JPH09325995A - Method and device for scheduling movement of carrier - Google Patents

Abstract

PROBLEM TO BE SOLVED: To enable the optimum movement of a carrier while efficiently considering uncertain factors concerning the carrier system of articles. SOLUTION: A probability table describing a value related to the probability for the carrier to respond to every carriage request for each station number of the last carriage destination is provided corresponding to each carrier. Based on the value related to the probability in the probability table of that carrier corresponding to the station number of the last carriage destination provided as a history up to the moment, each carrier determines which carriage request is to be responded to. Based on the determined carriage request, each carrier moves and carries parts, and the response to that carriage request is evaluated from the time of carriage and the number of stock at the carriage destination station. The probability table is updated so as to obtain the high probability to respond to the carriage request as the evaluation is high.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a system in which, in response to a transportation request generated from a plurality of stations, each transportation vehicle is moved to that station and a person or an object is transported, the transportation as a whole system is carried out. The present invention relates to a method and an apparatus for controlling an optimal movement allocation of each carrier in a current system state in order to improve the movement efficiency of a vehicle and the work efficiency of the entire system. INDUSTRIAL APPLICABILITY The present invention proposes a flow schedule of goods for efficiently moving the goods to each station in large-scale physical distribution, and an automatic guided vehicle (AG) for efficiently carrying the goods to a plurality of stations.
V) Transportation schedule planning, Taxi allocation to efficiently allocate taxis to multiple stations that require vehicle allocation, Distribution schedule planning to efficiently allocate elevators to multiple requested floors (stations), goods In order to efficiently deliver to a large number of stations, the present invention can be applied to a method and an apparatus relating to a vehicle dispatch schedule planning for allocating trucks at a physical distribution terminal.

[0002]

2. Description of the Related Art In order to convey goods from one station to another station, a dispatch command is given to each truck waiting at a different place, or a person is transported from one station to another station. Giving each taxi waiting at a different location a dispatch order currently means that the courier will rely on intuition and experience.
In order to enable efficient vehicle allocation, decisions are made in response to multiple vehicle allocation requests at that time.

However, the method based on the intuition and experience of the delivery staff has the following problems. 1. Experience and skill are required for the delivery staff. 2. Since it is necessary to respond to a large number of dispatch requests in a short time, the mental burden on the delivery staff is not small. 3. It is not always possible to create a vehicle allocation plan that reduces the objective function such as transportation cost, and it is not possible to evaluate whether the actual vehicle allocation was an efficient vehicle allocation plan for the entire system.

Further, it has been proposed that the vehicle allocation plan is performed by using an optimization method such as a branch and bound method or a simulated annealing method (SA), but this method has the following problems. . 1. It is necessary to know in advance the traffic conditions (road conditions) such as the time when the job occurs, the time required for processing, the vehicle traveling speed, etc., and it is necessary to assume that these data can be used as constraints. Is. 2. As in the case where a part of the equipment fails, the equipment temporarily fails, the work by workers etc. is delayed, the transfer resources increase or decrease, or the occurrence pattern of the transfer job changes, There is a problem in that heavy load calculations must be executed again each time the constraint conditions change. 3. The calculation cost becomes very high.

[0005]

On the other hand, in the field of artificial intelligence, there is a technical field called multi-agent aiming at distributed processing of information. In this technical field, a paper proposing a method for sequentially selecting production machines to which work is assigned in each process of a production system ("Approach to Dynamic Task Assignment Problem by Stochastic Rule Learning", The Japan Society of Mechanical Engineers) There is a collection of papers C, 58-551 (1992), P270). Here, each machine that performs processing such as machining is provided with a decision making mechanism (hereinafter referred to as an agent) independently, and the machine determines to which machine the job processed by itself is sent next. The configuration of the technology of the above paper is intended for factory machines, and for application to vehicle allocation and transportation methods such as AGVs, trucks and taxis,
The following problems occur.

