Route search method, route search device, and transportation system

Abstract

To provide a route search method capable of determining an optimal route for a battery-equipped transport vehicle.SOLUTION: A route search method includes: setting a destination for a battery-equipped transport vehicle; calculating a cost between base stations for reaching the destination, wherein input variables of a cost function for calculating the cost include an amount of charging and discharging that occurs between the base stations; and determining a route to the destination based on the calculated cost.SELECTED DRAWING: Figure 7

Description

Technical Field

[0001] The present invention relates to a path search method, a path search device, and a conveyance system.

Background Art

[0002] In the manufacturing process of semiconductor devices, substrates can be transferred by an unmanned conveyance system. In particular, the unmanned conveyance system can include a conveyance vehicle (for example, an OHT (Overhead Hoist Transport), an RGV (Rail Guided Vehicle), etc.) configured to be movable along a traveling rail installed on the ceiling or floor of a clean room. The operation control of the conveyance vehicle can be controlled by a host controller such as an OCS (OHT Control Server) device.

Summary of the Invention

Problems to be Solved by the Invention

[0003] The conveyance vehicle can receive power supply through a power line installed on the traveling rail and move. When the conveyance vehicle moves on a traveling rail without a power line with a battery inside, it can move using the energy stored in the battery.

[0004] The problem to be solved by the present invention is to provide a path search method capable of determining an optimal path for a conveyance vehicle equipped with a battery.

[0005] The problem to be solved by the present invention is to provide a path search device capable of determining an optimal path for a conveyance vehicle equipped with a battery.

[0006] The problem to be solved by the present invention is to provide a conveyance system capable of determining an optimal path for a conveyance vehicle equipped with a battery.

[0007] The problems of the present invention are not limited to the problems mentioned above, and other problems not mentioned will be clearly understood by those skilled in the art from the following description.

Means for Solving the Problems

[0008] The route search method according to some embodiments of the present invention for achieving the above problems includes setting a destination of a carrier vehicle using a battery, calculating a cost between bases for reaching the destination, where input variables of a cost function for calculating the cost include the charge and discharge amount generated between the bases, and determining a route to the destination based on the calculated cost.

[0009] The route search device according to some embodiments of the present invention for achieving the other problems includes a destination setting unit that sets a destination of a carrier vehicle using a battery; a cost calculation unit that calculates a cost between bases for reaching the destination, where input variables of a cost function for calculating the cost include the charge and discharge amount generated between the bases; and a route determination unit that determines a route to the destination based on the calculated cost.

[0010] The transport system according to some embodiments of the present invention for achieving the above and other problems includes a power supply section where a power line for supplying power is installed, a rail including a non-power supply section where the power line is not installed; a carrier vehicle that moves along the rail and has a battery for storing power supplied through the power line; and a host controller that controls the plurality of carrier vehicles, where the host controller sets a destination of the carrier vehicle, calculates a cost between bases for reaching the destination, input variables of a cost function for calculating the cost include the charge and discharge amount generated between the bases, and determines a route to the destination based on the calculated cost.

[0011] Specific contents of other embodiments are included in the detailed description and drawings.

Brief Description of the Drawings

[0012]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Embodiments for Carrying Out the Invention

[0013] Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings. The advantages and features of the present invention, as well as the methods for achieving them, will become clear by referring to the embodiments described in detail hereinafter together with the accompanying drawings. However, the present invention is not limited to the embodiments disclosed below and can be realized in various different forms. These embodiments are merely provided to complete the disclosure of the present invention and to fully inform those with ordinary knowledge in the technical field to which the present invention pertains of the scope of the invention. The present invention is defined only by the scope of the claims. Throughout the specification, the same reference numerals shall refer to the same components.

[0014] Spatially relative terms such as "below", "beneath", "lower", "above", "upper", etc. are used to easily describe the correlation between one element or component and another element or component as shown in the drawings. Spatially relative terms should be understood as terms that include different directions of elements relative to each other during use or operation in addition to the directions shown in the drawings. For example, when an element shown in the drawing is turned over, an element described as "below" or "beneath" another element can be placed "above" the other element. Therefore, the exemplary term "below" can include all directions of below and above. The element may be oriented in other directions, and thus spatially relative terms can be interpreted according to the orientation.

