JPS60137787A - Method of controlling group of elevator - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

[Detailed Description of the Invention] [Technical Field of the Invention] The present invention relates to a group management control method for a plurality of elevators,
In particular, it relates to improvements in the method of calculating predicted arrival times.

[Technical Background of the Invention] In group management control in which a plurality of elevators are centrally managed as a group, it has become common to use a small computer such as a microcomputer as a control device. In the case of such group management control, a comprehensive evaluation formula is calculated by obtaining various information on the elevator car status and landing status, and
Based on this result, the appropriate R car for the target hall call is determined.

In this case, in addition to assigned hall calls (hall calls that have already been assigned) and provisionally assigned hall calls (hall calls that have been provisionally assigned during evaluation calculations), it is predicted that future hall calls derived from these hall calls will occur. It is also necessary to register the car call and calculate the predicted arrival time and predicted waiting time a1. That is, when calculating the arrival time of an elevator, it is necessary to predict and perform glH not only the car calls and hall call assignments that currently exist, but also the car calls that will be derived when responding to the assigned hall call. However, since it can only be predicted probabilistically which floor a car call will be derived from, the arrival time may vary significantly depending on the location of the derived call floor, making accurate prediction impossible.

[Object of the Invention] The object of the present invention is to provide Tir! Efforts are made to average the derived car call predictions probabilistically, reduce prediction errors probabilistically, improve the accuracy of arrival time prediction calculations, and improve services for users such as preventing long car calls and reducing first-come-first-serve car calls. The object of the present invention is to provide a group management control method for elevators that enables the following.

[Summary of the Invention] The present invention generates random numbers by weighting uniform random numbers according to a predetermined derived car call occurrence probability distribution, and this weighting is used to pseudo-register derived car calls based on the random numbers to predict arrival. The operation of calculating the time is repeated multiple times, and the arithmetic average value of the results of these multiple calculations is used as the calculated value of the predicted arrival time for the target floor, and the service is based on the predicted waiting time based on this calculated value. It is characterized by determining the elevator.

[Embodiment of the Invention] FIG. 1 shows an example of the configuration of a group management control elevator system to which an embodiment of the present invention is applied.

In FIG. 1, reference numeral 2 denotes a hall call registration device in which hall calls from each floor landing, that is, the elevator hall are registered. 5A to 5D are individual control devices that control the movement of each elevator car (single unit), and the codes are A to D.
A to D@ machines are distinguished by the subscript . Reference numeral 30 denotes a group management control device, which controls the entire movement of each car based on the hall calls from each floor and information from each car.

Further, the signal line connecting these is illustrated as one signal line even when a plurality of parallel signal lines are used. When a hall call is registered from the hall call registration device 2, the hall condition information table (hereinafter referred to as rHCTJ) 24 (held in a part of memory) in the group management control device 30 corresponds to each hall call. Bit 11 of the data of the corresponding index is set, and the arithmetic unit (including input/output device) 26 consisting of a small computer such as a microcomputer in the group management control unit 30 searches for this. Semi-car call registration circuit 1 by selecting the optimal one from the four (numbers A to D)
3A to 13D are assigned (at this time, the hall lantern, that is, the forecast light is lit). At the same time, bit 10 of the index data corresponding to the hall call to be allocated to the HCT 24 and bits O to 3, which are the allocated bits corresponding to the car number, are set. Then, when the allocated car decelerates in response to the landing, the hall call registration is cleared by the car position signal from the car position selectors 14A to 14D, and the HCT2
Bit 11 of 4 is also reset. Further, when the assigned car decelerates, opens its door, and finishes serving the hall, the hall lantern is turned off by the car position signal from the car position selectors 14A to 14D, and at the same time, the assigned bit, that is, bit 1o of the HCT 24, is also reset. Month 1
Pis 1-0 to 3, which correspond to disasters, are also cleared. 11A~
11D is a car call registration circuit, and each floor of the in-car operation panel (not shown) is a car call registration switch 15A to 1.
When 5D is turned on, a call is registered and the car call status information table K sets the pin 1- (any one of bits 0 to 3) corresponding to the corresponding car number of CT23, and the registered floor is assigned to the car. When the service is completed, the car position signal is obtained from the car position selectors 14A to 14D, the car call registration is erased, and the bit corresponding to the corresponding car in the KCT 23 is also reset. 12'A to 12D are car status buffers, which receive car position signals from selectors 14A to 14D as well as contacts 16A to 16 such as relays that close when a door is opened.
D. Contact 17A of a relay etc. that closes when it becomes a non-directional state.
17D1 Contact points 18A to 18D such as relays that close while driving
Contacts 19A to 19 of relays etc. that close when 1 is in the ascending direction
Contacts 20A to 21, such as relays that close when D is in the downward direction; Contacts 21A to 21, such as relays that close when D1 group control is possible.
A signal such as D is input, and in response to these inputs, information such as car status information is stored in the car status information table (hereinafter referred to as rccTj) 22 (held in memory) in the group management control device 30. Stored.

