JPS6411547B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Technical Field of the Invention] The present invention relates to an elevator group management control method that selects and determines a service machine to respond to a generated hall call.

[Technical background of the invention]

In recent years, when multiple elevators are installed in parallel,
In order to improve elevator operation efficiency and improve services to elevator users, a small computer such as a microcomputer is used to rationally and quickly allocate answering machines for hall calls from the halls of each floor. is being carried out. In other words, when a hall call occurs, the system selects the elevator most suitable for handling that hall call, quickly assigns the elevator to respond to that hall call, and prevents other elevators from responding to that hall call. . Conventionally, it has been thought that the best method for selecting an assigned machine is to predict which elevator will arrive first at the floor where the hall call has occurred, and then allocate the hall call to that elevator. Therefore, various methods have been considered for predicting which elevator will arrive first. For example, prediction is performed by calculating the predicted time until the elevator arrives at each floor.

However, in the above-described method, since emphasis is placed on servicing newly generated hall calls, when considering the service for all hall calls, an inconvenient phenomenon occurs, especially during times of congestion.

For example, if hall calls that occur one after another are simply assigned to the elevator that can handle them first, the response to hall calls that have already been assigned will be delayed, resulting in hall calls with very long waiting times. become. Even if it is determined that this phenomenon can be dealt with early when the hall call is assigned, the operating status of the elevator may change as other hall calls are subsequently assigned or a new car call occurs from within the elevator. It occurs because it changes. Therefore, when considering the waiting times for all hall calls, significant non-uniformity occurs.
In particular, there is a high probability that long-waiting calls with extremely long waiting times will occur, leading to a decline in elevator service.

[Purpose of the invention]

The object of the present invention is to provide an extremely sophisticated elevator group management control method that takes hall call response times at all floors into consideration in response to actual traffic data.

[Summary of the invention]

The present invention calculates the statistical value of the response time of hall calls within the scheduled time (for example, the response rate within 30 seconds), and weights it with a weighting function that changes the minimum point (axis position of the function) according to that value. The feature is that the evaluation values are used to perform evaluation calculations for selecting the optimal machine.

[Embodiments of the invention]

FIG. 1 shows the configuration of a system to which an embodiment of the present invention is applied.

In Figure 1, 1 is a hall call registration circuit;
When a hall call is registered, the register for the corresponding stall and direction is set, and is reset when the car arrives at the floor corresponding to that hall call. 2A to 2C are elevator operation control devices provided for each of the three elevators A to C, including car status buffers 3A to 3C, car call registration circuits 4A to 4C, and semi-car call registration circuit 5A.
5C and signal synthesis circuits 6A to 6C are provided separately. Car status buffers 3A to 3C are buffers for inputting the car status to a wiper select circuit 7, which will be described later. Car call registration circuit 4
A to 4C are set at the time of car call registration, and are reset when the car arrives at the call registration floor. The semi-car call registration circuits 5A to 5C are
It stores the hall call assigned to that car and is reset when the car arrives at the floor corresponding to that hall call. Signal synthesis circuit 6A
6C outputs the logical sum of the outputs of the car call registration circuits 4A to 4C and the outputs of the quasi-car call registration circuits 5A to 5C.

7 is a wiper selection circuit, and 8 is a decoding circuit, which decodes an output signal from an output register 12, which will be described later, and sets secondary car call registration circuits 5A to 5C for the corresponding floor direction of the corresponding car. 9 is a small computer using, for example, a 12-bit microcomputer, and output register 1
0, an input register 11, and an output register 12. The output register 10 has a function of holding the same output until the next output is issued.

Note that the registers and interface devices having the same function, one for each elevator, are connected by a plurality of parallel signal lines, for example 12. All registers have a number of bits corresponding to one word of the small computer 9.