[0015] First, second, etc. are used to describe various elements, components, and / or sections, but of course these elements, components, and / or sections are not limited by these terms. These terms are merely used to distinguish one element, component, or section from another element, component, or section. Therefore, it goes without saying that the first element, the first component, or the first section referred to below can be the second element, the second component, or the second section within the technical idea of the present invention.

[0016] Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings. When describing with reference to the accompanying drawings, the same or corresponding components are given the same reference numerals regardless of the reference signs in the drawings, and duplicate descriptions thereof are omitted.

[0017] FIG. 1 is a conceptual diagram for explaining a transport system according to some embodiments of the present invention. FIG. 2 is a conceptual plan view for explaining a transport system according to some embodiments of the present invention. FIG. 3 is a diagram for explaining the transport vehicle of FIG. 1.

[0018] First, referring to FIG. 1, a transport system according to some embodiments of the present invention includes a host controller (OCS, OHT control system, 10) and a plurality of transport vehicles 20.

[0019] The transport vehicle 20 can be, but is not limited to, an OHT (Overhead Hoist Transport) that travels along a rail installed on the ceiling of a semiconductor manufacturing factory (i.e., a fab (FAB)).

[0020] The host controller 10 communicates with the plurality of transport vehicles 20, for example, by a wireless communication method, and controls the plurality of transport vehicles 20.

[0021] The host controller 10 determines the destination of the transport vehicle 20, calculates the cost between bases to reach the destination, and determines the route to the destination based on the calculated cost. The route search method will be specifically described with reference to FIGS. 5 to 7. The host controller 10 provides a transport command to each transport vehicle 20 according to the determined route. The transport command can include information such as the start position, arrival position, and inventory of the transport target, for example.

[0022] Here, referring to FIG. 2, in the transport system according to some embodiments of the present invention, the rail 110 can have a ceiling installed and can have a configuration in which a straight line and a curve are combined. The rail 110 can include, but is not limited to, a main passage 112 and a plurality of bay portions 111 branched from the main passage 112.

[0023] As shown in the figure, the main passage 112 can have a closed curve structure, but is not limited thereto. A plurality of bay portions 111 can be arranged on one side centered on the main passage 112, and a plurality of bay portions 111 can be arranged on the other side.

[0024] In addition, an article storage unit 130 for storing articles (for example, FOUP, FOSB, etc.) can be arranged in the main passage 112. A plurality of semiconductor manufacturing equipments 30 can be arranged in the bay portion 111.

[0025] While moving along the rail 110, the plurality of carrier vehicles 20 can, for example, transport articles from the article storage unit 130 to the manufacturing equipment 30 or transport articles from the manufacturing equipment 30 to the article storage unit 130.

[0026] On the other hand, the rail 110 includes a power supply section where the power line 140 is installed and a non-power supply section where the power line 140 is not installed. In FIG. 1, exemplarily, the power line 140 is installed in the straight section of the main passage 112, and it is shown that the power line 140 is not installed in the curved section of the main passage 112 and the bay portion 111, but is not limited thereto.

[0027] Here, referring to FIG. 3, the carrier vehicle 20 includes a housing 210, a traveling module 220, a battery 230, and the like.

[0028] The traveling module 220 is installed on the housing 210 and is installed to be movable along the rail 110. The traveling module 220 includes traveling wheels 222 and a motor 224 for rotating the traveling wheels 222. The "traveling wheel" in this specification means at least one of a front traveling wheel and a rear traveling wheel, and the "motor" means at least one of a front motor for driving the front traveling wheel and a rear motor for driving the rear traveling wheel.

[0029] Inside the internal space of the housing 210, a hoist module for gripping and lifting an article can be installed. For example, the hoist module may include a hoist unit for lifting and lowering the article, a slide unit for moving the hoist unit in the left - right direction, and a hand unit connected to the hoist unit for gripping the article.

