JP6695653B2 - Transfer planning method, transfer planning device, transfer system, computer program - Google Patents

Description

  The present invention relates to a method for formulating a transport plan from each transport start position to a target position when transporting an object to be transported from a plurality of transport start positions to a target position, an apparatus for formulating such a transport plan, and The present invention relates to a transport system that includes a transport plan formulation device and that transports an object to be transported from a plurality of transport start positions to target positions based on the formulated transport plan, and a computer program that causes a computer to perform the formulation of the transport plan.

  In a physical distribution warehouse functioning as a physical distribution base, there is a demand for flexible and efficient rearrangement of the plurality of conveyed objects to be stored and leaving them. In order to meet this demand, it is possible to introduce multiple transporters that can autonomously move and transport the transported objects from the multiple transport start positions in the distribution warehouse to the target position that is the rearrangement destination by each transporter. Although it has been studied, it is important to formulate a highly efficient transportation plan in a short time because each transportation start position and destination position change in such a distribution warehouse. Therefore, Patent Document 1 described below discloses the following method.

  That is, a plurality of nodes are defined in advance between each transport start position and the target position, and a plurality of transport route candidates connecting adjacent nodes are created to obtain an initial solution set. At this time, a transport route candidate is created by giving priority to a node having no time constraint over a node having a time constraint. Then, a genetic algorithm is applied to each of the obtained transport route candidates of the initial solution group and the improvement is repeated to create a transport route satisfying the time constraint and having a shorter total distance.

JP, 2003-157312, A

  However, in such a conventional method, it is possible to cope with a case where a so-called traffic jam occurs, in which a plurality of transport machines are overlapped between two common nodes on each created transport route. Therefore, the transportation efficiency was lowered as a whole.

  The present invention has been made in view of such circumstances, and even when congestion occurs between a plurality of transporters on a transport path, each transporter as a whole transports an object to be transported with higher efficiency. A method of formulating a transport plan that can be performed, an apparatus that formulates such a transport plan, and a transport machine including a plurality of transport plan formulation devices from a plurality of transport start positions to transport target positions based on the formulated transport plan Provides a transport system for transporting an object to be transported, and a computer program for causing a computer to execute the formulation of a transport plan.

(1) In the transport planning method according to the present invention, the objects to be transported respectively arranged at a plurality of different transport start positions on a plurality of transport paths are transported to respective transport target positions set on any one of the transport paths. When formulating a plan for transporting separately by multiple transporters that can move on the road, assign each transporter to the object to be transported, and for each transporter, from the transport start position to the transport target position. The optimum transport route is generated based on the distance to, and one or more suboptimal transports different from the optimum transport route are generated based on the direction of the generated optimal transport route and the value indicating the ease of transport on the transport route. A route is generated to obtain a plurality of transport route groups for each transport machine including the optimal transport route and one or more suboptimal transport routes, and an optimal transport route included in each of the transport route groups for each transport machine and Whenever you select any of the sub-optimal transport routes and select any of the transport routes, you can select any waiting time from a plurality of preset waiting times to make the transport machine stand by at an intermediate position on the transport route. Select, set the selected waiting time to one or more midway positions on the transport route, and repeat the operation of generating transport plan candidates as a transport plan candidate group. Perform these operations multiple times and wait. Obtain multiple transport plan candidate groups with different time settings, determine whether or not multiple transport machines exist on the same transport route at the same time for each of the obtained transport plan candidate groups , and transport from the transport start position. to the target position based on the time required for conveying the conveying object to calculate a goodness of fit, to compare the value of each goodness of fit is calculated, based on the comparison result, optimal from the plurality of conveying plan candidates It is characterized in that a proper transport plan candidate group is selected, and each transport plan candidate that constitutes the selected transport plan candidate group is used as a transport plan.

(2) Further, in the transportation plan formulation method according to the present invention, nodes are set at a plurality of positions on each of the transportation routes, and weights according to the distance between both nodes are previously set between adjacent nodes. By setting these weights and applying these weights to the equations (1) and (2) described later, the ease of transport to the transport target position by passing between adjacent nodes in the order of separating the position from the transport target position. When calculating the Q value indicating each of the above and generating the sub-optimal transport route, the first node is selected by selecting an arbitrary node from one or a plurality of nodes adjacent to the candidate node that should constitute the suboptimal transport route. Then, it is determined whether or not the direction from the candidate node to the first node is the same as the direction of the corresponding optimum transport route, and if not, the Q between the candidate node and the first node is determined. A value is specified, an arbitrary node is designated from one or a plurality of nodes adjacent to the first node as a second node, and a Q value between the first node and the second node is specified. If the latter Q value is smaller than the former Q value, the first node is set as a candidate node.

  (3) Furthermore, in the transportation planning method according to the present invention, the second node selects the first node before specifying the Q value between the first node and the second node. If it is at the same position as the candidate node at that time, and if it is at the same position, it is adjacent to the first node without specifying the Q value between the first node and the second node. It is characterized in that other nodes are specified.

  (4) On the other hand, in the transportation plan formulation method according to the present invention, when the direction from the candidate node to the first node is the same as the direction of the corresponding optimum transportation route, the candidate node and the first node are The first node is set as a candidate node without specifying a Q value between them.

(5) Further, in the transportation plan formulation method according to the present invention, when selecting an optimal transportation plan candidate group, each obtained transportation plan candidate group is ranked in accordance with the value of its suitability and is determined in advance. For each transport plan candidate group having a lower rank than the predetermined order, the transport routes included in each transport plan candidate group are crossed with each other up to a preset number of crossover frequencies to generate a new transport plan candidate group. After generating and / or selecting an arbitrary transfer plan candidate from each transfer plan candidate group and updating each transfer route of each transfer plan candidate group to each transfer route included in the selected transfer plan candidate, The waiting time set on the transfer route is updated with a predetermined mutation probability to generate new transfer plan candidate groups, and each transfer plan candidate group having a higher rank than a predetermined order and each new generated transfer plan. It is characterized in that the operation of ranking the candidate groups in accordance with the value of the suitability is repeated a predetermined number of times, and then the transport plan candidate group having a higher rank is set as the optimum transport plan candidate group.

(6) By the way, the transfer plan formulation method according to the present invention is characterized in that the compatibility is calculated by using the equation (3) described later .

(7) The transfer plan formulation device according to the present invention transfers an object to be transferred, which is respectively arranged at a plurality of different transfer start positions on a plurality of transfer paths, to a transfer destination position set on any one of the transfer paths. When formulating a plan for transporting separately by multiple transporters that can move on the road, a method for allocating each transporter to an object to be transported and transporting each transporter from the transport start position Based on the distance to the target position, an optimum transport route is generated, and one or more different from the optimum transport route based on the generated direction of the optimum transport route and the value indicating the ease of transport on the transport route. generates a suboptimal transport path, means for obtaining a plurality of conveyor by conveying path group including the optimal transport path and one or more sub-optimal transporting path, from each conveyor by conveying path group, there Conveyance route selection means for selecting an arbitrary conveyance route out of the optimum conveyance route and the suboptimal conveyance route included, and every time the arbitrary conveyance route is selected , the conveyance device is preliminarily set to stand by at an intermediate position on the conveyance route. Selecting a waiting time from a plurality of defined waiting times, setting the selected waiting time to one or a plurality of midway positions on the carrying route, and creating a carrying plan candidate generating unit, Each time the transport route selecting means selects an arbitrary transport route, the transport plan candidate generating means repeats the operation of generating transport plan candidates to form a transport plan candidate group, and these operations are performed a plurality of times to set the waiting time. Of each transfer plan candidate group different from each other, and whether or not a plurality of transfer machines exist on the same transfer route at the same time for each of the obtained transfer plan candidate groups , and transfer from the start position to the transfer target position compared with fitness calculating means for calculating a goodness of fit based on the time required for conveying the conveying object, the value of each goodness of fit is calculated, based on the comparison result, wherein An optimum transfer plan candidate group selecting means for selecting an optimum transfer plan candidate group from a plurality of transfer plan candidate groups, and a means for setting each transfer plan candidate constituting each selected transfer plan candidate group as a transfer plan. It is characterized by being provided.

(8) Further, in the transfer plan formulation device according to the present invention, nodes are set at a plurality of positions on each of the transfer routes, and weights corresponding to the distances between both adjacent nodes are set in advance. The weights are set , and these weights are applied to preset equations (1) and (2), which will be described later , to transfer to the transfer target position by passing between adjacent nodes in order of separating the position from the transfer target position. The sub-optimal transport route generating means is provided with a unit for calculating a Q value indicating ease of operation, and the sub-optimal transport route generation unit selects an arbitrary node from one or a plurality of nodes adjacent to the candidate node that should form the sub-optimal transport route. The means for selecting the first node and the direction determining means for determining whether or not the direction from the candidate node to the first node is the same as the direction of the corresponding optimum transport route, and the direction is not the same. In this case, the current node Q value specifying means for specifying the Q value between the candidate node and the first node and the second node by specifying an arbitrary node from one or a plurality of nodes adjacent to the first node Node specifying means to be a node, specified node Q value specifying means for specifying the Q value between the first node and the second node, and the latter Q value is smaller than the former Q value, And a candidate node determining unit that sets one node as a candidate node.

