Conveying device

Abstract

A transport facility is provided with a travelable path, a plurality of transport vehicles, and a control system. The control system is provided with: a state information acquisition unit that acquires state information indicating the arrangement of a plurality of transport vehicles present in a control region; an evaluation processing unit that outputs an evaluation value for each of the plurality of action patterns on the basis of the state information and an evaluation criterion; and a passage control unit that performs passage control of the plurality of transport vehicles in the control region in accordance with the action pattern having the highest evaluation value. The evaluation criterion is set so that the evaluation value increases as the average value of the required times for all the transport vehicles present in the control region decreases on the basis of the learning result.

Description

Technical Field

[0001] This invention relates to a conveying device. Background Technology

[0002] A conveying device is utilized, which includes a pre-set drivable path, multiple conveyor vehicles traveling on the drivable path, and a control system for controlling the multiple conveyor vehicles. An example of such a conveying device is disclosed in Japanese Patent Application Publication No. 2006-313463 (Patent Document 1).

[0003] In the article conveying device of Patent Document 1, when the conveyor (cart 5) approaches the junction (intersection with shortcut 3) in the drivable path (travel track 1), the conveyor sends a "blocking request" to the control system (area controller 11). Upon receiving the blocking request, the control system determines whether it is permissible to enter the junction and, if so, sends a "permission" back to the conveyor. By performing such an entry coordination process, collisions between conveyors at the junction can be avoided. Summary of the Invention

[0004] However, Patent Document 1 is a technology designed to prevent transport vehicles from colliding with each other. Even if it can avoid a significant drop in transport efficiency by avoiding collisions, it cannot improve the overall transport efficiency of multiple transport vehicles in the area including the confluence section.

[0005] Therefore, in conveying equipment, it is desirable to improve the overall conveying efficiency of multiple conveyor vehicles traveling in an area containing at least one confluence.

[0006] The conveying equipment disclosed herein is a conveying equipment having a pre-set drivable path, multiple conveyor vehicles traveling on the aforementioned drivable path, and a control system for controlling the multiple aforementioned conveyor vehicles. The aforementioned drivable path includes a junction where multiple paths converge. The aforementioned control system has: The status information acquisition unit acquires status information that shows the configuration of the multiple aforementioned transport vehicles present in the aforementioned control area when a control area is set up in a manner that includes at least one of the aforementioned merging units. The evaluation processing unit sets multiple action modes corresponding to combinations of whether or not the aforementioned transport vehicle is permitted to travel on each of the aforementioned paths, wherein the aforementioned paths are connected to the aforementioned merging points included in the aforementioned control area, and takes the aforementioned status information as input, and outputs an evaluation value for each of the aforementioned action modes based on an evaluation criterion obtained through learning; and The access control unit executes access control for the aforementioned transport vehicles within the aforementioned control area according to the aforementioned action mode with the highest aforementioned evaluation value. The required time is defined as the time from when the aforementioned transport vehicle enters the aforementioned controlled area until it passes through the aforementioned merging point. The aforementioned evaluation criteria are based on the results of learning the relationship between the aforementioned past state information, the aforementioned action patterns, and the aforementioned required time, and are set in such a way that the aforementioned evaluation value increases as the average of the aforementioned required time for all the aforementioned transport vehicles existing in the aforementioned control area decreases.

[0007] According to this configuration, the evaluation benchmark obtained through learning is set such that the evaluation value increases as the average time required for all transport vehicles existing in a control area containing at least one merging point decreases. Based on such an evaluation benchmark, the evaluation value of each of the multiple envisioned action modes is output, and traffic control is performed according to the action mode that achieves the highest evaluation value, thereby improving the overall transport efficiency of multiple transport vehicles traveling in an area containing at least one merging point.

[0008] Further features and advantages of the technology disclosed herein will become clearer from the following exemplary and non-limiting embodiments described with reference to the accompanying drawings. Attached Figure Description

[0009] Figure 1 This is a schematic diagram of the conveying equipment in the implementation method.

[0010] Figure 2 It is a block diagram of the control system.

[0011] Figure 3 It is a model diagram of the state information obtained from the status of multiple transport vehicles in the control area.

[0012] Figure 4 This is a schematic diagram showing the paths that each transport vehicle can travel at the confluence.

[0013] Figure 5 This is a diagram illustrating the permitted modes of action at the confluence.

[0014] Figure 6 It is a model diagram involving the input of the basic data used to calculate the evaluation value.

[0015] Figure 7 This is an example of the input used to calculate the basic data for the evaluation value. Detailed Implementation

[0016] The implementation method of the conveying equipment is described with reference to the accompanying drawings.

[0017] As in Figure 1and Figure 2 As shown, the conveying equipment 1 has a pre-set drivable path 2, multiple conveyor vehicles 4 that travel on the drivable path 2, and a control system 5 that controls the multiple conveyor vehicles 4.