When applied to the vehicle allocation method, the resources to be processed are AGVs, trucks and taxis. or,
The job is a transport job. In the above-mentioned conventional technology, there is a gap between the job allocation time and the job processing start time. In the case of vehicle dispatch and transportation, this gap is an obstacle. For efficient transportation, it is considered that AGVs, trucks, and taxis have less downtime. Therefore, due to the passage of time, A when the actual allocation was made
The position of GV or the like is different from that at the start of the transport job process. Actually, the time required for carrying out the carrying is different even for the same carrying job depending on the position of AGV or the like at the start of the job. The difference between the time at which the job is assigned and the processing start time of the job is that the job to be executed by one AGV or the like is different from that of another AGV.
This occurs because the configuration will be determined by the above.

The actual object to be conveyed includes information such as the time when the job is generated, the time required for the processing, the traveling speed of the vehicle, and other traffic conditions (road conditions) that cannot be accurately known. Therefore, it is an object of the present invention to efficiently consider an uncertain factor in actual transportation so that the transportation vehicle can be optimally moved.

[0008]

According to the method of claim 1 and the apparatus of claim 5, the calculation load is distributed in order to determine which transport request each transport vehicle responds to. Therefore, high-speed response can be realized. Since which transport request is responded to is determined with a probability corresponding to the state of the system at that time, optimal transport of each transport vehicle can be realized. In addition, as a result of the transport vehicle moving an object or a person in response to the transport request, the transport result is evaluated based on the transport state of the transport vehicle or the state of each station. The higher the evaluation,
Since the probability table is updated so that the probability of response to the transportation request is high, the optimal movement control of the transportation vehicle can be performed even if the system changes.
Further, since it is a learning method, it is possible to optimally deal with elements that cannot be predicted in advance. For example, it is possible to appropriately deal with a change in traffic conditions such as the traveling speed of a carrier during transportation. Further, since each transport vehicle determines the transport destination by itself, it is possible to easily cope with the increase or decrease of the transport vehicles, such as newly introducing a transport vehicle into the system or deleting a certain transport vehicle. In addition, since a probability table is provided corresponding to the state data that represents the history of possible states of the system, roles will be shared among the transport vehicles, and the transport will automatically be divided into several areas for efficient transport. Is done. In the method of claim 2, since the state data of the system that selects the probability table is the response history of the transport vehicle or the history of the transport result of the transport vehicle, it is optimal in relation to the past transport status of each transport vehicle. The next destination can be determined. Further, according to the method of claim 3, since the state data of the system for selecting the probability table is the data showing the state at each station, the optimum next destination according to the state such as the inventory quantity of each station. Can be determined.

In the method of the fourth aspect and the apparatus of the sixth aspect, the probability table of each transport vehicle is stochastically selected, and which transport request is responded to is determined according to the probability table. Therefore, as a result of learning, it is possible to imitate the response pattern to the transport request of the transport vehicle that is moving optimally, and the overall system is more efficient,
The carrier can be moved. In addition, since the probability tables of other transport vehicles are used in this way, even if the transport vehicle moves in an area where learning is not performed, the probability table of the transport vehicle that is learning in that area is calculated. Since it can be used, more efficient movement of the carrier can be realized.

In the method of claim 7 and the apparatus of claim 8, since the data indicating the past destination station of the carrier is used as the state data for determining the probability table, efficient movement or The transport order can be searched. Also, to update the probability table,
As the evaluation of the response to the transport request, the time required for transporting the transport vehicle and the stock status at the transport destination station are adopted, so that each transport vehicle can be transported in a shorter time and the optimal inventory can be secured. Is controlled to move.

[0011]

BEST MODE FOR CARRYING OUT THE INVENTION The present invention will be described below based on Examples. In the present embodiment, as shown in FIG. 1, a large number of mounting bases X1 to X are mounted on the belt conveyor 10 in the factory.
7, the assembling machines M1 to M7 are arranged along one side of the belt conveyor 10 along the belt conveyor 10. The assembling machines M1 to M7 are provided with buffer stations B1 to B7 for temporarily stocking assembled parts.