  (9) In the transfer plan formulation device according to the present invention, further, before the designated node Q value identifying means identifies the Q value between the first node and the second node, A means for determining whether or not the node is at the same position as the candidate node when the first node is selected is provided, and when the node is at the same position, the designated node Q value specifying means is provided for the first node and the second node. It is characterized in that the node designating means designates another node adjacent to the first node without specifying a Q value with the node.

  (10) On the other hand, when the direction determining means determines that the direction from the candidate node to the first node is the same as the direction of the corresponding optimum transfer route, the transfer plan formulation device according to the present invention is configured to: It is characterized in that the node Q value specifying means does not specify the Q value between the candidate node and the first node, and the candidate node determining means makes the first node the candidate node. ..

(11) By the way, in the transportation plan formulation device according to the present invention, the optimum transportation plan candidate group selecting means is a ranking means for ranking each of the obtained transportation plan candidate groups in accordance with the value of the compatibility. , For each transport plan candidate group having a lower rank than a predetermined rank, each transport route included in each transport plan candidate group is crossed with each other up to a preset number of crossover frequencies to obtain a new transport plan candidate. Each group is generated and / or an arbitrary transfer plan candidate is selected from each transfer plan candidate group, and each transfer route of each transfer plan candidate group is updated to each transfer route included in the selected transfer plan candidate. After that, the waiting time set for each transfer route is updated with a predetermined mutation probability, and a new transfer plan candidate group generation unit for respectively generating a new transfer plan candidate group is provided. The transfer plan candidate groups having a higher order than a predetermined order and the generated new transfer plan candidate groups are ranked according to the value of the degree of conformity, and the new transfer plan candidate group generation means is new. A means for generating the transport plan candidate groups respectively and repeating the operation of ranking by the ranking means a predetermined number of times, and then making the transport plan candidate group of a higher rank the optimum transport plan candidate group is provided. It is characterized by

(12) Further, in the transportation plan formulation device according to the present invention, the suitability calculating means is adapted to calculate the suitability based on a preset formula (3) described later. ..

  (13) A transfer system according to the present invention includes a plurality of transfer paths, a plurality of transfer devices that can move on the transfer paths, and the transfer plan formulation device according to any one of (7) to (12) above. And a transmitter for transmitting the transportation plan created by the transportation plan formulation apparatus to each of the transportation machines.

(14) The computer program according to the present invention causes the computer to transfer the objects to be conveyed, which are respectively arranged at a plurality of different conveyance start positions on the plurality of conveyance paths, to the conveyance target positions set on any one of the conveyance paths. A program for making a plan for individually carrying by a plurality of carriers that can move on a carrier path, a step of causing a computer to allocate each carrier to an object to be transported, and a program for each carrier Respectively, based on the step of generating an optimal transport route based on the distance from the transport start position to the transport target position, and the value indicating the direction of the generated optimal transport route and the ease of transport in the transport route , and suboptimal transport path generating step of generating a different one or more suboptimal transport path optimal transport route, the optimum transport path and one or more conveyor by conveying path group including a plurality of sub-optimal conveyance path generated And a transport route selecting step of selecting an optimal transport route and an optimal transport route included therein from each transport route group for each transport machine, and an optional transport route being selected. Each time, a desired waiting time is selected from a plurality of predetermined waiting times to make the carrier stand by at a midway position on the transportation route, and the selected waiting time is set to one or more midway positions on the transportation route. And a transfer plan candidate generating step of generating a transfer plan candidate, and the transfer plan candidate generating step repeats the operation of generating a transfer plan candidate each time an arbitrary transfer route is selected in the transfer route selecting step. Then, these operations are performed as a transfer plan candidate group, and these operations are performed a plurality of times to obtain a plurality of transfer plan candidate groups with different standby time settings. Regarding, regarding, whether or not there are multiple transporters on the same transport route at the same time, and the fitness calculation that calculates the fitness based on the time required to transport the transported object from the transport start position to the transport target position. Step, comparing the calculated values of each goodness of fit , based on the comparison result, an optimal transport plan candidate group selection step for selecting an optimal transport plan candidate group from the plurality of transport plan candidate groups , And a step of causing each of the transfer plan candidates forming the selected transfer plan candidate group to be a transfer plan, and causing the computer to formulate the transfer plan.

(15) Also, in the computer program according to the present invention, nodes are set at a plurality of positions on each of the transport paths, and weights corresponding to the distances between the adjacent nodes are set in advance between adjacent nodes. Therefore, by applying these weights to equations (1) and (2) described below , the computer can easily carry to the carrying destination position by passing between adjacent nodes in the order of separating the positions from the carrying destination position. And a step of calculating a Q value indicating each of the values, and the suboptimal transport route generation step causes an arbitrary node to be selected from one or a plurality of nodes adjacent to the candidate node which should form the suboptimal transport route. If the direction is not the same, the step of making the first node and the direction determining step of determining whether or not the direction from the candidate node to the first node is the same as the direction of the corresponding optimum transport route, A current node Q value specifying step of specifying a Q value between the node and the first node; and an arbitrary node is specified from one or a plurality of nodes adjacent to the first node to be a second node A node specifying step, a specified node Q value specifying step for specifying a Q value between the first node and the second node, and a first node if the latter Q value is smaller than the former Q value And a candidate node determining step for making the candidate node as a candidate node.

  (16) The computer program according to the present invention further causes the computer to specify the Q value between the first node and the second node in the specified node Q value specifying step. 2 node is determined to be at the same position as the candidate node when the first node is selected, and if it is at the same position, the designated node Q value specifying step is the same as the first node. It is characterized in that the node designating step allows the first node to designate another adjacent node without specifying the Q value with the second node.

  (17) On the other hand, the computer program according to the present invention causes the computer to determine in the direction determining step that the direction from the candidate node to the first node is the same as the direction of the corresponding optimum transport route. The present node Q value specifying step does not specify the Q value between the candidate node and the first node, and the candidate node determining step makes the first node the candidate node. Characterize.

(18) By the way, in the computer program according to the present invention, the optimal transportation plan candidate group selecting step causes the computer to rank each obtained transportation plan candidate group in accordance with the value of the matching degree. Then, for each transport plan candidate group having a lower rank than a predetermined rank, each transport route included in each transport plan candidate group is crossed with each other up to a preset number of crossover frequencies to create a new transport plan. A candidate group is generated respectively and / or an arbitrary transfer plan candidate is selected from each transfer plan candidate group, and each transfer route of each transfer plan candidate group is set to each transfer route included in the selected transfer plan candidate. After updating, the waiting time set for each transfer route is updated with a predetermined mutation probability, and a new transfer plan candidate group generation step for respectively generating new transfer plan candidate groups is provided. The step is arranged to rank each transport plan candidate group having a higher rank than a predetermined rank and each new generated transport plan candidate group in accordance with the value of the goodness of fit thereof. After the generation step generates new transport plan candidate groups and the operation of ranking by the ranking step is repeated a predetermined number of times, the transport plan candidate group of higher rank is set as the optimal transport plan candidate group. It is characterized by comprising steps.

(19) The computer program according to the present invention is characterized in that the adaptability calculating step causes the computer to calculate the adaptability based on an equation (3) described later .

  In the inventions of (1), (7), (13) and (14), the objects to be conveyed which are respectively arranged at a plurality of different conveyance start positions on the plurality of conveyance paths are set on any one of the conveyance paths. When formulating a plan to separately convey to the specified transportation target position by multiple transporters that can move on each transport path, assign each transporter to the object to be transported, and assign each transporter to each , An optimum transport path from the transport start position to the transport target position is generated. Here, the optimum transport route can be the shortest route from the transport start position to the transport target position.

  Further, one or more suboptimal transport routes are generated based on the generated optimal transport route. The suboptimal transport route is not the shortest route from the transport start position to the transport target position, but is a plurality of routes that can be set between the transport start position and the transport target position, for example, a reference such as a distance and corresponding. It can be determined by obtaining a route that is easier to pass through based on the direction of the optimum transport route and the like. In this way, by generating the optimum transport path and one or a plurality of suboptimal transport paths for each transport machine, a plurality of transport path groups for each transport machine is obtained.