[0018] The conveyor vehicle 4 transports goods. The conveyor vehicle 4 travels along the drivable path 2 and transports goods. For example, a covered conveyor vehicle can be used as the conveyor vehicle 4. Multiple conveyor vehicles 4 are configured to transport goods along the drivable path 2 respectively based on instructions from the control system 5.

[0019] Furthermore, various items exist as the items processed in the conveyor 1 and transported by the conveyor 4. For example, in the case where the conveyor 1 is used in a semiconductor manufacturing plant, items include wafer assemblies (so-called FOUPs: Front Opening Unified Pods) that hold wafers or mask assemblies (so-called mask cassettes) that hold masks. In this case, the conveyor 4 transports items such as wafer assemblies or mask assemblies between processes along the travel path 2.

[0020] The drivable path 2 is preset to a predetermined shape. In this embodiment, the drivable path 2 includes a main conveying path 21 formed in a loop, and multiple secondary conveying paths 22 each formed in a loop smaller than the main conveying path 21. The multiple secondary conveying paths 22 are connected to the main conveying path 21 via connecting paths 23. Furthermore, the drivable path 2 of this embodiment includes a shortcut path 24 that provides a shortcut along the relatively large loop of the main conveying path 21. In this embodiment, the main conveying path 21, secondary conveying paths 22, connecting paths 23, and shortcut path 24 are each equivalent to a "path".

[0021] Connecting path 23 branches off from one branch of the main transport path 21 and the secondary transport path 22, merging with another branch of the main transport path 21 and the secondary transport path 22. Shortcut path 24 branches off from one point on the main transport path 21, merging with another point on the main transport path 21. Thus, the drivable path 2 contains a merging point J where multiple paths converge. In this example, the point where the main transport path 21 or the secondary transport path 22 merges with connecting path 23, or the point where the main transport path 21 merges with shortcut path 24, corresponds to merging point J.

[0022] When a canopy-mounted transport vehicle is used as transport vehicle 4, the travel path 2 can be formed by a track suspended and supported from the canopy.

[0023] Control system 5 controls multiple conveyor vehicles 4. For example, in... Figure 2As shown, the control system 5 of this embodiment consists of a higher-level control device 50 and a terminal control device 59, wherein the terminal control device 59 is mounted on each of the plurality of transport vehicles 4 in a state capable of communicating with the higher-level control device 50. Furthermore, the control system 5 includes an integrated control unit 51, a status information acquisition unit 52, a required time measurement unit 53, an evaluation processing unit 54, and a passage control unit 55. In this embodiment, the higher-level control device 50 includes these integrated control unit 51, status information acquisition unit 52, required time measurement unit 53, evaluation processing unit 54, and passage control unit 55.

[0024] The integrated control unit 51 performs comprehensive control over the multiple transport vehicles 4 of the entire transport equipment 1. For example, when transporting items to their destination, the integrated control unit 51 performs the following controls: dispatching an empty transport vehicle 4 to the origin; allowing the transport vehicle 4 to receive items; or dispatching a loaded transport vehicle 4 to the destination. The integrated control unit 51 performs these controls for each item being transported. In addition, the integrated control unit 51 monitors the position of each of the multiple transport vehicles 4.

[0025] The status information acquisition unit 52 acquires status information F containing information showing the configuration of the transport vehicle 4 present in the control area A, wherein the control area A is configured to include at least one merging section J. The control area A is, for example, configured as an area including at least one merging section J and occupying a range within a predetermined distance upstream and downstream of each merging section J. As an example, in Figure 3 The upper part shows the control area A located near the junction J of the main transport path 21 and the shortcut path 24. In this example, the middle part of the main transport path 21 is connected by two shortcut paths 24.

[0026] exist Figure 3 In the example, the main transport path 21 includes a first straight track 26 and a second straight track 27 arranged in opposite directions and in parallel. Additionally, two shortcut paths 24 include a first connecting track 28 and a second connecting track 29, wherein the first connecting track 28 and the second connecting track 29 are respectively curved in an arc shape, arranged facing each other and connected to the first straight track 26 and the second straight track 27. A first entry position Pi1 is set upstream of the bifurcation point between the first straight track 26 and the first connecting track 28, and a second entry position Pi2 is set upstream of the bifurcation point between the second straight track 27 and the second connecting track 29. A first exit position Po1 is set downstream of the confluence J of the first straight track 26 and the second connecting track 29, and a second exit position Po2 is set downstream of the confluence J of the second straight track 27 and the first connecting track 28.

[0027] Furthermore, the area comprising the following portions is designated as control area A: the portion of the first straight track 26 extending from a predetermined distance upstream of the first entry position Pi1 to a predetermined distance downstream of the first exit position Po1; the portion of the second straight track 27 extending from a predetermined distance upstream of the second entry position Pi2 to a predetermined distance downstream of the second exit position Po2; the first connecting track 28; and the second connecting track 29. In addition, the predetermined distance in this case can be any distance within the range of 3 to 10 meters.

