A crane repositioning control method, electronic equipment, and media based on partitioned cooperation.

Abstract

This application provides a crane relocation control method, electronic device, and medium based on partitioned coordination, belonging to the field of crane control technology. The method includes: in each decision cycle, using minimizing the total cost of idle crane scheduling among multiple logical partitions as the objective function, solving for the number of idle cranes to be scheduled between each logical partition, and performing a first relocation of idle cranes among multiple logical partitions based on the solution results; determining a target parking node set based on the minimum weighted distance cost between the demand node set and the idle crane set within each logical partition; and performing a second relocation on the idle crane set based on the target parking node set to relocate the idle cranes in the idle crane set to the target parking nodes in the target parking node set. This application can achieve a balanced number of idle cranes at the partition level and refined matching of parking positions within partitions, significantly reducing the average pick-up waiting time of wafer fab production equipment.

Description

Technical Field

[0001] This application relates to the field of overhead crane control technology, and in particular to an overhead crane repositioning control method, electronic equipment, and medium based on partitioned coordination. Background Technology

[0002] In semiconductor wafer fabrication plants, the overhead crane system (OHT), a core component of the Automated Material Handling System (AMHS), plays a crucial role in transporting wafer casks in batches between hundreds of process equipment. The operation of the overhead crane system typically falls into two scenarios: when a handling request arises and an available idle overhead crane exists, the system assigns the task to the idle crane, which then travels to the requesting origin, loads the wafer cask, and transports it to its destination; when an idle overhead crane completes its handling task and there are no pending requests, the idle crane remains on its track, entering an idle state. As the scale of the overhead crane system continues to expand, the frequency or probability of handling tasks varies significantly between different process equipment due to differences in process flows.

[0003] To optimize the distribution of idle overhead cranes and improve the overall response efficiency to handling requests, various idle overhead crane scheduling schemes have been proposed in existing technologies. For example, Chinese patent CN120106702B proposes a dynamic idle overhead crane patrol method, which triggers a decision when the number of idle vehicles changes (such as adding an idle vehicle), selects a target area using priority rules, and moves idle vehicles according to dwell time rules to optimize the mobility of idle overhead cranes in the overall system. In addition, existing solutions aggregate multiple idle overhead cranes into a convoy and travel along a roaming path to respond to transportation tasks more quickly; or pre-plan a series of different standby position groups for the overhead cranes, and schedule the nearest idle group of overhead cranes when there is a handling task to shorten the response time. However, these existing solutions cannot dynamically adjust the distribution density of idle vehicles according to the actual needs of the wafer fab process flow, and also lack systematic and fine-grained optimization of the specific parking nodes of idle overhead cranes. This results in a severe shortage of idle overhead cranes in areas with high handling task demand and a large backlog of vehicles in areas with low demand, which in turn affects the overall response time of the handling needs of production equipment in the wafer fab process flow.

[0004] Therefore, how to achieve precise matching of idle overhead crane parking positions to reduce the average picking time of wafer fab production equipment has become an urgent technical problem to be solved. Summary of the Invention

[0005] The purpose of this application is to provide a crane repositioning control method, electronic device and medium based on partitioned cooperation to solve the above problems.

[0006] To achieve the above objectives, firstly, this application proposes a crane repositioning control method based on partitioned cooperation, the method comprising: In each decision cycle, with the objective function of minimizing the total cost of scheduling idle overhead cranes among multiple logical partitions in the overhead crane system's track network, the number of idle overhead cranes that need to be scheduled among each logical partition is solved, and the first repositioning of the idle overhead cranes among the multiple logical partitions is performed based on the solution results. For each of the plurality of logical partitions, the target parking node set is determined based on the minimum weighted distance cost between the set of demand nodes and the set of idle overhead cranes within each logical partition; A second relocation is performed on the set of available overhead cranes based on the set of target parking nodes, so as to relocate the available overhead cranes in the set of available overhead cranes to the target parking nodes in the set of target parking nodes.

[0007] In some implementations, the step of using minimizing the total cost of scheduling idle cranes among multiple logical partitions in the crane system's track network as the objective function, and solving for the number of idle cranes that need to be scheduled between each logical partition, includes: Based on the comparison results between the current number of idle cranes in each logical partition of the crane system's track network and the lower limit and upper limit of the number of idle cranes corresponding to each logical partition, the scarce partitions in the state of crane shortage and the surplus partitions in the state of crane surplus are determined. With the constraint of scheduling idle overhead cranes from the surplus partition to the shortage partition, and with the objective function of minimizing the total cost of scheduling idle overhead cranes, the number of idle overhead cranes that need to be scheduled between each logical partition is solved.

[0008] In some implementations, the constraint of scheduling idle cranes from the surplus partition to the shortage partition includes: The first constraint is that the total number of idle overhead cranes transferred from the surplus zone does not exceed the surplus amount of idle overhead cranes corresponding to the surplus zone, and the second constraint is that the total number of idle overhead cranes transferred into the shortage zone does not exceed the shortage amount of idle overhead cranes corresponding to the shortage zone. The objective function of minimizing the total cost of scheduling idle overhead cranes includes minimizing the sum of the unit cost of scheduling a single idle overhead crane from one logical partition to another by multiplying it by the number of idle overhead cranes scheduled from one logical partition to another.