  Next, each time the optimum transport route or the suboptimal transport route is selected from the transport route group for each transport machine, and each time the optimum transport route or the suboptimal transport route is selected, the optimum transport route or the suboptimal transport route is the transport route. The operation of setting the waiting time at one or more midway positions above and generating the transport plan candidates is repeated to form a transport plan candidate group. These operations are performed a plurality of times to obtain a plurality of transfer plan candidate groups. That is, one transfer plan candidate group is configured by each transfer plan candidate corresponding to each transfer device, and the transfer plan candidate is the optimum transfer route or the suboptimal transfer route corresponding to the transfer device and the optimum transfer route. The waiting time is set at one or more midway positions on the route or the suboptimal transport route.

  For each of the transfer plan candidate groups thus obtained, the degree of conformity to a predetermined standard is calculated, and the optimum degree of transfer plan candidate group is searched from each of the transfer plan candidate groups using the calculated degree of conformity. .. Then, each of the transportation plan candidates constituting the searched transportation plan candidate group is set as a transportation plan.

In this way, each transfer plan candidate that constitutes the transfer plan candidate group includes one or more suboptimal transfer routes in addition to the optimum transfer route, and further, the optimum transfer route and one or more midway positions on the suboptimal transfer route. Since the waiting time is set in and the optimal transfer plan candidate group is selected from the multiple transfer plan candidate groups, even if congestion occurs between multiple transfer machines on the transfer route, each transfer It is possible to formulate a transportation plan in which the machine as a whole can transport the transported object with higher efficiency. Here, if there are multiple transporters on the same transport route at the same time, the compatibility is low, and conversely, if there are multiple transporters on the same transport route at the same time, the compatibility is high. To do. Further, the shorter the time required to convey the object to be conveyed from the conveyance start position to the conveyance target position, the higher the adaptability. Therefore, even when congestion occurs between a plurality of transporters on the transport route, it is possible to formulate a transport plan in which each transporter can transport the transported object in a shorter time as a whole.

  In the inventions of (2), (8), (13), and (15), nodes are set at a plurality of positions on each of the above-described transport paths, and the transport machine passes through any of the nodes. It moves from the transport start position to the transport target position.

  Weights corresponding to the distances between both nodes are set in advance between the adjacent nodes, and these weights are used to determine the distance from the transport target position to the transport target position by passing between the adjacent nodes in the order of separating the positions. The Q value indicating the ease of transportation is calculated. As a result, the Q value corresponding to the transport target position is obtained. Using this Q value, the suboptimal transport path is generated as follows.

  An arbitrary node is selected from one or a plurality of nodes adjacent to the candidate node that should form the suboptimal transport route as the first node, and the direction from the candidate node to the first node is the corresponding optimum It is determined whether the direction is the same as that of the transport path. When the direction from the candidate node to the first node is different from the direction of the corresponding optimum transport route, the Q value between the candidate node and the first node is specified, and then the first node is adjacent to the first node. Alternatively, an arbitrary node is designated from a plurality of nodes to be the second node. Then, the Q value between the first node and the second node is specified, and when the latter Q value is smaller than the former Q value, the first node is set as the candidate node. This operation is repeatedly performed from the transport start position to the transport target position, and each candidate node is connected to generate the suboptimal transport route.

  When the direction from the candidate node to the first node is different from the direction of the corresponding optimum transport route, the transport route is a redundant route compared to the shortest route, but as described above, the candidate node and the first node When the Q value between the first node and the second node is smaller than the Q value between and, a more redundant route may be generated because the first node is the candidate node. This can be avoided, and thus a more appropriate suboptimal transport path can be efficiently generated.

  In the inventions of (3), (9), (13) and (16), as described above, before the Q value between the first node and the second node is specified, It is determined whether or not the node is at the same position as the candidate node when the first node is selected. That is, it is determined whether or not the route is from the candidate node to the first node and then back to the candidate node. Then, in the case of the same position, in order to discard such a route, another node adjacent to the first node is identified without specifying the Q value between the first node and the second node. It is specified. As a result, it is possible to avoid the generation of a low-efficiency suboptimal transport path that makes a round trip between the same nodes.

  In the inventions of (4), (10), (13), and (17), when the direction from the above-mentioned candidate node to the first node is the same as the direction of the corresponding optimum transport route, Since the route corresponding to the optimum transport route having high transport efficiency is generated, the first node is set as the candidate node without specifying the Q value between the candidate node and the first node. As a result, the suboptimal transport path can be efficiently generated.

  In the inventions of (5), (11), (13) and (18), the optimum transport plan candidate group is searched as follows. That is, the goodness of fit is obtained for each of the transfer plan candidate groups obtained as described above, and the transfer plan candidate groups are ranked according to the value of the obtained goodness of fit. Next, a genetic algorithm is applied to each transfer plan candidate group having a lower rank than a predetermined order to generate a new transfer plan candidate group. Genetic crossover and / or gene mutation can be used as the genetic algorithm.

  When a new transfer plan candidate group is generated, the goodness of fit is obtained for each of the transfer plan candidate groups having a higher rank than a predetermined order and each new generated transfer plan candidate group, and each transfer is performed according to the obtained value of the goodness of fit. Plan candidate groups. After repeating such an operation a predetermined number of times, the transport plan candidate group having a higher rank is set as the optimal transport plan candidate group.

  In this way, to generate a new transport plan candidate group by applying the genetic algorithm to each transport plan candidate group of a lower rank than a predetermined rank, and to re-rank with the high rank transport plan candidate group, It is possible to generate a transportation plan with higher efficiency relatively easily.

In the inventions of (6), (12), (13), and (19), the compatibility is calculated based on the equation (3) described later, and therefore, as before, there are a plurality of files on the same transport route at the same time. If there is a carrier, the compatibility is lowered, and conversely, if there are a plurality of carriers on the same transportation route at the same time, the compatibility is increased. Further, the shorter the time required to convey the object to be conveyed from the conveyance start position to the conveyance target position, the higher the adaptability. Therefore, even when congestion occurs between a plurality of transporters on the transport route, it is possible to formulate a transport plan in which each transporter can transport the transported object in a shorter time as a whole.

It is explanatory drawing explaining the structure of the conveyance system which concerns on this invention. It is a block diagram which shows the structure of the conveyance plan formulation apparatus shown in FIG. It is a block diagram which shows the structure of the conveyance machine shown in FIG. It is explanatory drawing explaining an example of the weight set between adjacent nodes. 6 is a table showing an intermediate result of repeatedly calculating a Q value using the transport path and the weight shown in FIG. 4. It is explanatory drawing explaining the generation method of a conveyance plan candidate. 3 is a flowchart showing a procedure for formulating a transportation plan by the transportation plan formulation unit shown in FIG. 2. It is a flow chart which shows the procedure of searching for the optimal transportation plan group. It is a flow chart which shows the procedure of searching for the optimal transportation plan group. It is a flow chart which shows the procedure of searching for the optimal transportation plan group. 9 is a flowchart showing a procedure for searching for an optimum transport route shown in FIG. 8. 9 is a flowchart showing a procedure for searching the suboptimal transport route shown in FIG. 8. 9 is a flowchart showing a procedure for searching the suboptimal transport route shown in FIG. 8. 9 is a flowchart showing a procedure for searching the suboptimal transport route shown in FIG. 8.

  Hereinafter, the present invention will be described in detail with reference to the drawings. It should be noted that the method and device, the transportation system, and the computer program relating to the transportation plan formulation described in the present embodiment are examples for explaining the gist of the present invention, and the present invention includes modifications within a range not departing from the gist thereof. Needless to say.

  FIG. 1 is an explanatory diagram for explaining the configuration of a transport system according to the present invention, in which W is a distribution warehouse. A plurality of transport machines 2, 2, ... That transport objects to be transported are installed in the distribution warehouse W, and a plurality of transport machines 2, 2, ... .. are provided in a grid pattern.

  In the present embodiment, one or a plurality of objects to be transferred are stored in a rack R formed by vertically stacking a plurality of storage shelves, and the transfer machine 2 stores the bottommost stage of the rack R. The height dimension is such that it can be inserted between the bottom of the shelf and the floor F.

  The carrier 2 is provided with a sensor for detecting an obstacle, a reading unit for reading position information, and a driving unit including a battery, a motor, driving wheels and casters, and a CPU for controlling these operations is built in. There is. The detection signal output from the sensor and the position information read by the reading unit are given to the CPU, and the CPU uses the given detection signal and position information to drive / stop the driving wheels via the motor. In addition to controlling the direction of the casters, the conveyor 2 can autonomously travel in a predetermined direction.