[0028] When there is only one conveyor vehicle 4 in control area A, the status information F acquired by the status information acquisition unit 52 becomes information containing the following: showing the position of the conveyor vehicle 4 in control area A. Furthermore, when there are multiple conveyor vehicles 4 in control area A, the status information F acquired by the status information acquisition unit 52 becomes information containing the following: showing the configuration of the multiple conveyor vehicles 4 present in control area A (i.e., the combination of their positions).

[0029] Status information acquisition unit 52 will indicate the location of the conveyor 4 in control area A, such as... Figure 3 The lower half of the diagram shows the model-like data acquisition. More specifically, the state information acquisition unit 52 divides the portion of the first straight track 26 upstream of the first entry position Pi1 at fixed intervals, and acquires it as positions Z11, Z12, Z13, Z14, and Z15. Additionally, the state information acquisition unit 52 acquires the portion of the first straight track 26 between the first entry position Pi1 and the first exit position Po1 as position Zm1. Furthermore, the state information acquisition unit 52 divides the portion of the first straight track 26 downstream of the first exit position Po1 at fixed intervals, and acquires it as positions Z16, Z17, Z18, Z19, and Z20. Moreover, position Z15 corresponds to the first entry position Pi1, and position Z16 corresponds to the first exit position Po1.

[0030] Furthermore, the status information acquisition unit 52 divides the portion of the second straight track 27 upstream of the second entry position Pi2 at fixed intervals, assigning it positions Z21, Z22, Z23, Z24, and Z25. Additionally, the status information acquisition unit 52 assigns the portion of the second straight track 27 between the second entry position Pi2 and the second exit position Po2 as position Zm2. Furthermore, the status information acquisition unit 52 divides the portion of the second straight track 27 downstream of the second exit position Po2 at fixed intervals, assigning it positions Z26, Z27, Z28, Z29, and Z30. Moreover, position Z25 corresponds to the second entry position Pi2, and position Z26 corresponds to the second exit position Po2. Furthermore, the status information acquisition unit 52 assigns the portion of the first connecting track 28 as position Zm3 and the portion of the second connecting track 29 as position Zm4.

[0031] exist Figure 3 In the example, if there are multiple transport vehicles 4 in the control area A, for example as shown in the upper part, the status information acquisition unit 52 acquires status information F as a combination of information including the positions of each transport vehicle 4. That is, in this case, the status information F is acquired as (Z11, Z14, Z15, Z25, Z28, Zm3, Zm4).

[0032] The status information F may simply contain information about the position of each conveyor 4, but it may also further include information indicating the degree of waiting time of each conveyor 4 in control area A. The information indicating the degree of waiting time can be either the length of the waiting time at that moment (i.e., the stopping time at a specific location) itself, or information that ranks the length of the waiting time. In this example, the status information F includes information about the length of the waiting time of each conveyor 4 in addition to the position information. Figure 3 In the example shown in the upper part, if a portion of the multiple transport vehicles 4 has been in standby for the displayed time, the status information F in this case is obtained as (Z11 (0 seconds), Z14 (10 seconds), Z15 (20 seconds), Z25 (50 seconds), Z28 (0 seconds), Zm3 (0 seconds), Zm4 (0 seconds)).

[0033] In this embodiment, the status information F is set to include the travel direction information of each transport vehicle 4. Here, the travel direction information means the information of the predetermined path of the transport vehicle 4. More specifically, the travel direction information includes information indicating whether the transport vehicle 4, which has traveled on the first straight track 26 and reached the first entry position Pi1, is destined to continue straight on the straight track 26, or is destined to change its path to the first connecting track 28. In addition, the travel direction information includes information indicating whether the transport vehicle 4, which has traveled on the second straight track 27 and reached the second entry position Pi2, is destined to continue straight on the second straight track 27, or is destined to change its path to the second connecting track 29.

[0034] Furthermore, the waiting time information of each transport vehicle 4 can be obtained, for example, as control information held by the integrated control unit 51, or as information actually measured by each transport vehicle 4 existing in the control area A. Additionally, the travel direction information of each transport vehicle 4 can be obtained, for example, as control information held by the integrated control unit 51.

[0035] The status information acquisition unit 52 acquires status information F at a predetermined period (e.g., every 1 second).

[0036] The required time measurement unit 53 measures the required time from the time the transport vehicle 4 enters the control area A until it passes through the confluence section J. In this embodiment, the required time measurement unit 53 measures the required time for each of the multiple transport vehicles 4 based on the position information of each of the multiple transport vehicles 4 held by the integrated control unit 51.

[0037] The evaluation processing unit 54 takes the status information F obtained by the status information acquisition unit 52 as input, and outputs the evaluation value for each of the multiple action modes based on the evaluation benchmark ST obtained through learning. Here, multiple action modes are set accordingly to the combination of whether the transport vehicle 4 is permitted to travel on each of the multiple paths, wherein the multiple paths are connected to the merging section J included in the control area A.