Further, in the factory, the assembly parts delivered to the factory are provided with the respective guided vehicles AGV1 to AGV5 (automatic guided vehicle "AG
V ”) are provided with warehousing conveyors G1 to G7, and the transporting vehicles AGV1 to AGV5 carry the assembly parts from the warehousing conveyors G1 to G7 and transport them to the buffer stations B1 to B7. . At this time, the assembly machine M
The assembled parts handled by No. 1 are carried in from the warehousing conveyor G1 by the transport vehicle, and the assembled parts handled by the assemble machine M2 are carried in by the courier vehicle from the warehousing conveyor G2. Similarly, the assembled parts handled by the assembling machine M7 are carried in from the warehousing conveyor G7 by the transport vehicle.

In each of the buffer stations B1 to B7,
The number of assembled parts in stock on the buffer station is C
If the number of stocks is less than the predetermined number C, the carry request is turned on. Each of the transport vehicles AGV1 to AGV5 selects one buffer station from the buffer stations for which the transport request is turned on by the method described later, and assembles the corresponding storage conveyor from the corresponding storage conveyors. Transport the attached parts.

On the other hand, the mounting base is moved by the belt conveyor 10, and the respective mounting parts are mounted on the mounting base by the respective mounting machines M1 to M7. At this time, the assembling machine required for assembling the parts on the assembling base is determined by the part number of the base. When the ratio of the product number of the base flowing on the belt conveyor 10 changes, the ON / OFF pattern of the transfer request generated by the buffer station changes, and each transfer vehicle AG
The movement route pattern of V1 to AGV5 also changes.

The movement control device mounted on the transport vehicle AGV1 is composed of a computer system as shown in FIG. That is, the CPU 2 that determines the destination station for a plurality of transport requests and updates the probability table
0, transmitting / receiving device 2 for communicating with the center device 40
1. A RAM 22 that stores various data, a ROM 23 that stores a control program, a buffer station that is a transfer destination that has been determined, and a transfer vehicle is automatically operated in order to transfer components from a determined storage conveyor to the buffer station. And an automatic driving device 24 for operating. Further, the RAM 22 stores a probability data received from the state data history storage area 31 and the center device 40, which stores, as state data, which buffer station the parts have been conveyed to the RAM 22 at a predetermined number of times in the past. And a table input area 32. The other transport vehicles AGV2 to AGV7 have exactly the same configuration.

Further, a center device 40 composed of a computer system is provided in the factory. That is, the center device 40 performs data communication with each of the transport vehicles AGV1 to AGV5 to receive a transport request from each of the buffer stations B1 to B7, a CPU 41 for performing data transmission / reception control, and various types. 43 for storing the data of ROM, ROM 4 for storing the control program
It is composed of four. Further, the RAM 43 stores a probability table corresponding to each transport vehicle in a probability table storage area 5
1 is formed.

Next, the operation of this apparatus will be described with reference to the flow charts of FIGS. FIG. 3 shows the CPU 2 of the carrier vehicle.
0 is a flowchart showing the operation procedure, and FIG. 4 is a flowchart showing the operation procedure of the CPU 41 of the center device 40.

First, in step 200, the CPU 41 of the center device 40 causes each of the buffer stations B1 to B1 to
The transport request transmitted from B7 is input and stored in the transport request storage area 52. Next, at step 202, it is judged if there is a data transmission request from the guided vehicle. If there is no data transmission request from the carrier, step 206
At, it is determined whether or not there is received data from the carrier, and if not, the process returns to step 200. In this way, the center device 40, when not performing data communication with each of the transport devices, has each of the buffer stations B1 to B7.
It means that it is monitoring the transportation request from the.

On the other hand, the CPU 20 of each carrier executes the program shown in FIG. The program shown in FIG. 3 is started after the execution of the initial setting routine after the power of the apparatus is turned on. In step 100, the state data stored in the state data history storage area 31, here,
The center device 40 is requested to transmit the probability table corresponding to the previous transport destination buffer station number and transport request data from each of the buffer stations B1 to B7.
Receiving the request for this data, the CPU 41 of the center device 40 determines in step 202 that there is a data request from the transport vehicle, and in step 204 transmits the probability table and the transport request requested by the transport vehicle. .

Probability table storage area 5 of the center device 40
The probability table stored in No. 1 is, as shown in FIG.
For each number i of the previous transfer destination buffer station, the probability of determining the next transfer destination buffer station (number k) is described. Therefore, based on the number i of the previous transfer destination buffer station transmitted from each transport vehicle, the center device 40 transmits the probability table corresponding to that number i to each transport vehicle.