  On the other hand, the ceiling of the carrier 2 is capable of moving up and down to form a carrying part 22 for carrying the rack R, and the carrier 2 has entered between the bottom of the lowest storage rack of the rack R and the floor F. Sometimes, the carrying section 22 is raised to carry the rack R and the object to be transferred stored in the rack R, and the object to be transferred is transferred by moving in that state, and the movement is stopped at a predetermined transfer target position. Then, by lowering the carrying unit 22, the rack R and the transported object housed in the rack R are rearranged to the transportation target position.

  The position information such as a mark or an electromagnetic signal determined according to the position is provided to each of the nodes N, N, ... The position information imparting devices 41, 41, ... Which are imparted to the carrier device 2 are arranged, and the position information imparting devices 41, 41 ,. Each section is provided. As the reading unit, for example, when the position information given by the position information giving unit 41 is a mark, the above-mentioned image pickup device can be applied. In this case, the mark is identified from the image data obtained by imaging the mark set in the position information adder 41 by the image pickup device, and the position information corresponding to the mark is obtained from the position information set for each mark in advance. Make a decision. When the position information given by the position information giving device 41 is an electromagnetic signal, the carrier device 2 is provided with a reading unit corresponding thereto.

  A transmitter 50 is arranged at a high position in the distribution warehouse W, and from the transmitter 50, all the transportation plans including the identification information of the transportation machines 2, 2, ... It is designed to be transmitted to the carrier machines 2, 2, .... The transport units 2, 2, ... Are each provided with a receiving unit, and the transport units 2, 2, ... Transport the transported objects 2, 2, ... to the target position based on the transport plan received by the receiving unit. To do.

  On the other hand, in the transmitter 50, identification information of the carrier machines 2, 2, ... And the carrier machines 2, 2 ,. The transportation plan of is given.

In the present embodiment, the case where the rack R that stores the transported object and the transporter 2 that carries the rack R and transports the transported object stored in the rack R is described. The present invention is not limited to this, and may have any configuration as long as one or a plurality of objects to be stored can be carried by the carrier. Further, in the present embodiment, the carrier 2 capable of autonomous traveling is used, but a non- autonomous carrier guided to a predetermined position by guide means such as a rail or a signal may be used. Needless to say.

  FIG. 2 is a block diagram showing the configuration of the transportation plan formulation device 1 shown in FIG. As shown in FIG. 2, the transfer plan formulation device 1 includes an input unit 11 such as a keyboard and a mouse, a display unit 12 such as a monitor, a memory 13 for storing various information, and a transfer plan for each transfer machine 2, 2 ,. Is provided as described later, a delivery plan formulation unit 14, an output unit 15 that outputs the formulated transport plan to the transmitter 50 (see FIG. 1) described above, and a CPU 10 that controls these operations.

  In the memory 13 described above, the configurations of the respective transport paths 30, 30, ..., The positions of the respective nodes N, N, ..., and the adjacent nodes N, N, N, N ,. The warehouse information including the weight, the type of the transported object, the transported object information relating to the positions of the racks R, R, ... Each storing them, and the identification information of the transporting machines 2, 2 ,. The carrier information including it is stored via the input unit 11. The CPU 10 determines whether or not a transportation plan formulation command has been input from the input unit 11. When the transportation plan formulation command has been input, the CPU 10 causes the transport plan formulation unit 14 to perform the following operation for each of the transport machines 2, 2 ,. Have a transportation plan formulated.

  FIG. 7 is a flowchart showing a procedure for formulating a transportation plan by the transportation plan formulation unit 14 shown in FIG. The procedure for formulating such a transportation plan is pre-read as a computer program in the transportation plan formulation device 1.

  When the transport plan formulation command is given, the transport plan formulation unit 14 reads out the warehouse information, the transported object information, and the transport machine information from the memory 13 (step S1), and then uses these information to perform a plurality of processes as described later. An optimum transport plan group is searched from a plurality of transport plan candidate groups including transport routes to be assigned to the transport machines 2, 2, ... (See FIG. 1) and standby information including a standby position and standby time (step S2). ), The transfer plans constituting the transfer plan group obtained by the search are assigned to the corresponding transfer machines 2, 2, ... (Step S6). Then, the transfer plan formulation unit 14 gives the allocated transfer plan and the identification information of the corresponding transfer machines 2, 2, ... To the CPU 10 (step S7).

  When the identification information of the carrier machines 2, 2, ... And the carrier plan assigned to the carrier machines are given, the CPU 10 sends them from the transmitter 50 via the output unit 15 to each carrier machine 2, 2 ,. (See FIG. 1).

  FIG. 3 is a block diagram showing the configuration of the carrier machine 2 shown in FIG. As shown in FIG. 3, the carrier device 2 includes a receiver 21 that receives a signal transmitted from a transmitter 50 (see FIG. 1), a memory 25 that stores information, and a timer unit that measures a standby time and the like. 26, a sensor 23 for detecting an obstacle, a carrying section 22 for carrying the rack R and the transported object, a reading section 24 for reading positional information, and a driving section 27 including a battery, a motor, driving wheels and casters. And a CPU 20 that controls these operations. When the receiving unit 21 receives a signal from the transmitter 50, the CPU 20 gives a transportation plan corresponding to a preset identification signal to a predetermined area of the memory 25. And store it there. Then, the CPU 20 controls the drive unit 27 and the carrier unit 22 using the information from the time counting unit 26, the sensor 23, and the reading unit 24 so that the transport route, the standby position, and the standby time are in accordance with the transport plan. , The rack R and the transported object are transported and rearranged from the transport start position to the transport target position.

  In the present embodiment, the case where the sensor 23 that detects an obstacle and the reading unit 24 that reads the position information are separately provided has been described, but the present invention is not limited to this, and, for example, an image pickup device takes an image. The sensor 23 and the reading unit 24 may be combined so that the information obtained by processing the obtained image and the position information can be detected.

  Next, the procedure for searching for the optimum transport plan group in step S2 shown in FIG. 7 will be described in detail.

  8 to 10 are flowcharts showing the procedure for searching for the optimum transport plan group. As shown in FIG. 8, the transfer plan formulation unit 14 sends an image prompting the selection of one or more transfer machines 2, 2, ... , And waits until it is determined that the conveyors 2, 2, ... Are selected (steps S21, S22). When the transport machines 2, 2, ... For which a transport plan is to be formulated are selected via the input unit 11, the transport plan formulation unit 14 uses the identification information of the selected transport machines 2, 2, ... It is given to the area and stored therein (step S23). Next, the transportation plan formulation unit 14 causes the display unit 12 to display an image prompting the user to input the transportation destination position, which is the transportation destination, and waits until it is determined that the transportation destination position has been input (step S24, S25). When the transport target position is input via the input unit 11, the transport planning unit 14 gives the transport target position to a predetermined area of the memory 13 and stores it in the predetermined area (step S26). The optimum transport route is searched for each of the transport machines 2, 2, ... (Step S30).

  FIG. 11 is a flowchart showing a procedure for searching for the optimum transport route shown in FIG. As shown in FIG. 11, the transport planning unit 14 has a sufficiently large number that the Q value does not exist as the initial value of the Q value between the adjacent nodes N, N, N, N, ... For example, an image prompting input of 1000 is displayed on the display unit 12 via the CPU 10 (step S301), and waits until the initial value is input (step S302). When the initial value is input, the transport planning unit 14 sets the value as the initial value of the Q value among the nodes N, N, N, N, ..., (step S303). The transportation plan formulation unit 14 configures the transportation destination position stored in the memory 13 and the warehouse information described above, and calculates the weights preset between the adjacent nodes N, N, N, N, ... According to the following equation (1), the Q value, which is the easiness (cost) of passing between adjacent nodes N, N, N, N, ... Obtain (step S304). The calculation of the Q value is repeatedly executed until the following expression (2), which is the convergence condition, is satisfied.

  Here, the weight set between the adjacent nodes N, N, N, N, ...

  FIG. 4 is an explanatory diagram illustrating an example of weights set between adjacent nodes N, N, N, N, ... In the case shown in FIG. 4, a horizontally-long rectangular ring-shaped conveying path 35 is provided, and the two short sides of the ring-shaped conveying path 35 are connected to the center position of the ring-shaped conveying path 35 in the vertical direction. A central transport path 36 is provided. Further, internal conveying paths 37, 37, 37, 37, 37 connecting between the two long sides of the annular conveying path 35 are provided so as to divide the inside of the annular conveying path 35 into six equal parts in the width direction. Further, a U-shaped external conveyance path 38 is provided around the annular conveyance path 35 such that one longitudinal conveyance path 39 is located at the center position of one long side of the annular conveyance path 35.

  Nodes N, N, ... Are at positions where the vertical transport path and the horizontal transport path forming the annular transport path 35, the central transport path 36, the internal transport paths 37, 37, 37, 37, 37, and the external transport path 38 intersect. Yes, serial numbers are sequentially provided in the horizontal direction step by step from the node N at the upper left position of the annular transport path 35. In the case shown in FIG. 4, the upper left node N is number 1, and the lower right position of the external transport path 38 is number 23.