General flowchart 1- of an embodiment of the present invention applied to a group management control elevator system having such a configuration.
FIG. 2(a) shows a detailed flowchart of the hall call allocation processing routine, which is the main part thereof, and FIG. 2(b) shows a detailed flowchart of the hall call allocation processing routine.

In Figure 3, when the program starts, the RAM
(Random access memory - here, it means read/write memory as usual) and human output buffer, etc., and then CG, T and KCT via repeat start point R8.
Input the car information, and if all the cars are out of the group, no processing is done, and if there is even one car in the group, the process associated with the car status, such as the process when the direction changes or the process of changing the full capacity assignment, etc. Let's do it. First, the hole index i is set to 0, and the word corresponding to that hole index in HCT (
data) and performs hall call assignment processing, the details of which are shown in FIG. 2(b). This hall call assignment process is performed by sequentially incrementing the hall index i by "1" and repeating the same process for each hall. When the processing for all the boarding halls is completed, distributed standby processing, advance command processing, and other processing are executed, and the process returns to the repeat start point R8, and the above processing is repeated.

Next, the hall call allocation process shown in FIG. 2(b) will be explained.

In FIG. 2(b), bins 1 to 11 of data I-ICT (+) corresponding to the hole index i of 1-ICT are checked, and if there is a hall call, HCT(i
), and if it is unallocated, the landing arrival time corresponding to car index j is set to HCT, KCT.
, CCT, etc., calculate the overall evaluation value, select the optimal machine, and allocate and output it.

The method for calculating the above comprehensive evaluation value is the main part of the present invention.

In this embodiment, the predicted arrival time Ti to the landing (hall index 1) is calculated using the following equation.

...(1) However, f (xk): Fixed running time on floor xk n: Predicted number of stops to be made up to landing call floor (1st floor) to be calculated, taO: Expected door opening/closing time (for simplicity, The case of stop and the case of car call stop are the same).

The above predicted arrival time differs depending on the car number to which the hall call is provisionally allocated. In addition, in this example, f (xk) in the above equation (1) is f (xk) = 4+2-x
Let it be k.

Then, the elapsed time T1 since the hall call was registered
' is stored in the RAM of the arithmetic unit 26 in the group management control unit 3o in a format as shown in FIG.

The predicted waiting time t1 can be calculated from the above-mentioned predicted arrival time T+ and the elapsed time Ti = since the hall call was registered, for example, by the following equation.

Predicted waiting time ti - elapsed time 1' - i - ten predicted waiting time T1 (2) Of course, when calculating the predicted waiting time t1 here, the above (
It goes without saying that calculation formulas other than formula 2) can be applied.

The predicted arrival time Ti in the above equation (2) is the predicted arrival time when provisional allocation is made to an unallocated newly generated (newly generated) hall call, so it differs depending on which car number is provisionally allocated. Further, for this reason, the predicted waiting time ti when provisionally allocated to a certain number of cars also differs depending on which number of cars the provisional allocation is made to. Therefore, the predicted waiting time to the landing with hall index i when car No.
x) [l=1.2. -=-,N], the predicted waiting time ti (X)[i
=1. -, N] average value t(x) and its standard difference 5
(X) is determined by the following formula.

(Here, N: the number of registered hall calls.

) ...(4) Here, t(x) indicates the average waiting time when machine X is assigned, and 5(X) obviously means the degree of variation in waiting time. .

Next, from the above-mentioned t(x) and S(x), the evaluation value E(x) when car No.

E (x) = 1 t (x) + 2 s (X)
... (5) K1, K2: constants (Kl, 2 > O) above (
5) Calculate the value E(x) of the equation for each @. That is,
X-A, B, C, ..., the evaluation value E (A
), E(B) provisionally assigned to B@machine, E(C) corresponding to C@machine, and so on are sequentially obtained. Then, the machine with the smallest evaluation value E(x>) is assigned to the newly generated hall call.

The first term in equation (5) above is an increasing function based on the average value of waiting time, and the second term is an increasing function of variation in waiting time. Realize the transformation. In this case, if K1 is set to a large value, the waiting time will be more effectively shortened, and if K2 is set to a large value, the waiting time will be equalized more effectively.

Furthermore, the above equation (5) is an evaluation equation that includes parameters (average, standard deviation, etc.) representing the characteristics of the waiting time distribution, and it can be seen that this allows the allocation process to explicitly control the waiting time distribution.

Figure 4 shows an example of the probability distribution of derivative car calls in this embodiment. In this case, we are considering a probability distribution in which statistical value data based on derivative car calls that actually occurred in the past are normalized to 1. . In the figure, the value of a uniform random number is plotted on the ring, and the floor of occurrence is plotted on the horizontal axis.

In other words, the probability density P (J,
(4) (M"I+1, and usually 1+1>J.