[0030] Also, a battery 230 is installed inside the housing 210. While the carrier 20 moves along the rail 110, it receives power supply from a power line (refer to 140 in FIG. 2). The supplied power can be used to drive the traveling module 220 or stored in the battery 230. When the carrier 20 moves along a rail 110 where the power line 140 is not installed, the traveling module 220 is operated using the energy stored in the battery 230.

[0031] According to some embodiments of the present invention, the upper controller 10 calculates the cost between bases to reach the destination in order to determine the route to the destination of the carrier 20. In particular, the input variables of the cost function for calculating the cost include not only the distance between bases, the traffic congestion degree between bases, but also the charge - discharge amount occurring between bases.

[0032] FIG. 4 is a block diagram for explaining the carrier of the conveying system according to some embodiments of the present invention.

[0033] Referring to FIG. 4, the carrier 20 includes an energy storage module 201, a drive module 229, a communication module 260, a power supply module 270, a power reception module 280, and a control module 290, etc.

[0034] The power receiving module 280 receives power supply from a power line (refer to 140 in FIG. 2) installed on the rail. For example, the power receiving module 280 receives power transmission from the power line 140 in a non-contact manner. Examples of the non-contact method may include, but are not limited to, HID (High efficiency Inductive power Distribution technology) or CPS (Contactless Power Supply technology).

[0035] The power supply module 270 provides the power received from the power receiving module 280 to the drive module 229. The drive module 229 includes at least one module that requires power within the carrier vehicle. For example, the drive module 229 may be a traveling module (refer to 220 in FIG. 3), a hoist module, or various sensors, but is not limited thereto.

[0036] In addition, the power supply module 270 can transmit power to the energy storage module 201. The power supply module 270 can selectively supply a static current or a variable current to the energy storage module 201. Also, the power supply module 270 can selectively adjust the supply and non-supply of power to the energy storage module 201.

[0037] The battery 230 can provide the stored energy to the drive module 229 via the power supply module 270 when the carrier vehicle 20 operates in a non-powered section or when an output above a certain level is required.

[0038] The battery 230 can be managed by a BMS (Battery Management System, 240). The BMS 240 can measure the remaining energy of the battery 230 and perform charging and discharging. Also, the BMS 240 can measure the energy efficiency and internal resistance of the battery 230, and based on this, can determine the lifespan of the battery 230.

[0039] The control module 290 can manage and control the power operation state and the running state of the carrier vehicle 20. Although the control module 290 generally controls the carrier vehicle 20, for example, it can control the power reception module 280, the power supply module 270, the drive module 229, the communication module 260, and the energy storage module 201.

[0040] The communication module 260 can be for communicating with a plurality of objects. For example, it can include a communication module for communicating with a host controller (refer to 10 in FIG. 1), a communication module for organizing the internal state of the carrier vehicle 20 and reporting it to a diagnostic server, and a communication module for communicating with other carrier vehicles.

[0041] FIG. 5 is a block diagram for explaining a host controller OCS of a transport system according to some embodiments of the present invention. FIG. 6 is a conceptual diagram for explaining a path search method.

[0042] First, referring to FIG. 5, the host controller 10 includes a processor 310, a communicator 360, and a memory 370.

[0043] The processor 310 communicates with an external object via the communicator 360 and calculates a cost using the data stored in the memory 370.

[0044] The processor 310 sets a destination for a carrier vehicle using a battery, calculates a cost between bases for reaching the destination, and determines a path to the destination based on the calculated cost.

[0045] The data necessary for the processor 310 to perform route search is stored in the memory 370. The data can include, for example, the distances between bases and the amount of charge and discharge between bases (i.e., the average value of the charge and discharge amount). Further, the memory 370 can further include a cost function for determining the cost between bases, a weighting value of the cost function, and the like. Further, the memory 370 can store instructions for the processor to perform route search operations.