The CPU 20 of the transport vehicle executes step 102.
Then, the probability table is stored in the probability table input area 32, and the transport destination buffer station (number j) is determined in step 104. The transfer destination buffer station is determined in proportion to the probability of the probability table so that the transfer destination buffer station for which the transfer request is made is selected. For example, a numerical value range is set in proportion to the probability of a transfer destination buffer station (number k) that has a transfer request, random numbers are generated with equal occurrence probabilities, and the next transfer destination buffer is generated according to the numerical value range in which the random number exists. The station (number j) can be determined.

Next, the CPU 20 of the transport vehicle takes out the component from the corresponding receiving conveyor in order to transport the component to the determined transport destination buffer station (number j).
The parts are transferred to the transfer destination buffer station (number j). Next, at step 108, in order to evaluate that the component has been transported in response to the transport request from the buffer station (number j), it is determined whether or not the time required for transport is a predetermined time H or less, and the predetermined time If it is below, in step 110, it is determined whether or not the stock quantity of the transfer destination buffer station (number j) is less than or equal to a predetermined value W. If the stock quantity is less than or equal to the predetermined value W, the probability table is updated in step 112.

The update of this probability table is performed as follows. Last destination buffer station (number i)
To the next destination buffer station (number j)
Let P _ij be the probability that is selected

[0024]

## EQU1 ## When k = j P _ij ← P _ij + a (1-P _ij ) (1)

## EQU2 ## When k ≠ j P _ik ← (1-a) P _ik (2) where a is a constant (0 ≦ a ≦ 1).

In the equation (1), if the response to the transport request from the transport destination buffer station (j) is highly evaluated, then the transport destination buffer station (i) is followed by the transport destination buffer station (i). in order to increase the probability that the j) is selected, means increasing the probability P _ij by an amount proportional to the probability that the buffer station (j) is not selected (1-P _ij). In addition, in the equation (2), the probability P _ik of the buffer station (k, k ≠ j) is calculated in order to reduce the probability that other transport destination buffer stations (numbers k, k ≠ j) are selected. It means to reduce by an amount proportional to P _ik .

Then, the updated probability table is transmitted from the carrier to the center device 40, and the center device 40
In step 206, it is determined that there is received data from the transport vehicle, and in step 208, the updated probability table is stored in the probability table storage area 51. And then,
In step 114, the number j of the transfer destination buffer station that has transferred the component is stored in the state data history storage area 31 as the state data.

In this way, if the result of responding to the transport request from each station is good, the probability of selecting a transport request having a good result with respect to subsequent responses to the transport request is high. In addition, since the probability table is updated, each transport vehicle can determine the next transport destination buffer station so that the state of the entire system will be in a better state with respect to subsequent transport requests. Therefore, each transport vehicle will eventually move according to the optimal transport order by self-learning. Further, when the ratio of the mounting base flowing on the belt conveyor 10 changes, the ON / OFF patterns of the transfer requests output from the buffer stations B1 to B7 change, and the optimum order of the transfer vehicles also changes. As described above, since each transport vehicle learns its own probability table, each transport vehicle can determine the optimal transport order even if there is such an external change. Also new,
When a carrier vehicle is put into the system, the probability table of the carrier vehicle is set to a predetermined initial value or the value of the current probability table of another carrier vehicle as an initial value, and the above probability table learning is advanced. Good. Therefore, it is possible to easily insert or delete the carrier into the system.

In the above embodiment, the center device 40 monitors the transport request from each buffer station, but the transport request may be issued by each transport vehicle. Further, the probability table is provided in the center device 40, and each transport vehicle receives its own probability table from the center device 40 through data communication and transmits the updated probability table to the center device 40. The table may be stored in each transport device. In that case, the processing flow of the CPU of each carrier vehicle and the center device (FIGS. 3 and 4) does not require the processing of requesting or transmitting the probability table. Further, each transport vehicle may monitor transport requests from each buffer station, store the transport requests, and store a probability table. In that case, the center device itself becomes unnecessary.