[0038] exist Figure 4 The conveyor 4 is shown. Figure 3The illustrated control area A offers multiple selectable routes. As shown in the figure, the transport vehicle 4 within control area A can select any one of the following: a first straight route Ws1, a second straight route Ws2, a first curved route Wc1, and a second curved route Wc2. The first straight route Ws1 is the route taken by the transport vehicle 4 traveling on the first straight track 26. The second straight route Ws2 is the route taken by the transport vehicle 4 traveling on the second straight track 27. The first curved route Wc1 is the route where the transport vehicle 4, traveling on the first straight track 26, branches off from the first straight track 26, travels on the first connecting track 28 (curve), and then rejoins the second straight track 27. The second curved route Wc2 is the route where the transport vehicle 4, traveling on the second straight track 27, branches off from the second straight track 27, travels on the second connecting track 29 (curve), and then rejoins the first straight track 26.

[0039] In this embodiment, whether passage is permitted on the first connecting track 28 is represented by "a1", and whether passage is permitted on the portion between the first entry position Pi1 and the first exit position Po1 of the first straight track 26 is represented by "a2". Furthermore, whether passage is permitted on the second connecting track 29 is represented by "a3", and whether passage is permitted on the portion between the second entry position Pi2 and the second exit position Po2 of the second straight track 27 is represented by "a4". These passage permissions "a1", "a2", "a3", and "a4" represent either "permitted" or "prohibited".

[0040] If passage is permitted for the first connecting track 28 (a1; permitted), the transport vehicle 4 can proceed along the first curved roadway Wc1. If passage is permitted downstream of the first entry position Pi1 of the first straight track 26 (a2; permitted), the transport vehicle 4 can proceed along the first straight roadway Ws1. If passage is permitted for the second connecting track 29 (a3; permitted), the transport vehicle 4 can proceed along the second curved roadway Wc2. If passage is permitted downstream of the second entry position Pi2 of the second straight track 27 (a4; permitted), the transport vehicle 4 can proceed along the second straight roadway Ws2.

[0041] In this embodiment, among the 16 combinations involving whether passage is possible at any of the four points, excluding the 7 combinations where there is a possibility of collision at any confluence point J and the 5 combinations where excessive passage prohibition significantly reduces transport efficiency, 4 combinations are set as the action mode for output evaluation values. Specifically, as in Figure 5As shown, the first action mode permitted by "a1" and "a2" but prohibited by others, the second action mode permitted by "a1" and "a3" but prohibited by others, the third action mode permitted by "a2" and "a4" but prohibited by others, and the fourth action mode permitted by "a3" and "a4" but prohibited by others are set as the action modes of the object of the output evaluation value.

[0042] Furthermore, in the first operating mode, the multiple transport vehicles 4 located in control area A can proceed along either the first curved road line Wc1 or the first straight road line Ws1, respectively. In the second operating mode, the multiple transport vehicles 4 located in control area A can proceed along either the first curved road line Wc1 or the second curved road line Wc2, respectively. In the third operating mode, the multiple transport vehicles 4 located in control area A can proceed along either the first straight road line Ws1 or the second straight road line Ws2, respectively. In the fourth operating mode, the multiple transport vehicles 4 located in control area A can proceed along either the second curved road line Wc2 or the second straight road line Ws2, respectively.

[0043] As described above, the evaluation processing unit 54 takes the state information F as input and, based on the evaluation benchmark ST obtained through learning, outputs an evaluation value for each action mode among multiple action modes. Figure 6 The diagram shows the model involved in the input of the basic data used by the evaluation processing unit 54 to calculate the evaluation value.

[0044] In the model diagram of 12 rows × 12 columns (cells A1~L12), each cell shown below represents the following position in control area A.

[0045] Units A6~E6; Positions Z11~Z15 at the first curve road line Wc1 Units F5~F1; Positions Z26~Z30 at the first curve (road alignment Wc1). Units A7~E7; Positions Z11~Z15 when the first straight route Ws1 is in progress. Units F8~F12; Positions Z16~Z20 when the first straight route Ws1 is in progress. Units L7~H7; Positions Z21~Z25 at the second curve road line Wc2 Units G8~G12; Positions Z16~Z20 at the second curve (road alignment Wc2). Units L6~H6; Positions Z21~Z25 when traveling the second straight route Ws2 Units G5~G1; Positions Z26~Z30 on the second straight route Ws2 Units A1~D4; Location Zm3 Units A12~D9; Location Zm1 Units L12~I9; Location Zm4 Units L1~I4; Location Zm2.

[0046] Furthermore, in each of the above units, based on the obtained state information F, the input values ​​are input according to the following rules.

[0047] • Default value → 0 • When there is a transport vehicle 4 at this location → +0.5 • When there is a waiting time → +f(waiting time) ※However, 0 < f(waiting time) ≤ 0.5.

[0048] Additionally, for each of the cells shown below, enter information representing the following content.