[0046] The memory 370 can include a volatile memory (e.g., DRAM, SRAM, or SDRAM) and / or a non-volatile memory (e.g., OTPROM (one time programmable ROM), PROM, EPROM, EEPROM, mask ROM, flash ROM, flash memory, PRAM, RRAM, MRAM, hard drive, or solid state drive (SSD)). The memory 370 can include, but is not limited to, an internal memory and / or an external memory.

[0047] Note that the processor 310 can functionally include a destination setting unit 312, a cost calculation unit 314, and a route determination unit 316.

[0048] The destination setting unit 312 sets the destination of the carrier 20 that uses the battery 230. Specifically, for example, in FIG. 6, it is assumed that the start position of the operation of the carrier 20 that uses the battery 230 is node A and the arrival position (destination) is node I.

[0049] The cost calculation unit 314 calculates the cost between bases for reaching the destination.

[0050] Specifically, in FIG. 6, it is assumed that the bases for reaching from node A to node I are nodes B, C, D, E, F, G, and H. The edges between nodes A, B, C, D, E, F, G, H, and I are indicated by lines.

[0051] Calculate the cost between adjacent bases. The adjacent bases can be A-B, A-C, A-E, B-D, B-F, B-G, B-E, C-E, C-H, D-F, E-F, E-G, H-G, F-I, G-I. Here, "A-B" means the edge between node A and node B.

[0052] For example, the cost function for calculating the cost includes the charge and discharge amount that occurs between adjacent bases. The cost function further includes the distance between the bases and the degree of congestion between the bases.

[0053] The reason why the "charge and discharge amount" is included in the cost function is as follows. The cost function is for searching for a path for the carrier 20 that uses the battery 230. Also, since the rail 110 includes a power supply section and a non-power supply section, if the charge and discharge amount is not considered, the consumption of the battery 230 of the carrier 20 may be accelerated.

[0054] Specifically, the cost function cost(A,B) between node A and node B is as shown in Equation 1. The cost function cost(A,B) means the cost that occurs on the edge between node A and node B.

[0055] cost(A,B)=α×(distance)+β×(degree of congestion)+γ×(charge and discharge amount) …(Equation 1) α is the weighting value for the distance variable, β is the weighting value for the degree of congestion variable, and γ is the weighting value for the charge and discharge amount variable.

[0056] (distance) in Equation 1 is the distance between node A and node B, and (degree of congestion) is proportional to the number of carriers located between node A and node B. (Charge and discharge amount) represents the average of the total charge amount and discharge amount performed between node A and node B.

[0057] As described above, the distance between node A and node B, the average of the total charge amount and discharge amount performed between node A and node B, etc. are stored in the memory 370. Therefore, the processor 310 can receive the provision of (distance) and (degree of congestion) in Equation 1 from the memory 370.

[0058] The processor 310 receives the provision of the current positions from a plurality of carrier vehicles 20 via the communicator 360. Accordingly, the processor 310 can know how many carrier vehicles are there between node A and node B. Thereby, the processor 310 can calculate the number 1 (degree of congestion).

[0059] Also, the weighted values α, β, γ of the cost function can be determined by a machine learning method for maintaining the remaining amount (SoC, State of Charge) of the battery of the carrier vehicle 20 as a target remaining amount. Alternatively, the weighted values α, β, γ of the cost function can be determined by a machine learning method for minimizing the transport time of the carrier vehicle and maximizing the remaining amount of the carrier vehicle 20. A method for generating learning data for performing machine learning will be specifically described later with reference to FIGS. 8 and 9.

[0060] Similar to the method using the cost function cost(A,B), costs are also calculated for each of the other adjacent bases, A-C, A-E, B-D, B-F, B-G, B-E, C-E, C-H, D-F, E-F, E-G, H-G, F-I, G-I respectively.

[0061] The route determination unit 316 determines a route to the destination based on the calculated cost.

[0062] Specifically, as a method for determining a route to the destination, for example, Dijkstra's algorithm, Warshall-Floyd algorithm, Bellman-Ford algorithm, etc. are used.