Next, an embodiment of the present invention will be described with reference to flowcharts shown in FIGS. 2 to 6. 2, 4, and 5 are the overall flowcharts of this embodiment, FIG. 3 is a detailed flowchart of the responding machine determination routine, and FIG. 6 is a detailed flowchart of the weighting function table creation routine. It is a flowchart.

First, starting from terminal P1 in FIG. 2, the writable memory in the computer is initialized in process P2, and the status of each car (direction, position, door status, etc.) is read for all cars in process P3. Then, in process P4, the hole index I is set to 0.

Next, in process P5, the state of the hall (new hall call generated, response to hall call completed, hall call generated but service not completed, no hall call) is determined by the method described below.

That is, when a hall call is registered in the hall call registration circuit 1 shown in FIG. 1, the corresponding bit in the table storing the hall call status shown in FIG. 5 becomes "1", and when there is no hall call, it becomes "0". ” becomes. Therefore, when the corresponding bit changes from "0" to "1", it means that a new call has occurred, and the process advances to process P6. Also, the corresponding bit is “1”
When it changes from to "0", it means that the response to the hall call has been completed, and after storing the unanswered time T ₁ of the hall call in process P7, T ₁ =
0 and moves on to process P9. The corresponding bit is “1” →
If it is "1", it means that there is a hall call but the service has not been completed, so in step P10, the unresponse time _T1 for the hall call is incremented by "1" and the process goes to step P9. If the corresponding bit changes from "0" to "0", there is no hall call and there is no change, so T ₁ is set to 0 in process P8 and the process moves to process P9. In the process P6, the answering machine for the newly generated hall call is determined by the method shown in the flowchart of FIG. 3. First, the allocation of the hall call on the i floor will be explained using a mathematical formula.

In other words, the predicted arrival time (predicted waiting time) T _j i of car j to the i-floor hall is determined by the time required for car j to travel from its current position to the i-floor hall, and the time required for car j to travel from its current position to the i-floor hall, as well as the time required to stop on the way to the i-floor hall. It is calculated as the sum of the loss time (mainly acceleration/deceleration time, door opening/closing time, and opening time) spent on the vehicle. Next, when the hall call on the i floor is assigned to the car j, the predicted waiting time of the already assigned hall call of the car j is T _j k ₁ ,...,T _j k _o (k _o = number of already assigned hall calls) is as follows. It is determined by the formula. However, i
The predicted waiting time changes (delays) only for calls that stop after the floor (assigned hall calls).

T _j kl = (time elapsed since the hall call for the kl floor) + (estimated waiting time until the car arrives at the kl floor) + (time required for car j to stop at the i floor)... (1) However, the KL floor is a floor that stops after the i floor,
For floors that stop first, the time required for the car to stop at the i floor is not required in the above equation.

Next, the predicted waiting time T _j kl obtained by equation (1) is set appropriately in advance using a weighting function with a single minimum value (any function with one minimum value is effective). However, here we use a quadratic function as an example.). (Of course, this weighting function may be used in the form of a broken line approximation.)
In other words, in Figure 9, which will be described later, the horizontal axis shows the predicted waiting time.
As shown by taking T _j and the weighted evaluation value (T _j ) on the vertical axis, the corresponding evaluation value (T _j kl) is obtained from the predicted waiting time T _j kl. Therefore, the evaluation value of the newly generated hall call and the previously allocated hall call for each machine is determined by the following formula.

E _j =1/kn+1 { _ko 〓 ^l=1 (T _j l) + (T _j i)} ...(2) After calculating the average evaluation value of all three aircraft using this formula (2), the minimum The average evaluation value E _MIN is calculated using equation (3).

E _MIN = min (E _A , E _B , E _C ) ...(3) The machine corresponding to E _MIN in equation (3) becomes the service machine for the hall on the i floor and displays the forecast in the hall.

Note that FIG. 9 is a characteristic diagram showing the relationship between the predicted waiting time and the weighted evaluation value when calculating the evaluation value for each call.