  The weight set between the adjacent nodes N, N, N, N, ..., Means the ease of passing the carrier machine between the adjacent nodes N, N, N, N ,. And is determined by the distance between adjacent nodes N, N, the degree of failure between the nodes N, N, and other external factors. In the case shown in FIG. 4, from the 1st node N to the 2nd node N, from the 2nd node N to the 1st node N, and from the 22nd node N When transporting to the 23rd node N, the distance between both nodes N and N is two blocks apart, and there is no obstacle between both nodes N and N and other external factors. It is set to 2 ". Similarly, when carrying from the first node N to the eighth node N, when carrying from the eighth node N to the first node N, and when carrying from the eighteenth node N to the twenty-second node N. In this case, the distances between the nodes N and N are 4 blocks, 4 blocks, and 3 blocks apart, and there are no obstacles between the nodes N and N and other external factors. It is set to "4" and "3". Here, it is assumed that each of the blocks described above is generated by dividing adjacent nodes N and N according to a predetermined unit.

  On the other hand, when transporting from the 22nd node N to the 18th node N, and when transporting from the 23rd node N to the 22nd node N, it is impossible to transport due to a failure or an external factor, and thus it is predetermined. A maximum value weight (for example, 1000) is set, which prohibits transportation from the 22nd node N to the 18th node N and transportation from the 23rd node N to the 22nd node N. ..

  FIG. 5 is a table showing an intermediate result of repeatedly calculating the Q value using the transport path and the weight shown in FIG. Here, the final transport destination position is the 22nd node, and the Q value of this 22nd node is defined as zero.

  Now, if n = 0 in the equation (1), the Q value of the 22nd node will be obtained, so the Q value is zero from the above definition. Therefore, in the table, the node is 22 and the orientation is 22. 0 is entered as the Q value.

  Next, assuming that n = 1 in the equation (1), the 18th and 23rd nodes adjacent to the 22nd node are targeted, and the Q value to be calculated is Q (18, 22) or Q. (23,22). However, as shown in FIG. 4, since the transport from the 23rd node to the 22nd node is prohibited, the Q value to be calculated is only Q (18, 22). Q (18,22) is calculated from the equation (1) as the sum of MinQ (22, k) and 3 which is the weight between the 18th node and the 22nd node. , K) corresponds to the 22nd node, which is the final target position, so that only Q = 0 exists, and MinQ (22, k) = 0, so that Q (18,22) = 3 + 0 = 3 is obtained. Therefore, in the table, 3 is entered as the Q value when the node is 18 and the orientation is 22.

  Next, assuming that n = 2 in the equation (1), the 17th node, the 19th node, and the 11th node that are adjacent to the 18th node are targeted, so the Q value to be calculated is Q (17, 18), Q (19,18) and Q (11,18). Here, Q (17,18) is calculated as the sum of MinQ (18, k) and 2 which is the weight between the 17th node and the 18th node from the equation (1). Here, the Q value of the 18th node has four values Q (18,11), Q (18,17), Q (18,22), and Q (18,19). In addition to Q (18,22) = 3, the initial value is set to, for example, 1000 as described above, and thus the minimum value is 3 and MinQ (18, k) = 3. .. Therefore, since Q (17,18) = 2 + 3 = 5 is obtained, 5 is entered as the Q value when the node is 17 and the orientation is 18 in the table.

  Since Q (19,18) = 2 + 3 = 5 is obtained by performing the same calculation, 5 is entered as the Q value when the node is 19 and the orientation is 18 in the table.

  The Q value between the adjacent nodes is calculated by repeatedly executing such calculation until the convergence condition (2) is satisfied.

  When the Q values between adjacent nodes are obtained in this way, the transport planning unit 14 reads the positions of all racks R, R, ... (See FIG. 1) included in the warehouse information. It is determined whether or not there is a node position of the rack R that has not been read (step S305), and the position of the node of the rack R that has not been read is read (step S306), and the position of the node is the transport destination position. It is determined whether there is any (step S307). When the transfer plan formulation unit 14 determines that the position is not the transfer target position, it determines the positions of all the nodes adjacent to the node (step S308), and determines the position of the node having a smaller Q value with the node. A selection is made (step S309). The transport plan formulation unit 14 repeats the operations from step S307 to step S309 until it determines that the position of the node is the transport target position, thereby generating the optimal transport route. Then, the transport planning unit 14 repeats such an operation until it determines in step S305 that there is no position of the rack R that has not been read out, thereby obtaining the optimal transport routes for all the racks R, R, ....

  In this way, when the optimum transport route is obtained for each of the racks R, R, ..., The transport plan formulation unit 14 searches for the suboptimal transport route as follows (step S50).

  12 to 14 are flowcharts showing the procedure for searching the suboptimal transport route shown in FIG. As shown in FIG. 12, the transportation plan formulation unit 14 determines the position of each node N, N, ... (Both of which are shown in FIG. 1) in which the racks R, R ,. All the positions of are referred to as nodes are acquired (step S501).

  The transfer plan formulation unit 14 determines whether or not there is an unselected node corresponding to the racks R, R, ... (Step S502), and when there is an unselected node, selects an arbitrary node. It is set as the current node (candidate node) (steps S503 and S504), and the optimum transport route corresponding to the selected node is specified from the respective optimum transport routes obtained in step S30 (step S505).

  Next, the transportation plan formulation unit 14 specifies all nodes adjacent to the current node based on the warehouse information (step S506). The transport plan formulation unit 14 determines whether or not there is an unadopted node for the specified node (step S507), and if there is an unadopted node, selects an arbitrary node from the specified nodes (step S507). S508).

  Whether the direction from the current node to the adopted node (first node) is the same as the direction from the current node to the adjacent node in the corresponding optimum transport route is determined by the transport plan formulation unit 14. (Step S509), and if the directions are the same, the adopted node is set as a candidate node (step S510).

  On the other hand, when it is determined in step S509 that they are not in the same direction, the transport plan formulation unit 14 acquires the Q value between the current node and the adopted node as shown in FIG. 13, and sets it as a. Together with (step S601), all nodes adjacent to the adopted node are recognized (step S602). The transport plan formulation unit 14 determines whether or not an undesignated node exists for the certified node (step S603), and if an undesignated node exists, designates an arbitrary node (step S604). The transport planning unit 14 determines whether the designated node (second node) is the same as the current node (step S605), and if the designated node is the same, the designated node should be excluded from the candidates. Returning to step S603, it is determined whether or not there is an unspecified node. As a result, it is possible to avoid backtracking on the same route, and it is possible to prevent the generation of a suboptimal transport route whose transport efficiency is obviously low.

  When the transport plan formulation unit 14 determines that the node designated in step S605 is different from the current node, the transport plan formulation unit 14 acquires the Q value between the adopted node and the designated node and sets this as b (step S606). Then, it is determined whether or not this b is less than the a (step S607). When b is less than a, the transfer plan formulation unit 14 sets the node adopted in step S508 as a candidate node (step S608), while when b is a or more, excludes the designated node from the candidates. In order to do so, the process returns to step S603, and it is determined whether or not there is an unspecified node. When it is determined in step S509 that a node in a direction different from the direction from the current node to the adjacent node in the optimal transport route is adopted, the suboptimal transport route including the node becomes redundant than the optimal transport route. However, by detecting a candidate node in which b is less than a in this way, it is possible to prevent the redundancy from being increased, and it is possible to search for a suboptimal transport route with higher transport efficiency. ..

  The transport plan formulation unit 14 repeats the operations from step S603 to step S608 until it determines in step S603 that there is no unspecified node. Depending on the node adopted in step S507, no node that satisfies the condition of step S607 may be designated at all, and in such a case, the adopted node is excluded from the candidate nodes.

  When the node selected in step S510 is set as a candidate node, or when it is determined in step S603 that there is no unspecified node, the transportation plan formulation unit 14 returns to step S507 and the unselected node exists. By repeating the operations from step S507 to step S608 until it is determined not to be performed, it is determined whether or not each node adjacent to the current node becomes a candidate node.

  Next, when the transport plan formulation unit 14 determines in step S507 that there is no unadopted node, the transfer plan formulation unit 14 proceeds to step S701 and has not yet acquired candidate nodes for all candidate nodes generated in steps S510 and S608. To determine if exists. When determining that there is an unacquired candidate node, the transfer plan formulation unit 14 acquires an arbitrary candidate node (step S702) and determines whether the acquired candidate node is the node at the final transfer destination position. It is determined (step S703). Then, when the transfer plan formulation unit 14 determines that the acquired candidate node is not the node at the final transfer destination position, it updates the current node to the candidate node (step S704).