[0063] Dijkstra's algorithm is used when finding the shortest path from a certain node to another specific node. On the contrary, the Warshall-Floyd algorithm is used when finding all the shortest paths from all nodes to all other nodes.

[0064] Dijkstra's algorithm selects the node with the shortest distance among the unvisited nodes at each step to find the shortest distance. On the contrary, the Bellman - Ford algorithm finds the shortest distance between all nodes while checking all edges at each step. Different from Dijkstra's algorithm, the Bellman - Ford algorithm can find the shortest path even if there are edges with negative costs.

[0065] The route determination unit 316 uses such an algorithm to determine the route to the destination based on the calculated costs between the bases.

[0066] The upper controller 10 provides the carrier vehicle 20 with a conveyance command corresponding to the route determined by the communicator 360.

[0067] FIG. 7 is a flowchart for explaining a route search method according to some embodiments of the present invention. For convenience of explanation, the description will focus on the points different from the content described with reference to FIGS. 1 to 6.

[0068] Referring to FIGS. 6 and 7, the destination of the carrier vehicle 20 using the battery 230 is set (S510).

[0069] For example, the starting position of the carrier vehicle 20 using the battery 230 by the upper controller 10 is node A, and the arrival position (destination) is set to node I.

[0070] Next, the costs between the bases for reaching the destination are calculated (S520).

[0071] For example, when the bases for reaching Node I from Node A are Nodes B, C, D, E, F, G, and H, the upper controller 10 calculates the costs between adjacent bases (A - B, A - C, A - E, B - D, B - F, B - G, B - E, C - E, C - H, D - F, E - F, E - G, H - G, F - I, G - I). The cost function for calculating the cost includes the charge and discharge amounts that occur between adjacent bases. Further, the cost function further includes the distance between the bases and the degree of congestion between the bases.

[0072] Next, a route to the destination is determined based on the calculated cost (S530).

[0073] The upper controller 10 determines a route to the destination using an algorithm such as, for example, Dijkstra's algorithm, the Floyd - Warshall algorithm, or the Bellman - Ford algorithm.

[0074] Hereinafter, a method for machine - learning a cost function will be described with reference to FIGS. 8 to 10.

[0075] FIG. 8 is a flowchart for explaining one method of accumulating learning data for machine - learning a cost function. A method of acquiring learning data on the field line will be described using FIG. 8.

[0076] Referring to FIG. 8, the operator changes the weighting values (S610).

[0077] Specifically, the operator arbitrarily determines the weighting values α, β, γ of the cost function based on empirical values or the like. For example, the operator determines (α, β, γ) to be (α1, β1, γ1). That is, the cost function is determined as cost(A,B)=α1×(distance)+β1×(congestion degree)+γ1(charge and discharge amount).

[0078] Next, the field line is operated according to the changed weighting values (S620).

[0079] Specifically, the cost between bases for reaching the destination is calculated using the determined cost function. Based on the calculated cost, a path to the destination is determined using a predetermined algorithm (for example, Dijkstra's algorithm).

[0080] Here, operating the on-site line can also mean running the transport vehicle on the on-site line to produce an actual semiconductor device, or intentionally running the transport vehicle on the on-site line to obtain learning data.

[0081] Next, the results of operating the on-site line are accumulated as learning data (S630).

[0082] The results of operating the on-site line are accumulated as learning data based on the weighting values (α1, β1, γ1) determined by the operator.

[0083] Again, return to the S610 step, where the operator changes the weighting values to (α2, β2, γ2), and operates the on-site line according to the changed weighting values to accumulate learning data.

[0084] In this way, the operator can accumulate learning data while changing the weighting values N times (where N is a natural number of 2 or more). The learning data can include, for example, weighting values, path data, machine numbers, departure SOC, arrival SOC, etc. Path data means the section where the transport vehicle 20 moves (for example, moving from node A to node B). The machine number means the ID of the transport vehicle 20. The departure SOC means the remaining battery level at the start position of the transport vehicle 20 (for example, node A). The arrival SOC means the remaining battery level at the arrival position of the transport vehicle 20 (for example, node B). If the departure SOC is greater than the arrival SOC, it means that more discharging than charging is performed while moving along the path (that is, while moving from node A to node B). Also, if the arrival SOC is greater than the departure SOC, it means that more charging than discharging is performed while moving along the path (that is, while moving from node A to node B).