Next, another embodiment will be described. In the above embodiment, each transport vehicle determines which station responds to the transport request based on its own probability table, but in this embodiment, each transport vehicle includes its own probability table. The probability table of another carrier can be selected, and the selected probability table is used to determine which transport request is to be responded to. Therefore, as shown in FIG. 6, a table selection probability table for selecting a probability table is provided corresponding to each transport vehicle.

The table selection probability table shown in FIG. 6 corresponding to each carrier is stored in the table selection probability table storage area 53 of the RAM 43 of the center device 40, as shown in FIG. Each carrier AGV1 to AGV5
As shown in FIG. 2, the RAM 22 has a table selection probability table input area 33. The center device 40 inputs the latest table selection probability table of its own carrier by data communication, and the table selection probability table is input. It can be stored in the table input area 33.

Next, the operation of this apparatus will be described with reference to the flowcharts of FIGS. FIG. 7 shows the CPU 2 of the carrier vehicle.
8 is a flowchart showing the operation procedure of 0, and FIG. 8 is a flowchart showing the operation procedure of the CPU 41 of the center device 40.

Step 400 is step 200 of FIG.
In response to the above, the transfer request transmitted from each of the buffer stations B1 to B7 is input and stored in the transfer request storage area 52. Steps 402 and 41
When no data communication is performed with respect to each carrier after 0, the carrier requests from the buffer stations B1 to B7 are monitored.

On the other hand, the CPU 20 of each carrier executes the program shown in FIG. In step 300, the center device 40 is requested to generate its own table selection probability table. Upon receiving this data request, the CPU 41 of the center device 40 determines in step 402 that there is a data request from the transport vehicle, and in step 404, the requested transport vehicle from the table selection probability table storage area 53. The table selection probability table of is read and transmitted to the transport vehicle. As shown in FIG. 6, the table selection probability table stored in the table selection probability table storage area 53 of the center device 40 is a table that defines the probability for selecting the probability table for each transport vehicle. For example, if there is a data request from the carrier AGV3,
The table selection probability table of the row of AGV3 is A
Sent to GV3.

In step 302, each transport vehicle stores the table selection probability table received from the center device 40 in the table selection probability table storage area 53. Next, in step 304, a probability table for determining the next destination buffer station is selected using the table selection probability table. This selection method is the same as the method of determining the next transfer destination buffer station using the above-described probability table, and each probability table is selected in proportion to the probability described in the table. is there.

Steps 306 to 320 are the same as steps 100 to 110 in FIG. 3, and the probability table received from the center apparatus 40 in response to the transmission request of the probability table determined by the table selection probability table in step 304. Based on, the next transport destination buffer station is determined, the transport vehicle is moved to the determined transport destination buffer station, and the movement result is evaluated. That is, in step 306, the determined probability table, the own probability table, and the transport request data of each buffer station are received from the center device 40, these data are stored in step 308, and determined in step 310. The transfer destination station is determined from the probability table, the parts are transferred in step 312, and the step 3
The evaluation of the transportation is performed at 14, 316.

Next, when the evaluation of the movement result in steps 314 and 316 is high, the table selection probability table of its own is updated in step 318. This update is performed as follows. In the transport vehicle (number n), the probability that the probability table of the transport vehicle (number q) is generally selected is S _nq . Further, the transport destination buffer station is determined according to the selected probability table, and the number of the transport vehicle to which the probability table belongs, which has been highly evaluated as a result of transporting the parts to the transport destination buffer station, is m.

[0037]

## EQU3 ## When q = m S _nm ← S _nm + b (1-S _nm ) (3)

## EQU4 ## When q ≠ m, S _nq ← (1-b) S _nq (4) where b is a constant (0 ≦ b ≦ 1).

The equation (3) uses the probability table of the transport vehicle (m) to determine the transport destination buffer station, and if the response to the transport request of the station is highly evaluated, the transport is performed thereafter. In order to increase the probability that the probability table of the vehicle (m) is selected, the probability S _{nm is set} to the probability that the probability table of the carrier vehicle (m) is not selected (1-S _nm ).
It means increasing by an amount proportional to. or,
(2), in the transport vehicle (n), in order to lower the probability that the random table of transport vehicle other than the transport vehicle (m) is selected, it is decreased by an amount proportional to the probability S _nq in probability S _nq Means that.