  On the other hand, when the transfer plan formulation unit 14 determines that the candidate node acquired in step S702 is the node at the final transfer destination position, it generates a route passing through each candidate node related to the detection of the candidate node. , Which is the second best transport route (step S710).

  The transport plan formulation unit 14 repeats the operations of steps S701 to S710 until it determines in step S701 that there is no unacquired candidate node, and then determines whether or not the current node has been updated ( (Step S720), by repeating the operations from Step S506 to Step S710 until it is determined that the current node is not updated, the plurality of suboptimal transports related to the node selected in Step S503, that is, any rack R. Generate a route.

  At this time, a sub-optimal transport route having the same route as the optimum transport route may be generated, but in this case, the sub-optimal transport route is excluded. Further, the generation of the sub-optimal transport route may be stopped when a predetermined number of sub-optimal transport routes are generated, or the number of sub-optimal transport routes generated may be changed by an external factor.

  The transport plan formulation unit 14 repeats such an operation until it determines in step S502 that there is no unselected node, thereby generating a plurality of suboptimal transport routes for each rack R, R ,. ..

  In this way, when the optimum transport route and the plurality of suboptimal transport routes are generated for each rack R, R, ..., The transport plan formulation unit 14 assigns to each rack R, R, ... As shown in FIG. The optimum transport path and each suboptimal transport path respectively corresponding to the transport machines 2, 2, ... Which are located close to each other are assigned to obtain a transport path group for each transport machine (step S81). The transfer plan formulation unit 14 determines whether or not a preset number of transfer plan candidates have been generated (step S82). If the transfer plan candidates have not reached the predetermined number, as shown in FIG. An optimal transport route or suboptimal transport route included therein is randomly selected from the transport route group (step S83), and, for example, 0, 1 or 2 is selected in a plurality of locations of each selected optimal transport route or suboptimal transport route. By setting the waiting time at random (step S84), the optimal transport route or the suboptimal transport route and the transport plan candidate in which the standby time is set are generated for each of the transport machines 2, 2, ... (Step S85). The transportation plan formulation unit 14 repeats the operations of steps S82 to S85 until it determines in step S82 that the predetermined number has been reached, and generates a predetermined number of transportation plan candidates. It should be noted that, in the generated transport plan candidates, if there are overlapping transport plan candidates, they are excluded in advance, and the optimum transport route constituting each transport machine-specific transport route group is generated by any of the generated transport plan candidates. It is designed to be reflected in the transportation plan candidates.

  FIG. 6 is an explanatory diagram illustrating a method of generating a transportation plan candidate. In the example shown in FIG. 6, transporter-specific transport route groups P1, P1, P1 are generated for each of the transporters A, B, C, and the transporter-specific transport route groups P1, P1, P1 are included in the transporter-specific transport route groups P1, P1, P1. The optimal transport route and two suboptimal transport routes are included. In each of the optimum transport path and the suboptimal transport path, the position where the node is arranged is represented by a square first block 31, 31, ... To facilitate understanding, and adjacent nodes are adjacent to each other. The conveyance path between them is represented by the rectangular second blocks 32, 32, ....

  In order to generate the transport plan candidates, for example, one suboptimal transport route is selected from the optimal transport route and the suboptimal transport route for the transport device A, and the optimal transport route and the suboptimal transport route for the transport device B are selected. The optimum transport route is selected from the above, and the optimum transport route is selected from the optimum transport route related to the carrier C and the second best transport route. Then, by setting a standby time of 0, 1 or 2 at random in each of the second blocks 32, 32, ... Which respectively constitute the selected suboptimal transport path and both optimum transport paths, the transport machines A to C are set. A first transfer plan candidate group including three corresponding transfer plan candidates is generated. In addition, in the case shown in FIG. 6, the second blocks 32, 32, ... With the waiting time 0 set are shown as blanks, and the second blocks 23, 23 ,. Are indicated by diagonal lines, and the second blocks 32, 32, ... In which the waiting time 2 is set are indicated by dots.

  Next, the optimum transport route is selected from the optimum transport route and the second best transport route related to the carrier A, and the one best transport route is selected from the optimum transport route and the second best transport route related to the transport device B, One of the second best transport route is selected from the optimal transport route and the second best transport route related to the carrier C. Then, by setting a standby time of 0, 1 or 2 at random in each of the second blocks 32, 32, ... Which respectively constitute the selected suboptimal transport path and both optimum transport paths, the transport machines A to C are set. A transfer plan candidate group P2, P2, P2 composed of three corresponding transfer plan candidates is generated. One unit of the waiting time can be appropriately set according to the configuration of the transport path, the specifications of the transporting machine, and the maximum waiting time can be appropriately set.

  By repeating such an operation until a preset number of transport plan candidate groups are generated, a predetermined number of transport plan candidate groups are generated.

  Next, as shown in FIG. 10, the transfer plan formulation unit 14 calculates the conformity for each of the generated transfer plan candidate groups using the following equation (3) (step S90), and calculates the calculated conformance. The transport plan candidate groups are rearranged in accordance with (step S91). In the equation (3), the shorter the time required to reach the target transportation position on the optimum transport path and the suboptimal transport path, and the more the number of overlaps in which a plurality of transport machines exist on the same transport path at the same time. The smaller the number, the higher the fitness value.

  The transfer plan formulation unit 14 determines whether or not the transfer plan candidate group, which will be described later, has been updated a predetermined number of times (step S92). If the transfer plan has not reached the predetermined number of times, for example, the highest suitability is five. In addition, after taking a measure to temporarily store the transport plan candidates having the higher suitability (step S93), for example, the medium transport plan candidate group having the middle suitability and the suitability lower. It divides into a subordinate transportation plan candidate group (step S94), applies a genetic algorithm to both transportation plan candidate groups, and generates a new transportation plan candidate group as follows.

  First, the gene crossover is applied to the candidate group for the intermediate transportation plan. That is, all the transportation plan candidates belonging to the medium-level transportation plan candidate group are divided into two groups, and one transportation plan candidate is selected from each group to form a pair. For each transport route included in the paired transport plan candidates, the operation of switching the transport route corresponding to each transport machine so that the preset crossover frequency is performed is performed for each of the transport plan candidates of all pairs ( Step S95).

  On the other hand, the gene mutation is applied to the lower transportation plan candidate group. That is, one transportation plan candidate is randomly selected from all the transportation plan candidates belonging to the lower transportation plan candidate group, and other transportation plan candidates belonging to the lower transportation plan candidate group are selected so that each transportation route constitutes the selected transportation plan candidate. Each transportation route that constitutes all transportation plan candidates is updated. Then, the standby time is updated with a preset mutation probability for each of the transport routes that respectively constitute all of these transport plan candidates (step S96).

  The transport plan formulation unit 14 repeats the operations from step S92 to step S96 until it determines in step S92 that the transport plan candidate group has been updated a predetermined number of times. When it is determined that the transfer plan candidate group has been updated a predetermined number of times, the transfer plan formulation unit 14 selects the transfer plan candidate group having the highest degree of conformity and selects each transfer plan that constitutes the transfer plan candidate group. The optimum transportation plan is given to the CPU 10 (step S100).

  In this way, in order to search for the optimal transport plan for each transport machine from a plurality of transport plan candidate groups with waiting times set for the optimum transport path or the suboptimal transport path, congestion caused by multiple transport machines on the transport path It is possible to prevent the occurrence of the above, and each transporting machine can transport the transported object to the target position with higher efficiency as a whole.

  Next, the transport plan obtained by the method of the present invention, that is, the optimal transport route and the suboptimal transport route are generated, and the transported object is transported according to the transport plan searched using the generated optimal transport route and the suboptimal transport route. The simulation result is compared with the result of simulating the transport of the transported object according to the transport plan searched using only the optimal transport route without generating the suboptimal transport route.

  The configuration of the transport path was the configuration shown in FIG. Then, the third node N, the twelfth node N, and the twenty-first node N shown in FIG. 4 are respectively set as the transport start positions, and the 22nd node N is set as the transport target position. Three corresponding transport machines were set up. The weights between adjacent nodes N, N, N, N, ..., Are also set as shown in FIG.

  In the example of the present invention, after the optimum transport route and the two suboptimal transport routes are generated separately for each transport device as described above, the optimum transport route or the next transport route group is generated from the generated transport route group for each transport device. 60 sets of transport plan candidates were generated by randomly setting the waiting time of 0 to 2 while selecting a good transport route, and 30 was set as PENALTY in the above-mentioned equation (3) to obtain the suitability. In addition, the frequency of gene crossover was set to 2, the probability of gene mutation was set to 10, and a genetic algorithm was applied over 201 generations to search for an optimal transfer plan group. Each transport plan that constitutes the optimal transport plan group was assigned to each of the three corresponding transport machines, and the transport behavior of each transport machine was simulated according to the assigned transport plan, and the average of the total traveling time of each transport machine was obtained. ..