Therefore, when the uniform random number IJ is generated, since ■J corresponds to the probability distribution of the Mth floor, a pseudo derived car call is registered at the ~1st floor. ■J is a uniform random number and takes a value in the range 0≦IJ≦1, and the generation floor M takes a value in the range T+1≦M≦N stochastically.

FIG. 5 is for explaining the arrival time prediction calculation method in this embodiment. Here, a case is illustrated in which the predicted arrival time is calculated a total of four times, from the first to the fourth times. However, it goes without saying that the greater the number of times, the higher the accuracy.

Now, suppose the car is stopped at 1F (1st floor) and 3u (3
When a 9u call (9u hall call) is assigned and 5K (5th floor car call) is already registered,
Calculate the predicted arrival time until arrival at the 9th floor ascending hall (until arriving at the 9th floor hall in our direction) when a hall call in the ascending direction occurs.

The first time was when the derived car call was predicted and pseudo-registered using the method described above (shown in the explanation of Figure 4), 8K (
If a pseudo-derived call (8th floor car call) is registered as a call, the predicted arrival time for the 9th floor ascending platform will be the same as the above (1)
) formula, =8+8+10+6+3x5 -47 (seconds). Next, since the second time is predicted to be derived from 5, = 8 +8-4-12 +2 x 5 = 38 (seconds), and the third time is predicted to be derived from 10K (10th floor car call). Then, it becomes =8+8+12+2x5 -38 (seconds), and the 4th call is predicted to be 4K (4th floor car call), so T4=Σf (xk) + (4-1) td.

k-! =8+8+1+12+3X5 -47 (seconds) becomes.

Therefore, the arithmetic mean T of T1 to T4 is -LT=(□TI)/L-(TI+T2+T3+T4)/4=42.
5 (seconds), and the predicted arrival time for the 9th floor ascending platform is 42 seconds.
．． It will be 5 seconds.

It takes 3 minutes after the 9u call occurs to obtain the above calculation result.
Assuming that 0 seconds have elapsed, the predicted waiting time TW for the 9u call is calculated as Tw-30+T=72.5 (seconds).

In this way, the predicted arrival time for any target landing can be set with high accuracy, and as a result, the waiting time can be shortened, and the average value of the waiting time can be controlled with high precision. This is extremely effective in improving service in group control elevators, such as long service life and prevention of first-come, first-served car calls.

Further, if the information on the predicted arrival time obtained in this way is used for displaying waiting time, arrival time, etc., highly accurate display can be realized.

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.

[Effects of the Invention] According to the present invention, probabilistic derived car call predictions are averaged, prediction errors are probabilistically reduced, arrival time prediction calculation accuracy is improved, car calls are prevented from running out long, car calls are It is possible to provide a group management control method for elevators that makes it possible to improve services for users, such as reducing first-arrival numbers.

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing a schematic configuration of an elevator control system to which an embodiment of the present invention is applied, and FIGS. 2(a) and (b) are flowcharts 1 for explaining the embodiment.
・,Figure 3 is a diagram showing an example of the format of a table used in the same embodiment, Figure 4 is a diagram showing an example of the probability distribution of derived car calls applied to the same embodiment, and Figure 5 is a diagram showing an example of the probability distribution of derived car calls used in the same embodiment. FIG. 3 is a schematic diagram for explaining example processing. 2...Hall call registration capacity, 5A-5D...Elevator individual control machine, 11A-IID...Car call registration circuit, 12A-12D...Car status buffer, 13A
~13D... Semi-car call registration circuit, 14A~14D.
...Selector, 15A-15D...Car call registration button (switch), 16A-16D...Contact that closes when opening/closing the door, 17A-17D...Contact that closes when there is no direction, 18A-18D...When driving Close contact, 19A~
19D... Contact that closes in the upward direction, 20A to 20D.
...Contacts that close when in the descending direction, 21A to 21D...Contacts that close when group management control is possible, 22...Cage status information table (OCT), 23...Cage call status information table (KCT), 24...・Hall call status table (H
CT), 26... Arithmetic device, 30... Group management control device. Applicant's representative Patent attorney Takehiko Suzue Figure 1 2 Figure 2 (b) Figure 3 Figure 4 Figure 5

Claims (1)

[Claims] When multiple elevators are put into service on multiple floors and each elevator is provisionally assigned to a new hall call, the predicted waiting time based on the predicted arrival time for the hall call floor is considered. In the elevator group management control method in which the optimal elevator is determined using a comprehensive evaluation formula and assigned to the hall call, the occurrence probability distribution is determined in accordance with the predetermined occurrence probability distribution of derived car calls. Register the generated pseudo-derived car call, calculate the predicted arrival time for the target floor, repeat this calculation multiple times, calculate the arithmetic mean of each calculated value, and calculate the predicted arrival time for the target floor. A group management control method for an elevator, characterized in that the predicted waiting time is obtained.