[0085] The learning data can be as shown in Table 1 below, but is not limited thereto. For example, for the first data, the weighting values (α, β, γ) are (α1, β1, γ1), and when the carrier 001 moves from node A to node B, the starting SOC is 90% and the arrival SOC is 70%. The Nth data indicates that the weighting values (α, β, γ) are (α10, β10, γ10), and when 099 moves from node A to node I for conveyance, the starting SOC is 80% and the arrival SOC is 75%.

[0086] [Table 1]

[0087] FIG. 9 is a flowchart for explaining another method of accumulating learning data for machine learning of a cost function. Referring to FIG. 9, a method of obtaining learning data by simulation not on the field line will be described. For convenience of explanation, the description will focus on the content not described with reference to FIG. 8.

[0088] Referring to FIG. 9, the weighting values are randomly changed (S612).

[0089] As described with reference to FIG. 8, when obtaining learning data on the field line, since the operator mainly changes the weighting values reflecting empirical values, the weighting values are not set to various values. On the contrary, if simulation is used, the weighting values can be set to various values that are difficult for the operator to predict. A cost function is determined according to the randomly determined weighting values.

[0090] Next, a simulation is performed according to the changed weighting values (S622).

[0091] Specifically, a cost function is determined according to the changed weighting values, and the cost between bases for reaching the destination is calculated using the determined cost function. Based on the calculated cost, a route to the destination is determined using a predetermined algorithm. A simulation of moving the carrier along the determined route is performed.

[0092] Next, the simulation results are accumulated as learning data (S630).

[0093] Again, return to S612, randomly change the weight values, perform simulations according to the changed weight values, and accumulate learning data. Repeat such a process N times. The learning data accumulated in such a manner can be in the form as shown in Table 1.

[0094] FIG. 10 is a flowchart for explaining a machine learning method using learning data.

[0095] Referring to FIG. 10, start machine learning (S640).

[0096] Specifically, start machine learning using the learning data accumulated with reference to FIGS. 8 and 9. That is, start machine learning for determining the weight values α, β, γ in the cost function.

[0097] As the machine learning algorithm, supervised learning can be used, but it is not limited thereto. In some cases, unsupervised learning or semi-supervised learning may be used.

[0098] The result of machine learning may vary depending on the purpose of machine learning.

[0099] For example, the cost function can be optimized so as to optimize in the direction of "maintaining the battery remaining amount as the target remaining amount" (Yes at the S651 stage). The target remaining amount can be a default value or a value received by an operator.

[0100] Alternatively, the cost function can be optimized so as to optimize in the direction of "minimizing the conveyance time and maximizing the battery remaining amount" (Yes at the S652 stage).

[0101] For example, an iterative method is used to optimize the cost function, but it is not limited to this.

[0102] In this way, machine learning is performed to determine the optimal weighting value (S650).

[0103] The embodiments of the present invention have been described with reference to the above and the accompanying drawings. However, those skilled in the art to which the present invention pertains can understand that the present invention can be implemented in other specific forms without changing its technical idea and essential features. Therefore, it should be understood that the above-described embodiment is illustrative in all aspects and not restrictive.

Explanation of Reference Numerals

[0104] 10 Upper controller 20 Carrier vehicle 110 Rail 111 Bay section 112 Main passage 201 Energy storage module 210 Housing 220 Travel module 229 Drive module 230 Battery 260 Communication module 270 Power supply module 280 Power reception module 290 Control module 310 Processor 312 Destination setting unit 314 Cost calculation unit 316 Route determination unit 360 Communicator 370 Memory

Claims

1. Set the destination of the battery-powered transport vehicle, A cost between the base stations for reaching the destination is calculated, and an input variable of a cost function for calculating the cost includes a charge / discharge amount occurring between the base stations; determining a route to the destination based on the calculated cost.