  On the other hand, in the comparative example, the average of the total traveling time of each carrier was obtained by performing the same operation as in the example of the present invention except that the suboptimal transport route was not generated.

  The following table shows the results of performing the same operation 10 times for both the inventive example and the comparative example.

  As is clear from the table, in all cases, the average of the total running time is equal to or less than that of the comparative example, and therefore, even when congestion occurs between a plurality of transporters on the transport route, each transport is performed. The machine as a whole was able to convey the conveyed object with higher efficiency.

1 Transport Plan Formulation Device 2 Transport Machine 14 Transport Plan Formulation Section 22 Carrying Section 30 Transport Path 50 Transmitter R Rack N Node

Claims (19)

A plurality of transportable objects, which are respectively arranged at a plurality of transport start positions different from each other on a plurality of transport paths provided in a lattice, can be moved to each of the transport target positions set on any of the transport paths. When formulating a plan to carry each by a carrier,
Assign each carrier to the object to be transported,
A value that indicates the ease of transportation to the transport target position by passing between a plurality of intersecting positions where the transport paths from the transport start position to the transport target position intersect , for each transport machine. determined using the distance, on the basis of the values determined, sequentially selects prone intersections relatively transported to generate optimal conveyance path, based on the direction and the value of the generated optimal conveyance path, wherein The optimum position is generated by selecting an intersecting position that includes an intersecting position different from the intersecting position forming the optimum conveying route and is relatively easy to convey, and generating one or a plurality of suboptimal conveying routes different from the optimum conveying route. Obtaining a plurality of carrier-by-machine carrier path groups including a carrier path and one or more suboptimal carrier paths,
Each time an arbitrary transport route is selected from the optimal transport route and the suboptimal transport route included therein from each transport route group for each transport machine, and each time the arbitrary transport route is selected, each intersection that constitutes the transport route is selected. A transportation plan is selected by selecting an arbitrary waiting time from a plurality of preset waiting times for waiting the carrier between positions and setting the selected waiting time between one or a plurality of intersecting positions forming the carrier route. Repeat the operation of generating candidates to create a group of transport plan candidates,
Performing these operations a plurality of times to obtain a plurality of transport plan candidate groups with different waiting positions and / or intersection positions where waiting times are set ,
For each transport plans candidates respectively obtained, the number of times there are multiple conveyor between the same intersection at the same time,及beauty, to the transfer target position from conveyance start position on the time it takes to transport the transported object In calculating the goodness of fit on the basis of the number of times is small, and the shorter the time is to calculate the goodness of fit of the higher value, comparing the values of the calculated goodness of fit, based on the comparison result, the A transport plan formulation characterized by selecting an optimal transport plan candidate group with a higher degree of conformity from a plurality of transport plan candidate groups and setting each transport plan candidate that constitutes the selected transport plan candidate group as a transport plan. Method. Nodes that are the intersecting positions are set at a plurality of positions on each of the transport paths, respectively , and it is easy to pass the transport machine between the two nodes according to the distance between the two adjacent nodes. Set a certain weight in advance,
By applying these weights to the following equations (1) and (2), it is the value of the ease of transport to the transport target position by passing between adjacent nodes in the order of separating the position from the transport target position. Then, the smaller the value is, the easier it is to convey, and the Q value is calculated.
When a sub-optimal transport route is generated, an arbitrary node is selected from one or a plurality of nodes adjacent to the candidate node that should form the sub-optimal transport route as a first node, and the first node is selected from the candidate nodes. It is determined whether the direction to the node is the same as the direction of the corresponding optimum transport route, and if it is not the same direction, the Q value between the candidate node and the first node is specified, and the first node is determined. A second node by designating an arbitrary node from one or a plurality of nodes adjacent to each other, specifying the Q value between the first node and the second node, and comparing the Q value of the former with the Q value of the latter. The transportation planning method according to claim 1, wherein when the value is smaller, the first node is set as a candidate node.

Prior to identifying the Q value between the first node and the second node,
It is determined whether the second node is in the same position as the candidate node when the first node is selected, and if it is in the same position, the Q value between the first node and the second node 3. The transportation planning method according to claim 2, wherein another node adjacent to the first node is specified without specifying the.   When the direction from the candidate node to the first node is the same as the direction of the corresponding optimum transport route, the first node is set without specifying the Q value between the candidate node and the first node. The transportation plan formulation method according to claim 2, wherein the transportation node is a candidate node. When selecting the optimal transport plan candidate group,
The obtained transportation plan candidate groups are ranked according to the value of their suitability,
For each transport plan candidate group having a lower rank than a predetermined rank, each transport route included in each transport plan candidate group is divided into two groups, and transport plan candidates related to an arbitrary transport machine are selected from both groups. As a pair, perform the operation of crossing the transport paths between the plurality of intersection positions respectively constituting the two transport plan candidates with each other up to a preset number of crossover frequencies for all the transport machines. , A new transportation plan candidate group is generated, and / or an arbitrary transportation plan candidate is selected from each transportation plan candidate group, and each transportation route of each transportation plan candidate group is included in the selected transportation plan candidates. After updating to the transport route, by applying the genetic algorithm to the waiting time set in each transport route, to generate different waiting time with a preset mutation probability, by updating to the generated waiting time , Generate new transport plan candidate groups,
After repeating a predetermined number of operations for ranking each transfer plan candidate group having a higher rank than the predetermined rank and each new transfer plan candidate group generated, according to the value of the conformity degree, a higher rank 5. The transportation plan formulation method according to claim 1, wherein the transportation plan candidate group is an optimal transportation plan candidate group. The transportation plan formulation method according to claim 1, wherein the suitability is calculated using the following equation (3).

A plurality of transportable objects, which are respectively arranged at a plurality of transport start positions different from each other on a plurality of transport paths provided in a lattice, can be moved to each of the transport target positions set on any of the transport paths. When formulating a plan to carry each by a carrier,
A means for allocating each carrier to the object to be transported,
A value that indicates the ease of transportation to the transport target position by passing between a plurality of intersecting positions where the transport paths from the transport start position to the transport target position intersect , for each transport machine. It determined using the distance, based on the values obtained, and means for generating an optimal transporting path sequentially selects easily intersection relative transport, based on the direction and the value of the generated optimal transport path , Generating an one or a plurality of suboptimal transport routes different from the optimal transport route by selecting an intersecting position that is relatively easy to transport, including an intersecting position different from the intersecting position forming the optimal transport route, A means for obtaining a plurality of carrier-by-machine transport path groups including the optimum transport path and one or more suboptimal transport paths;
Conveying route selecting means for selecting an arbitrary conveying route out of the optimum conveying route and the suboptimal conveying route included therein from each conveying route group for each conveying machine, and the conveying route each time the arbitrary conveying route is selected. The waiting time is selected from a plurality of waiting times set in advance to make the carrier stand by between the intersecting positions forming the crossing position , and the selected waiting time is set between the one or more intersecting positions forming the carrying route. And a transportation plan candidate generation means for setting and generating a transportation plan candidate,
Every time the conveyance path selection means for selecting any of the conveying path, the conveying plan candidates by repeating an operation of conveying plan candidate generating means generates a conveying plan candidates, performed a plurality of times these operations, the waiting time and / Or , it is designed to obtain a plurality of transportation plan candidate groups having different waiting positions, which are different from each other in the intersection position ,
Further, for each of the transport plan candidates obtained, required to transport number及beauty, to the transfer target position from conveyance starting position the transported object in which a plurality of conveyor is present between the same intersection at the same time In calculating the goodness of fit based on time, the smaller the number of times, the shorter the time, the higher the goodness of fit is calculated , the calculated values of the goodness of fit are compared, and based on the comparison result. , An optimum transfer plan candidate group selecting means for selecting an optimum transfer plan candidate group having a higher degree of conformity from the plurality of transfer plan candidate groups, and each transfer plan candidate forming the selected transfer plan candidate group, An apparatus for formulating a transportation plan, comprising: means for making a transportation plan. Wherein the plurality of positions of the conveying path are respectively the said intersection and a is in node is set positions, depending on the distance between each two nodes between Aitonaru nodes conveyor between both nodes The weight that is easy to pass is preset,
By applying these weights to the following preset expressions (1) and (2), it is easy to carry to the transfer target position by passing between adjacent nodes in order of separating the position from the transfer target position. And a means for calculating a Q value for which it is determined that the smaller the value, the easier it is to carry ,
The sub-optimal transport route generation means,
A means for selecting an arbitrary node from one or a plurality of nodes adjacent to a candidate node to form the suboptimal transport route as a first node, and a direction from the candidate node to the first node, Direction determining means for determining whether or not the direction is the same as the direction of the corresponding optimum transport path,
If not in the same direction, current node Q value specifying means for specifying the Q value between the candidate node and the first node,
Node designating means for designating an arbitrary node from one or a plurality of nodes adjacent to the first node as a second node;
Designated node Q value specifying means for specifying a Q value between the first node and the second node,
The transportation plan formulation device according to claim 7, further comprising: a candidate node determining unit that determines the first node as a candidate node when the latter Q value is smaller than the former Q value.