[0049] Unit E5; Is "a1" allowed to pass at this moment? Unit E8; Is "a2" passable at this moment? Unit H8; Is "a3" passable at this moment? Unit H5; Is "a4" allowed to pass at this moment? Unit F6; Whether the behavior pattern of the evaluation subject is acceptable ("a1") Unit F7; Whether the behavior pattern of the evaluation subject is acceptable ("a2") Unit G7; Whether the behavior pattern of the evaluation subject is acceptable ("a3") Unit G6; Whether the behavior pattern of the evaluation object is acceptable "a4".

[0050] Specifically, the input values ​​are entered in each of the above units according to the following rules.

[0051] • Is it permissible to pass through "a1" to "a4" when "permitted"? → 0.8 • Whether passage is allowed when “a1”~“a4” are “prohibited” → 0.2.

[0052] exist Figure 7 Quoting again Figure 3 The configuration example is provided, and the input of the basic data used to calculate the evaluation value is explained in more detail. Furthermore, in Figure 7 To distinguish the multiple transport vehicles 4 from one another, they are designated as “4A” to “4G”. In addition, transport vehicles 4 that will advance on the first straight route Ws1 or the second straight route Ws2 are indicated in white, and transport vehicles 4 that will advance on the first curved route Wc1 or the second curved route Wc2 are indicated in black.

[0053] In this example, passage of the first connecting track 28 and the second connecting track 29 is permitted at this moment. That is, passage of "a1" and "a3" is "permitted", and passage of "a2" and "a4" is "prohibited". Therefore, "0.8" is entered in cells E5 and H8 corresponding to passage of "a1" and "a3", and "0.2" is entered in cells E8 and H5 corresponding to passage of "a2" and "a4".

[0054] Conveyor 4A is a vehicle scheduled to travel on the first curved road line Wc1, located at position Z11 on the first straight track 26. At this moment, there is no waiting time for conveyor 4A. Therefore, in cell A6 corresponding to position Z11 when traveling on the first curved road line Wc1, input "0.5".

[0055] Either conveyor 4B or 4C is a conveyor 4 scheduled to travel on the first straight route Ws1, located at positions Z14 and Z15 on the first straight track 26, respectively. At this moment, the waiting times for conveyors 4B and 4C are 10 seconds and 20 seconds, respectively. In this example, the sum of values ​​for a waiting time of 10 seconds is "0.03", and the sum of values ​​for a waiting time of 20 seconds is "0.13". Therefore, in cells D7 and E7 corresponding to positions Z14 and Z15 on the first straight route Ws1, respectively, input "0.53 (=0.5+0.03)" and "0.63 (=0.5+0.13)".

[0056] Transport vehicle 4D is currently traveling on the first curve road line Wc1, located at position Zm3 on the first connecting track 28. Transport vehicle 4D is in motion and has not incurred any waiting time. Therefore, in cell A1, which is one of the cells A1~D4 corresponding to position Zm3, enter "0.5".

[0057] Conveyor 4E is a scheduled vehicle 4 traveling on the second straight route Ws2, located at position Z25 on the second straight track 27. At this moment, the waiting time for conveyor 4E is 50 seconds. In this example, the sum of the waiting times of 50 seconds is "0.5". Therefore, in cell H6 corresponding to position Z25 on the second straight route Ws2, enter "1 (=0.5+0.5)".

[0058] Transport vehicle 4F is currently traveling on the second curve road line Wc2, located at position Zm4 on the second connecting track 29. Transport vehicle 4F is in motion and has not incurred any waiting time. Therefore, in cell L12, which is one of the cells L12~I9 corresponding to position Zm4, enter "0.5".

[0059] Conveyor 4G is currently traveling on the second straight route Ws2, located at position Z28 on the second straight track 27. Conveyor 4G is in motion and has not incurred any waiting time. Therefore, in cell G3 corresponding to position Z28 when traveling on the second straight route Ws2, input "0.5".

[0060] In this embodiment, the action mode of the evaluation object is limited to the first to fourth action modes as described above. Therefore, in units F6 to G7 corresponding to whether the action mode of the evaluation object is passable "a1" to "a4", "0.8" or "0.2" is input accordingly to the action mode of the evaluation object.

[0061] When evaluating the first action mode, input "0.8" in cells F6 and F7 corresponding to whether "a1" and "a2" can pass, and input "0.2" in cells G7 and G6 corresponding to whether "a3" and "a4" can pass.

[0062] When evaluating the second action mode, input "0.8" in cells F6 and G7 corresponding to whether "a1" and "a3" can pass, and input "0.2" in cells F7 and G6 corresponding to whether "a2" and "a4" can pass.

[0063] When evaluating the third action mode, input "0.8" in cells F7 and G6 corresponding to whether "a2" and "a4" can pass, and input "0.2" in cells F6 and G7 corresponding to whether "a1" and "a3" can pass.

[0064] When evaluating the fourth action mode, input "0.8" in units G7 and G6 corresponding to whether "a3" and "a4" can pass, and input "0.2" in units F6 and F7 corresponding to whether "a1" and "a2" can pass.