2. The route search method according to claim 1 , wherein input variables of the cost function further include a distance between base stations and a congestion degree between the base stations.

3. The cost function between the first site (A) and the second site (B) is cost(A,B)=α×(distance)+β×(congestion level)+γ×(charge / discharge amount), 3. The route search method according to claim 2, wherein the weight α is a weight for a distance variable, the weight β is a weight for a congestion degree variable, and the weight γ is a weight for a charge / discharge amount variable.

4. 4. The method of claim 3, wherein the weights of the cost function are determined by a machine learning method to maintain a state of charge (SoC) of the vehicle's battery as a target state of charge.

5. The method of claim 3 , wherein the weights of the cost function are determined by a machine learning method that minimizes a travel time of the vehicle and maximizes a remaining amount of the vehicle.

6. The method of claim 1 , wherein the learning data for learning the cost function includes weights input by an operator when operating a production line.

7. The route search method according to claim 1 , wherein the learning data for learning the cost function includes weights that are randomly changed by simulation.

8. The transport vehicle moves along a rail, The route search method according to claim 1 , wherein the rail includes a powered section in which a power line for supplying electric power is installed, and a non-powered section in which the power line is not installed.

9. a destination setting unit that sets a destination of a transport vehicle that uses a battery; A cost calculation unit that calculates a cost between base stations for reaching the destination, and an input variable of a cost function for calculating the cost includes a charge / discharge amount occurring between the base stations; and and a route determination unit that determines a route to the destination based on the calculated cost.

10. The route search device according to claim 9 , wherein input variables of the cost function further include a distance between bases and a congestion degree between the bases.

11. The cost function between the first site (A) and the second site (B) is cost(A,B)=α×(distance)+β×(congestion level)+γ×(charge / discharge amount), The route search device according to claim 10, wherein the weight α is a weight for a distance variable, the weight β is a weight for a congestion degree variable, and the weight γ is a weight for a charge / discharge amount variable.

12. The path search device according to claim 11 , wherein the weights of the cost function are determined by a machine learning method to maintain a remaining battery charge of the vehicle as a target remaining charge.

13. The route search device according to claim 11 , wherein the weights of the cost function are determined by a machine learning method that minimizes a travel time of the vehicle and maximizes a remaining amount of the vehicle.

14. The path search device according to claim 9 , wherein the learning data for learning the cost function includes weights input by an operator when operating a production line, or weights randomly changed by a simulation.

15. a rail including a power supply section in which a power line for supplying electric power is installed and a non-power supply section in which the power line is not installed; a transport vehicle that moves along the rail and has a battery that stores power supplied through the power line; and A host controller that controls the plurality of transport vehicles, The upper controller is Set a destination of the transport vehicle; A cost between the base stations for reaching the destination is calculated, and an input variable of a cost function for calculating the cost includes a charge / discharge amount occurring between the base stations; A transportation system that determines a route to the destination based on the calculated cost.

16. The transportation system according to claim 15 , wherein the input variables of the cost function further include a distance between base stations and a congestion degree between the base stations.

17. The cost function between the first site (A) and the second site (B) is cost(A,B)=α×(distance)+β×(congestion level)+γ×(charge / discharge amount), The transportation system according to claim 16, wherein the weight α is a weight for a distance variable, the weight β is a weight for a congestion degree variable, and the weight γ is a weight for a charge / discharge amount variable.

18. 20. The transportation system of claim 17, wherein the weights of the cost function are determined by a machine learning method to maintain a battery charge of the transportation vehicle at a target charge.

19. 20. The transportation system of claim 17, wherein the cost function weights are determined by a machine learning method that minimizes the transportation time of the vehicle and maximizes the remaining capacity of the vehicle.

20. The conveyance system according to claim 15 , wherein the learning data for learning the cost function includes weighting values ​​input by an operator when operating a line, or weighting values ​​randomly changed by simulation.

Images