Further, in step 320, the buffer station selected as the next transfer destination this time is also selected for the next probability on the basis of the same state data, as described above, also for its own probability table. Update of its own probability table is performed so that the probability of being performed is high. And the updated table selection probability table,
The updated own probability table is transmitted from the carrier vehicle (n) to the center device 40. The center device 40 determines in step 410 that there is received data from the transport vehicle, and in step 410, stores the updated table selection probability table in the table selection probability table storage area 53, and stores the probability table in the probability table storage area. Store in 51. Then, in step 322, the number j of the transfer destination buffer station that has transferred the component is stored in the state data history storage area 31 as the state data.

In this way, if the result of responding to the transport request from each station is good, the selection probability of the probability table that the result is good will be high for subsequent responses to the transport request. In addition, the table selection probability table is updated, and the probability table of the own transport vehicle is updated so that the selection probability of the transport request is increased. , Each transport vehicle can determine the next destination buffer station so that it is in better condition. Therefore, each transport vehicle will eventually move according to the optimal transport order by self-learning.

As described above, one transport vehicle (n) can use the result obtained by the other transport vehicle through learning. Then, it is further determined by learning which transport vehicle the probability table should be used for. Therefore, it is possible to quickly find the optimum transfer order in response to more complicated system changes.

In the present embodiment, the table selection probability table is stored in the center device 40, and each transport device receives data from the center device 40 by data communication when necessary, and the updated table selection probability table is created. Although the table is transmitted to the center device 40, the table selection probability table may be stored in each transport vehicle. Further, the probability table may be stored in each transport vehicle instead of being stored in the center device 40. In this case, each carrier carries out data communication with other carriers,
The probability table determined by the table selection probability table is received from the carrier having the probability table, and the next transfer destination buffer station is determined.
Further, as a result of the response to the transportation request, the own probability table updated is stored in the own transportation vehicle. Further, each transport vehicle may monitor transport requests from each buffer station, store the transport requests, and store the table selection probability table and its own probability table. In that case, the center device itself becomes unnecessary.

In this embodiment, the present invention is implemented in a factory.
The present invention is applied to the moving method and moving device of the transport vehicle for transporting the parts to the station, but it can also be applied to the system for transporting the parts by truck to factories scattered in various places.
Further, it can be used in a taxi dispatch system in which taxis are dispatched and people are transferred to destinations in response to requests from stations scattered in various places. further,
Generally, it can be applied to a transportation system for people and goods. The evaluation for the response of the transport request is performed in the above-mentioned embodiment.
Although the transportation time, the number of stocks at the station, and the like are used, the state data that can evaluate the distribution efficiency of the transportation vehicle in the system such as the fuel consumption amount per unit distance and the vehicle allocation time can be used. In addition, as the history of the state data stored by each transport vehicle, the past transport destination buffer station is used, but the data indicating the system state such as the history of the number of parts in stock at each buffer station and the history of the past transport time are stored. You may use.

[Brief description of drawings]

FIG. 1 is a configuration diagram showing a transfer system controlled by a transfer control device for a transfer vehicle according to a specific embodiment of the present invention.

FIG. 2 is a block diagram showing a configuration of a movement control device.

FIG. 3 is a flowchart showing a processing procedure of a CPU mounted on a carrier vehicle of the movement control device.

FIG. 4 is a CPU mounted on a center device of the movement control device.
5 is a flowchart showing the processing procedure of FIG.

FIG. 5 is an explanatory diagram illustrating an example of a probability table.

FIG. 6 is an explanatory diagram showing an example of a table selection probability table.

FIG. 7 is a flowchart showing a processing procedure of a CPU mounted on a carrier vehicle of a movement control device according to another embodiment.

FIG. 8 is a C mounted on a center device of the movement control device.
The flowchart which showed the processing procedure of PU.