[0065] The evaluation processing unit 54 outputs an evaluation value based on the basic input data and the evaluation benchmark ST obtained through learning. Here, the evaluation benchmark ST is defined as a function that performs calculations using the input values ​​and returns the result. The evaluation benchmark ST is set based on the result of learning the relationship between past state information F, action modes, and required time, with the evaluation value increasing as the average required time for all transport vehicles 4 existing in control area A decreases. By calculating the evaluation value based on this evaluation benchmark ST, and selecting the action mode assigned a high evaluation value, subsequent transport efficiency can be improved.

[0066] In this embodiment, the evaluation criterion ST is set such that the evaluation value increases as the variance of the required time for all transport vehicles 4 present in control area A decreases. That is, the evaluation criterion ST is set such that the evaluation value increases as the average value and variance of the required time for all transport vehicles 4 present in control area A decrease. Based on this evaluation criterion ST, an evaluation value is calculated, and by selecting an action mode that assigns a high evaluation value, subsequent transport efficiency can be improved while simultaneously reducing the deviation in required time.

[0067] Furthermore, the improvement in conveying efficiency in these situations can also be limited to maximizing the conveying efficiency at that particular moment and in that particular situation.

[0068] In this embodiment, the evaluation benchmark ST, which aims to improve delivery efficiency in the long term, is defined by including an Action-Value Function. This Action-Value Function calculates the expected value of the cumulative reward when the reward increases as the required time decreases. Furthermore, in this embodiment, the evaluation benchmark ST is defined by including an Action-Value Function, which calculates the expected value of the cumulative reward when the reward increases as the required time decreases and the variance decreases. Such an Action-Value Function can be derived using reinforcement learning, such as Deep-Q Network (DQN).

[0069] Thus, when the evaluation criterion ST includes an action value function, the evaluation value output by the evaluation processing unit 54 becomes the value (Q-value, Q-value) of the action value function for each of the multiple action modes given the state information F at that moment as input. By calculating the evaluation value based on the evaluation criterion ST including the action value function, and continuously selecting action modes that assign high evaluation values, the average time required for the multiple transport vehicles 4 can be reduced not only in that specific instance, but continuously. Therefore, the transport efficiency can be significantly improved in the long term.

[0070] The evaluation and processing unit 54 outputs evaluation values ​​at predetermined intervals.

[0071] The passage control unit 55 compares the evaluation values ​​(in this embodiment, the values ​​of the action value function) output for each action mode of the multiple action modes of the evaluation object, and performs passage control on the multiple transport vehicles 4 in the control area A according to the action mode with the highest evaluation value.

[0072] When the evaluation value for the first operation mode is maximized, the traffic control unit 55 controls the target transport vehicle 4 to proceed in sequence along the first straight route Ws1 and the first curved route Wc1. When the evaluation value for the second operation mode is maximized, the traffic control unit 55 controls the target transport vehicle 4 to proceed in sequence along the first curved route Wc1 and the second curved route Wc2. When the evaluation value for the third operation mode is maximized, the traffic control unit 55 controls the target transport vehicle 4 to proceed in sequence along the first straight route Ws1 and the second straight route Ws2. When the evaluation value for the fourth operation mode is maximized, the traffic control unit 55 controls the target transport vehicle 4 to proceed in sequence along the second straight route Ws2 and the second curved route Wc2.

[0073] The traffic control unit 55 performs traffic control at a predetermined cycle.

[0074] In this way, by executing the passage control of multiple transport vehicles 4 in control area A according to the action mode that assigns the maximum evaluation value (the maximum value of the action value function in this embodiment), the transport efficiency can be substantially improved in the long run.

[0075] [Other Implementation Methods] (1) In the above embodiment, specific examples are given regarding the input of the basic data used to calculate the evaluation value. However, the numerical values ​​corresponding to various phenomena (e.g., "+0.5" when the conveyor 4 is present, "+f (waiting time)" when there is a waiting time, "0.8" when passage is permitted, "0.2" when passage is prohibited, etc.) are given as examples. These can be appropriately set, for example, according to the general specifications required for the conveyor 1 or the specific circumstances of each conveyor 1.

[0076] (2) In the above embodiment, the input of the basic data used to calculate the evaluation value will be described as an example of the following configuration: when there is a waiting time, "+f(waiting time)" is determined discretely in relation to the waiting time. However, this configuration is not limited to this; "+f(waiting time)" may also be determined linearly in relation to the waiting time. Alternatively, the values ​​may not be added regardless of whether there is a waiting time, in which case the status information F becomes information that does not include information showing the degree of waiting time of each transport vehicle 4 in the control area A.

[0077] (3) In the above embodiment, an example in which the waiting time is defined as the stopping time at a specific position is described. However, it is not limited to this configuration, and the stopping time may also be defined, for example, as the time from the initial stopping in the control area A until passing through the first exit position Po1 or the second exit position Po2. In addition, the waiting time can be defined in various other ways.