This characteristic diagram shows the state function set in the P-POM (Programmable Read Only Memory) table, and shows the predicted waiting time that takes the minimum value.
The value of T _j , that is, the minimum value point is initially set to, for example, 10 seconds. Furthermore, even when the minimum point is not 10 seconds, as shown in FIG. 9, the curves with the same shape simply move horizontally in the direction of the time axis. Furthermore, in the function of FIG. 9, when the predicted waiting time T _j is smaller than the local minimum point, the evaluation value (T _j ) becomes large and becomes difficult to assign. Furthermore, the average non-response time (minimum value point) when the predicted waiting time T _j takes a minimum value
Even if it becomes larger than , the evaluation value (T _j ) becomes large and becomes difficult to assign.

The above-mentioned processing will be explained with reference to the flowchart of FIG.

First, in process Q1, the predicted waiting time T _j of the hall call new generation floor and the allocated hall call command floor is calculated for car A using equation (1), and the process proceeds to process Q2.

In process Q2, the predicted waiting time calculated in this way
An evaluation value is obtained by weighting T _j based on the characteristics shown in FIG. Next, in process Q3, the average evaluation value E _j is calculated. When the A machine is completed, it is checked in process Q4 whether the above processing has been completed for all machines, and if it is not completed for all machines, the process returns to process Q1 again, and the calculations are made for B and C machines in the same way, and all the processes are completed. When the process is completed for each machine, the minimum value of the average evaluation value E _j is determined in process Q5, and the elevator corresponding to the minimum value is selected as a service machine in process Q6, and a forecast is displayed in the hall.

It is determined in process P9 of FIG. 2 whether the above operations have been performed for all floors, and if the process has not been completed, the process returns to process P4, and if it has been completed, the process proceeds to connector B.

FIG. 4 shows the processing after connector B. First, in process S1, it is determined whether a certain period of time has elapsed or not, and if it has not elapsed, the process returns to process P3 in FIG.
The statistical value of the non-response time is calculated in step S3, the minimum value point is updated in step S3, and the process passes through connector D.
After re-creating the function conversion table, return to connector A.

FIG. 5 shows details of the non-response time statistical value calculation process S2 in FIG. 4 and the minimum value point update process S3 using the statistical value.

That is, in FIG. 5, in process T1, the number N of newly generated hall calls up to the elapse of a certain period of time is stored, and in process T2, among the hall calls that have been answered within the same time, the unanswered time _T1 is within 30 seconds, for example. The number M of objects is determined, and in process T3, the response rate (percentage) L (=M/N×100) within 30 seconds is determined, and a preset constant value L _u (usually 70% to 75%) is determined.
In process T4, if the response rate L exceeds L _u , it is considered that the service level has improved, so in process T5, a constant value L ₀ is added to the minimum point B, and (as L ₀ is preset to a value in the range of 5 to 10, for example), and as a result, the service level is lowered and the process proceeds to connector D. In addition, if the response rate L is less than or equal to L _u , it is compared with the allowable lower limit value L _D in processing T6 via the connector C1, and if it is larger than L _D , it is compared with the minimum value point B.
Returns to connector A without updating LD, and if it is smaller than _LD , the service level has deteriorated, so process T7
Then, subtract L ₁ (L ₁ is set to a value in the range of 5 to 10, similar to L ₀ ) from the minimum value B (if the result of subtraction is negative, set it to 0), and proceed to connector D.

Then, as the set value of the non-response time that becomes the minimum value is updated, the function conversion table is re-created in step S4 shown in FIG. 4 as described above, and the details of this process are shown in FIG. 6.

That is, in process R1 of FIG. 6, the non-response time that takes the minimum value calculated every time a certain period of time has elapsed, that is, the minimum value point B, is stored in D, E=0 is set in process R2, and the minimum value is set in process R3. The P-ROM table (TBL (0)
(corresponds to the function value when is the minimum value) and stored in the area corresponding to the non-response time (average non-response time when the value is the minimum value) of the RAM table (TBLA) in process R4. For example, if the average non-response time at the minimum value is 15 seconds, TBLA
Store TBL (0) in (15). and processing R5
Then, B=B+1, E=E+1, and the above-mentioned operation is carried out in process R6 to extract the value corresponding to TBLA from TBL and store it until B reaches 180.

