is detected, measure the landing call duration from the point of call registration to the arrival of the assigned car or the landing call duration from the point of call registration to the point of passage when a passage is detected. Means to
Means for storing the duration;
Means for processing the stored data for each landing call duration;
An elevator group management system, characterized in that the value of the processed data exceeds a predetermined standard.

Images