[0078] (4) In the above embodiment, the following configuration is used as an example: the status information F includes information showing the configuration of the plurality of transport vehicles 4 present in the control area A, information showing the degree of waiting time of each transport vehicle 4 in the control area A, and the travel direction information of each transport vehicle 4 (predetermined travel path information). However, the configuration is not limited to this. The status information F may include at least the information showing the configuration of the plurality of transport vehicles 4 present in the control area A, and may not include either the information showing the degree of waiting time or the travel direction information. Alternatively, the status information F may further include other information. Examples of other information include, for example, the speed information of each transport vehicle 4 or the direction of the transport vehicle 4 at that moment. Alternatively, examples of other information include, for example, information related to whether or not a permit has been obtained or information related to the transport task (whether or not there is cargo or task priority).

[0079] (5) In the above embodiment, the following configuration will be described as an example: the evaluation benchmark ST obtained by learning is set such that the evaluation value increases as the average time required for all transport vehicles 4 existing in the control area A decreases. However, this configuration is not limited to this one. When calculating the evaluation value, in addition to the average time required for all transport vehicles 4, phenomena that affect the required time, such as the travel speed, stopping time, and number of stops of each transport vehicle 4, can also be considered.

[0080] (6) In the above embodiment, the following configuration will be described as an example: the status information acquisition unit 52 grasps the portion between the first entry position Pi1 and the first exit position Po1 in the first straight track 26 as a single position (position Zm1). However, it is not limited to this configuration; the status information acquisition unit 52 may also grasp the portion between the first entry position Pi1 and the first exit position Po1 in the first straight track 26 as subdivided positions (positions Zm11, Zm12, ...). Similarly, the status information acquisition unit 52 may also grasp the portion between the second entry position Pi2 and the second exit position Po2 in the second straight track 27 as subdivided positions (positions Zm21, Zm22, ...). If this is done, the accuracy of the status information F (in particular, information on the configuration of the multiple transport vehicles 4 is shown) can be improved, and the evaluation value can be calculated more accurately.

[0081] (7) In the above embodiment, the following configuration will be described as an example: the status information acquisition unit 52 grasps a portion of the first connecting track 28 as a single position (position Zm3). However, it is not limited to this configuration; the status information acquisition unit 52 may also grasp a portion of the first connecting track 28 as subdivided positions (positions Zm31, Zm32, ...). Similarly, for a portion of the second connecting track 29, the status information acquisition unit 52 may also grasp that portion as subdivided positions (positions Zm41, Zm42, ...). If this is done, the accuracy of the status information F (here, in particular, information on the configuration of the multiple transport vehicles 4 is shown) can be improved, and the evaluation value can be calculated more accurately.

[0082] (8) In the above embodiment, the following configuration is described as an example: the upper control device 50 is equipped with a required time measuring unit 53, which measures the required time for each of the plurality of transport vehicles 4. However, it is not limited to this configuration. For example, the terminal control device 59 of each of the plurality of transport vehicles 4 may be equipped with a required time measuring unit 53, and the results measured by each required time measuring unit 53 may be sent to the upper control device 50.

[0083] (9) In the above embodiment, the following configuration is used as an example: the control system 5 is composed of a higher-level control device 50 and a terminal control device 59, wherein the terminal control device 59 is mounted on each of the plurality of transport vehicles 4 in a state that enables communication with the higher-level control device 50. However, it is not limited to this configuration, for example, the control system 5 may be composed of a terminal control device 59, wherein the terminal control device 59 is mounted on each of the plurality of transport vehicles 4 in a state that enables communication with each other.

[0084] (10) In the above embodiment, it is mainly envisioned that a covered transport vehicle be used as the configuration of the transport vehicle 4. However, it is not limited to this configuration, and the transport vehicle 4 may also be a rail-guided trolley or an automated guided vehicle (AGV). In the latter case, the travel path 2 may also be composed of a magnetic tape or the like installed on the ground.

[0085] (11) The configurations disclosed in the above embodiments (including the above embodiments and other embodiments; the same applies below) can also be combined and applied with the configurations disclosed in other embodiments, provided that there is no contradiction. Regarding other configurations, the embodiments disclosed in this specification are exemplified in all respects and can be appropriately modified without departing from the spirit of this disclosure.

[0086] [Summary of Implementation Methods] In summary, the conveying equipment disclosed herein suitably comprises the following components.