Next, in process R7, after setting E=181 and F=0,
In process R8, D is stored in B. Then, in process R9, B=B-1 is set, and the process proceeds to process R10. Processing R1
0, compare whether B is larger or smaller than 0, and
If ≧0, proceed to process R11 and set E=E+F. Then, in process R12, data is loaded from the P-ROM table, and TBL(E) is stored in C. Then, in process R13, C is stored in TBLA(B),
Proceed to process R14. In process R14, F=F+1 is set, and the process returns to process R9, where B=B-1 is again set.

In this way, the above-described operation is repeated until the minimum value point B (average non-response time when the minimum value is present) is from -1 to B=0.

By the above operations (flowchart in FIG. 6), TBLA is created as a table with unresponse time as an index. The function conversion table created in this way is as shown in FIG.

After creating the table in this way, the connector A
After that, the process returns to process P3 in FIG. 2, and the same process is repeated thereafter.

In this way, when determining a service machine from among multiple elevators, the predicted unanswered time for an already allocated hall call is calculated for each elevator, and the predicted unanswered time is calculated using a weight function having a minimum value point. The elevator with the lowest evaluation value is selected as the service elevator, and the hall call response status is monitored from time to time, and the minimum value point is corrected accordingly. When the level of service becomes extremely poor, control is carried out in accordance with the actual service situation, such as by minimizing the non-response time. In addition, since the system monitors the hall call response status in buildings under control, it is possible to take measures such as suspending free cars when the service level is extremely good.

It should be noted that the present invention is not limited to the embodiments described above and shown in the drawings, but can be implemented with various modifications without changing the gist thereof.

For example, in the above example, the response rate within 30 seconds was used as the statistical value of hall call response time, but the response rate within a short time or a long time may be used, or the response rate exceeding the scheduled time may be used. Alternatively, a value corresponding to the maximum response time may be used.

[Effects of the Invention] According to the present invention, it is possible to provide a sophisticated elevator group management control method that takes hall call response time on all floors into account and provides an extremely high level of service in accordance with actual traffic conditions. .

[Brief explanation of drawings]

FIG. 1 is a block diagram showing the configuration of an elevator system to which an embodiment of the present invention is applied, and FIGS.
FIG. 6 is a flowchart of an embodiment of the present invention, FIGS. 7 and 8 are schematic diagrams of a conversion table for explaining the embodiment, and FIG. 9 is a weighted table used in the embodiment. FIG. 3 is a diagram showing an example of an evaluation function. 1... Hall call registration circuit, 2A to 2C... Elevator operation control device, 3C to 3C... Car status buffer, 4A to 4C... Car call registration circuit, 5A, 5C
...Semi-car call registration circuit, 6A to 6C...Signal synthesis circuit, 7...Wiper selection circuit, 8...Decoding circuit, 9...Small computer.

Claims (1)

[Claims] 1 Targeting multiple elevators installed in parallel,
In an elevator group management control method that selects and responds to a service machine for a hall call that has occurred, the predicted unanswered time for unassigned hall calls and assigned hall calls is calculated for each elevator machine, and a weight function having a minimum value is used. In order to weight the predicted unanswered time and convert it into an evaluation value, and select the elevator with the minimum evaluation value as the service machine, the statistical value of the hall call response time for each scheduled time is determined, and this statistical value is set. When the statistical value exceeds the set value, the minimum value point of the weighting function used for calculating the evaluation value is moved in the direction in which the predicted non-response time increases, and when the statistical value becomes smaller than the set value, the minimum value point is moved to the A group management control method for elevators, characterized in that the elevators are moved in a direction in which predicted non-response time is reduced.