Further, the specified node Q value specifying means, prior to specifying the Q value between the first node and the second node, is a candidate node when the second node selects the first node. Equipped with means for determining whether or not it is in the same position as
In the case of the same position, the specified node Q value specifying means does not specify the Q value between the first node and the second node, and the node specifying means does not specify another Q value adjacent to the first node. 9. The transportation plan formulation device according to claim 8, wherein the node is designated.   When the direction determining unit determines that the direction from the candidate node to the first node is the same as the direction of the corresponding optimum transport route, the current node Q value specifying unit determines that the candidate node and the first node are the same. 10. The transportation plan formulation device according to claim 8 or 9, wherein the candidate node determination means sets the first node as a candidate node without specifying a Q value between them. The optimum transportation plan candidate group selecting means,
A ranking means for ranking each obtained transport plan candidate group according to the value of its suitability,
For each transport plan candidate group having a lower rank than a predetermined rank, each transport route included in each transport plan candidate group is divided into two groups, and transport plan candidates related to an arbitrary transport machine are selected from both groups. and a pair, a crossover to replace the conveying path which is between a plurality of intersections which constitutes both the conveying plan candidates each, performed the operations that were crossed with each other up to the predetermined cross frequency number to all the conveyor , A new transportation plan candidate group is generated, and / or an arbitrary transportation plan candidate is selected from each transportation plan candidate group, and each transportation route of each transportation plan candidate group is included in the selected transportation plan candidates. After updating to the transport route, by applying the genetic algorithm to the waiting time set in each transport route, to generate different waiting time with a preset mutation probability, by updating to the generated waiting time, And a new transportation plan candidate group generation means for respectively generating new transportation plan candidate groups,
The ranking means is configured to rank each transport plan candidate group having a higher rank than a predetermined rank and each generated new transport plan candidate group in accordance with the value of the compatibility.
The new transportation plan candidate group generation means generates new transportation plan candidate groups respectively, and the ranking means repeats the ranking operation a predetermined number of times, and then a higher ranking transportation plan candidate group is optimized. The transportation plan formulation device according to any one of claims 7 to 10, further comprising: a unit that sets a transportation plan candidate group. 12. The transportation plan formulation device according to claim 7, wherein the suitability calculating means is configured to calculate the suitability based on the following equation (3) set in advance.

A plurality of transport paths, a plurality of transport machines that can move on each transport path, the transport plan formulation apparatus according to any one of claims 7 to 12, and the transport plan created by the transport plan formulation apparatus for each transport machine. And a transmitter for transmitting to the carrier system. On the computer, move the objects to be conveyed, which are respectively arranged at a plurality of different conveyance start positions on the plurality of conveyance paths arranged in a grid pattern, to the conveyance target position set on any one of the conveyance paths. It is a program that allows you to formulate a plan to carry separately by multiple carriers to obtain,
On the computer,
Assigning each carrier to the object to be transported,
A value that indicates the ease of transportation to the transport target position by passing between a plurality of intersecting positions where the transport paths from the transport start position to the transport target position intersect , for each transport machine. Based on the obtained values, the step of sequentially selecting relatively easy crossing positions to generate an optimum transport path, and the direction of the generated optimum transport path and the value Then, an intersecting position different from the intersecting position forming the optimum conveying route and relatively easy to convey are selected to generate one or a plurality of suboptimal conveying routes different from the optimum conveying route. A good transport route generation step, and a step of obtaining a plurality of transport route groups for each transport machine including the generated optimal transport route and one or more suboptimal transport routes,
A transport route selecting step of selecting an optimal transport route and a sub-optimal transport route included therein from each transport route group for each transport machine, and the transport of each time the optional transport route is selected. An arbitrary waiting time is selected from a plurality of waiting times set in advance to make the carrier stand by between the intersecting positions forming the route , and the selected waiting time is set between one or a plurality of intersecting positions forming the conveying route. And a transport plan candidate generation step for generating a transport plan candidate by setting to
Each time any transport route is selected in the transport route selection step, the transport plan candidate generation step repeats the operation of generating transport plan candidates to form a transport plan candidate group, and these operations are performed a plurality of times, and the standby. It is designed to obtain a plurality of transport plan candidate groups having different intersection positions where time and / or waiting time are set ,
Further, for each of the transport plan candidates obtained, required to transport number及beauty, to the transfer target position from conveyance starting position the transported object in which a plurality of conveyor is present between the same intersection at the same time In calculating the goodness of fit on the basis of time, the smaller the number of times, the goodness of fit calculating step of calculating the goodness of fit as the time is shorter, and the calculated value of each goodness of fit are compared. Based on the comparison result, an optimum transfer plan candidate group selecting step for selecting an optimum transfer plan candidate group having a higher degree of conformity from the plurality of transfer plan candidate groups, and each of the selected transfer plan candidate groups. And a step of causing each of the transportation plan candidates to be a transportation plan, and causing a computer to formulate the transportation plan. Wherein the plurality of positions of the conveying path are respectively the said intersection and a is in node is set positions, depending on the distance between each two nodes between Aitonaru nodes conveyor between both nodes The weight that is easy to pass is preset,
On the computer,
By applying these weights to the following equations (1) and (2), it is the value of the ease of transport to the transport target position by passing between adjacent nodes in the order of separating the position from the transport target position. And a step of calculating a Q value in which it is determined that the smaller the value is, the easier it is to carry ,
The suboptimal transport route generation step,
A step of selecting an arbitrary node from one or a plurality of nodes adjacent to a candidate node which should form the suboptimal transport route to be a first node; and a direction from the candidate node to the first node, A direction determining step of determining whether or not the direction is the same as that of the corresponding optimum transport route, and a current node Q value identifying step of identifying a Q value between the candidate node and the first node if the direction is not the same. , A node designating step of designating a second node by designating an arbitrary node from one or a plurality of nodes adjacent to the first node, and specifying a Q value between the first node and the second node 15. The computer program according to claim 14, further comprising: a designated node Q value specifying step of causing the first node to be a candidate node when the latter Q value is smaller than the former Q value.

On the computer,
Further, in the designated node Q value specifying step, when the second node selects the first node before specifying the Q value between the first node and the second node, A step of determining whether or not the position is the same as that of the candidate node,
In the case of the same position, the specified node Q value specifying step does not specify the Q value between the first node and the second node, and the node specifying step determines whether another Q node adjacent to the first node is adjacent to another node. 16. The computer program according to claim 15, adapted to specify a node. On the computer,
When it is determined in the direction determining step that the direction from the candidate node to the first node is the same as the direction of the corresponding optimum transport route, the current node Q value specifying step is performed in the candidate node and the first node. The computer program according to claim 15 or 16, wherein the candidate node determination step causes the first node to be a candidate node without specifying a Q value between and. The optimal transport plan candidate group selection step is
On the computer,
A ranking step for ranking each obtained transport plan candidate group according to the value of its conformity, and each transport plan candidate group having a rank lower than a predetermined rank is included in each transport plan candidate group. Each transport route is divided into two groups, transport plan candidates relating to arbitrary transport machines are selected from both groups to form a pair, and transport routes between a plurality of intersecting positions respectively configuring both transport plan candidates are switched. An operation of causing the crossovers to cross each other up to a preset number of crossover frequencies is performed on all the transporting machines to generate new transporting plan candidate groups, and / or from each transporting plan candidate group. After selecting a transport plan candidate and updating each transport route of each transport plan candidate group to each transport route included in the selected transport plan candidates , the genetic algorithm is applied to the waiting time set for each transport route. A new transfer plan candidate group generation step for generating new transfer plan candidate groups by applying them to generate different standby times with preset mutation probabilities and updating the generated standby times. ,
The ranking step is arranged to rank each transport plan candidate group having a higher rank than a predetermined rank and each generated new transport plan candidate group in accordance with the value of the degree of conformity,
After the new transportation plan candidate group generation step generates new transportation plan candidate groups respectively, and the ranking step repeats the ranking operation a predetermined number of times, a higher ranking transportation plan candidate group is optimized. The computer program according to any one of claims 14 to 17, further comprising the step of causing the transfer plan candidate group. The computer program according to any one of claims 14 to 18, wherein the adaptability calculating step causes a computer to calculate the adaptability based on the following equation (3).

Images