[0087] A conveying device comprising a pre-set drivable path, multiple conveyor vehicles traveling along the drivable path, and a control system for controlling the multiple conveyor vehicles. The aforementioned drivable path includes a junction where multiple paths converge. The aforementioned control system has: The status information acquisition unit acquires status information that shows the configuration of the multiple aforementioned transport vehicles present in the aforementioned control area when a control area is set in a manner that includes at least one of the aforementioned merging units. The evaluation processing unit sets multiple action modes corresponding to combinations of whether or not the aforementioned transport vehicle is permitted to travel on each of the aforementioned paths, wherein the aforementioned paths are connected to the aforementioned merging points included in the aforementioned control area, and takes the aforementioned status information as input, and outputs an evaluation value for each of the aforementioned action modes based on an evaluation criterion obtained through learning; and The access control unit executes access control for the aforementioned transport vehicles within the aforementioned control area according to the aforementioned action mode with the highest aforementioned evaluation value. The required time is defined as the time from when the aforementioned transport vehicle enters the aforementioned controlled area until it passes through the aforementioned merging point. The aforementioned evaluation criteria are based on the results of learning the relationship between the aforementioned past state information, the aforementioned action patterns, and the aforementioned required time, and are set in such a way that the aforementioned evaluation value increases as the average of the aforementioned required time for all the aforementioned transport vehicles existing in the aforementioned control area decreases.

[0088] According to this configuration, the evaluation benchmark obtained through learning is set such that the evaluation value increases as the average time required for all transport vehicles existing in a control area containing at least one merging point decreases. Based on such an evaluation benchmark, the evaluation value of each of the multiple envisioned action modes is output, and traffic control is performed according to the action mode that achieves the highest evaluation value, thereby improving the overall transport efficiency of multiple transport vehicles traveling in an area containing at least one merging point.

[0089] As a preferred option, The aforementioned evaluation criteria are then set such that the evaluation value increases as the variance of the required time for all the aforementioned transport vehicles existing in the aforementioned control area decreases.

[0090] According to this configuration, the evaluation criterion obtained through learning is set such that the evaluation value increases as the average and variance of the required time for all transport vehicles existing in the controlled area decrease. Traffic control is performed based on this evaluation criterion, thereby improving the overall transport efficiency of multiple transport vehicles traveling in an area containing at least one junction, while simultaneously reducing the deviation in required time.

[0091] As a preferred option, The aforementioned status information further includes information indicating the degree of waiting time of each of the aforementioned transport vehicles in the aforementioned control area.

[0092] Based on this configuration, an evaluation criterion is established that considers not only the configuration of multiple transport vehicles in the controlled area, but also the degree of waiting time for each transport vehicle. This increases the likelihood of reducing the deviation in the required time for multiple transport vehicles traveling in an area containing at least one confluence.

[0093] As a preferred option, The aforementioned evaluation criteria include an action value function, which calculates the expected value of the cumulative reward when the reward is increased as the required time decreases. The aforementioned evaluation processing unit outputs the value of the aforementioned action value function for each of the aforementioned action modes when the aforementioned state information at that moment is used as input, as the aforementioned evaluation value.

[0094] Based on this configuration, under constantly changing transport conditions, traffic control is repeatedly executed according to the action pattern that achieves the highest evaluation value (i.e., the highest value of the action value function) at each time point. In the long run, this can improve the overall transport efficiency of multiple transport vehicles traveling in an area containing at least one confluence point.

[0095] The conveying equipment involved in this disclosure only needs to be able to achieve at least one of the above-mentioned effects.

[0096] Explanation of reference numerals in the attached figures 1. Conveying equipment 2. Driving routes 4. Conveyor vehicle 5 Control System 52 Status Information Acquisition Department 54 Evaluation and Processing Department 55. Traffic Control Department J Convergence Department A control area ST Evaluation Benchmark F Status Information.

Claims

1. A conveying device comprising a pre-set drivable path, a plurality of conveyor vehicles driving along the drivable path, and a control system for controlling the plurality of said conveyor vehicles. The conveying equipment has the following characteristics: The drivable path includes a junction where multiple paths converge. The control system includes: A status information acquisition unit acquires status information that shows the configuration of the multiple transport vehicles present in the control area when multiple transport vehicles exist in a control area configured in a manner including at least one of the merging units. An evaluation processing unit sets multiple action modes corresponding to combinations of whether or not the transport vehicle is permitted to travel on each of the multiple paths, wherein the multiple paths are connected to the merging point included in the control area, and outputs an evaluation value for each of the multiple action modes based on the evaluation benchmark obtained through learning, using the status information as input. as well as The access control unit executes access control for multiple transport vehicles in the control area according to the action mode with the highest evaluation value. The required time is defined as the time from when the transport vehicle enters the control area until it passes through the merging section. The evaluation benchmark is based on the result of learning the relationship between the past state information and the action pattern and the required time, and is set in such a way that the evaluation value increases as the average required time for all the transport vehicles present in the control area decreases.

2. The conveying device according to claim 1, wherein, The evaluation criterion is then set such that the evaluation value increases as the variance of the required time for all the transport vehicles present in the control area decreases.

3. The conveying device according to claim 1 or 2, wherein, The status information further includes information indicating the degree of waiting time of each of the plurality of transport vehicles in the control area.

4. The conveying device according to claim 1 or 2, wherein, The evaluation criterion is defined by an action value function, which calculates the expected value of the cumulative reward when the required time decreases and the reward becomes larger. The evaluation processing unit outputs the evaluation value as the value of the action value function for each of the multiple action modes when the state information at that moment is used as input.

Images