[Explanation of symbols]

 10 ... Belt conveyor X1 to X7 ... Assembly base B1 to B7 ... Buffer station M1 to M7 ... Assembly machine G1 to G7 ... Storage conveyor AGV1 to AGV5 ... Transport vehicle 31 ... Status data history storage area 32 ... Probability table input area 33 ... table selection probability table input area 51 ... probability table storage area 52 ... transport request storage area 53 ... table selection probability table storage area

Claims (8)

[Claims] 1. A system in which, in response to each of transport requests generated from a plurality of stations, each transport vehicle moves to each station to transport an object or person, and transports the object or person, A probability table in which values relating to the probability that the transport vehicle responds to each of the transport requests is described for each state data that represents a history of states that can be taken by the system is provided corresponding to each of the transport vehicles. Determines which transport request of the transport requests is to be responded to, based on the value associated with the probability in the probability table of the transport vehicle corresponding to the state data of the system obtained as a history up to that time. , Each of the transport vehicles moves based on the determined transport request to transport an object or person, and as a result, the response to the transport request is evaluated from the state of the system, and the evaluation is high. A movement scheduling method for a guided vehicle, characterized in that the probability table is updated so that the probability of responding to the transportation request becomes higher. 2. The state data is data indicating a state of each of the transport vehicles such as a history of responses of the transport vehicle to the transport request or a history of movement results of the transport vehicle. The method according to claim 1, wherein the movement scheduling method is for a transport vehicle. 3. The movement scheduling method for a guided vehicle according to claim 1, wherein the status data is data indicating a status at each of the stations that generate the transportation request. 4. The table selection probability table according to claim 1, further comprising a table selection probability table in which values associated with selected probabilities of each probability table are described, the table selection probability table being provided for each of the transportation vehicles. Based on a value related to the probability in the table selection probability table, select the probability table for determining the transport request to respond, the higher the evaluation of the response, the higher the probability table A movement scheduling method for a guided vehicle, comprising updating the table selection probability table so that the probability of selection is increased. 5. In response to each of the transport requests generated from each of the plurality of stations, each transport vehicle moves to each station to transport an object or person, and the transport vehicle transports the object or person. In the scheduling device, a probability table in which a value related to a probability that the transport vehicle responds to each transport request is described for each state data that represents a history of states that the system can store is stored corresponding to each transport vehicle. Probability table storage means, state data history storage means for storing the state data of the system for a certain period in the past corresponding to each of the transport vehicles, and the probability of the transport vehicle provided in each of the transport vehicles. Determining means for determining which of the transport requests the transport request responds to, based on a value related to the probability in the table; and the determining means provided in each of the transport vehicles. Detection means for detecting the state of the system in the result of movement of the transport vehicle based on the determined transport request; and evaluation of response to the transport request from the state detected by the detection means, which is provided in each of the transport vehicles. Evaluating means for performing, and the higher the evaluation by the evaluating means provided in each of the transport vehicles, the higher the probability of responding to the transport request,
A movement scheduling device for a guided vehicle, comprising: a first updating means for updating the probability table. 6. The table selection probability table storage unit according to claim 5, further comprising a table selection probability table storing a value associated with a selection probability of each probability table in association with each of the transport vehicles. Probability table selecting means provided in each of the transport vehicles for selecting the probability table for determining a transport request to respond based on a value associated with the probability in the table selection probability table; And a second updating means for updating the table selection probability table so that the higher the evaluation of the response is, the higher the probability that the probability table is selected becomes. Mobile scheduling device. 7. The system is a system for transporting an article to each of a plurality of stations by an automatic transport vehicle, and the status data for determining the probability table is data indicating a past transport destination station of the transport vehicle. 5. The state of the system for performing the evaluation is a state such as a time required for the transportation of the transportation vehicle or an inventory status at a transportation destination station, and the like. A method for scheduling the movement of a carrier vehicle. 8. The movement scheduling device is a device that automatically conveys articles to a plurality of stations, and the state data that determines the probability table is data indicating a past conveyance destination station of the conveyance vehicle. 7. The state of the system for performing the evaluation is the time required for the transport of the transport vehicle, or the inventory status at the transport destination station, or the like. A movement scheduling device for a guided vehicle as described.