[0009] In some implementations, determining the target parking node set based on the minimum weighted distance cost between the demand node set and the idle crane set within each logical partition includes: For each logical partition, an initial set of parking nodes is selected from the set of candidate parking nodes based on the selection probability value of each candidate parking node. Calculate the weighted distance cost between the set of demand nodes within each logical partition and the initial set of parking nodes; Replace the initial parking node in the initial parking node set with the remaining parking nodes in the candidate parking node set to obtain an updated parking node set. Repeat the calculation of the weighted distance cost between the demand node set and the updated parking node set until the minimum weighted distance cost value is reached. The updated parking node set corresponding to the minimum weighted distance cost value is taken as the target parking node set.

[0010] In some implementations, selecting an initial set of parking nodes from the set of candidate parking nodes based on the selection probability value of each candidate parking node includes: The demand node with the highest task occurrence probability weight in the set of demand nodes is selected as the first initial parking node. Calculate the demand coverage and the shortest path distance between each candidate parking node and the currently selected initial parking node in sequence, and use the product of the demand coverage and the shortest path distance as the weighted distance; The selection probability value of each candidate parking node is determined based on the normalized result of the weighted distance of each candidate parking node; The remaining initial parking nodes are selected sequentially from the candidate parking node set in descending order of the selection probability values ​​to form the initial parking node set.

[0011] In some implementations, calculating the weighted distance cost between the set of demand nodes within each logical partition and the initial set of parking nodes includes: For each set of demand nodes within a logical partition, calculate the shortest path distance from each demand node in the set of demand nodes to the initial set of parking nodes. The weighted distance cost is obtained by summing the products of the shortest path distance from each demand node to the initial set of parking nodes and the task occurrence probability weight of each demand node.

[0012] In some implementations, the step of replacing the initial parking nodes in the initial parking node set with the remaining parking nodes in the candidate parking node set to obtain an updated parking node set, and repeatedly calculating the weighted distance cost between the demand node set and the updated parking node set until the minimum weighted distance cost value is reached, includes: Traverse the remaining parking nodes in the candidate parking node set and the initial parking nodes in the initial parking node set. Replace one initial parking node with one remaining parking node to obtain an updated parking node set. Calculate the weighted distance cost between the demand node set and the updated parking node set. If the calculated weighted distance cost is less than the weighted distance cost corresponding to the initial parking node set, then the updated parking node set is used as the new initial parking node set. Repeat the process of traversing the remaining parking nodes in the candidate parking node set and the initial parking node in the initial parking node set, replacing an initial parking node with a remaining parking node to obtain an updated parking node set, until the minimum weighted distance cost value is reached.

[0013] In some implementations, performing a second relocation of the set of idle overhead cranes based on the set of target parking nodes includes: Calculate the shortest travel distance from the current position of each idle crane in the idle crane set to each target parking node in the target parking node set in turn; According to the ascending order of the shortest travel distance, each idle crane is assigned a target parking node from the target parking node set, and each target parking node is assigned only one idle crane.

[0014] Secondly, to achieve the above objectives, this application also proposes an electronic device, comprising: one or more processors; and a memory for storing one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors execute the overhead crane repositioning control method based on partitioned collaboration as described above.

[0015] Thirdly, to achieve the above objectives, this application also proposes a computer storage medium storing executable instructions, which, when executed by a processor, cause the processor to perform the overhead crane repositioning control method based on partitioned collaboration as described above.

[0016] Compared with the prior art, the beneficial effects of this application include: On the one hand, the adoption of a zone-coordinated overhead crane relocation strategy achieves dynamic balance optimization of the number and location of idle overhead cranes among logical zones. This ensures that zones with high handling needs obtain sufficient idle overhead crane resources, while zones with low handling needs avoid overhead crane backlog, and achieves a balance of overhead crane resources among zones with minimal scheduling costs. This avoids unnecessary large-scale empty runs that would waste energy and track resources, and improves the overall resource utilization and system throughput of the overhead crane system.

[0017] On the other hand, by optimizing the parking node positions within the partition through micro-relocation within the partition, idle overhead cranes are concentrated at target parking nodes with high task occurrence probability and good coverage, which greatly reduces the average pickup distance and can significantly reduce the average pickup time of production equipment in the wafer fab process flow, thereby improving the overall production efficiency of the wafer fab process flow.

[0018] Therefore, the embodiments of this application achieve a balance between global scheduling efficiency and local precision from two levels: macro-level quantity balance and micro-level position optimization. This collaboratively reduces the overall response time of the overhead crane system to the handling requests of wafer fab production equipment, improves the utilization efficiency of overhead crane resources, avoids unnecessary long-distance empty runs, and ultimately improves the overall production efficiency of the wafer fab process. Attached Figure Description

[0019] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments will be briefly described below. It should be understood that the following drawings only show some embodiments of this application and should not be regarded as a limitation on the scope of this application.

[0020] Figure 1 This is a flowchart illustrating a crane repositioning control method based on partitioned collaboration in one embodiment. Figure 2 This is a detailed flowchart illustrating the process of determining the target parking node set based on the minimum weighted distance cost value between the demand node set and the idle crane set within each logical partition, as shown in one embodiment. Figure 3 This is a detailed flowchart illustrating the process of selecting an initial set of parking nodes from the set of candidate parking nodes based on the selection probability value of each candidate parking node in one embodiment. Figure 4 This is a detailed flowchart illustrating the process of calculating the weighted distance cost between the set of demand nodes within each logical partition and the initial set of parking nodes in one embodiment. Figure 5 This is a detailed flowchart illustrating the process of replacing the initial parking node in the initial parking node set with the remaining parking nodes in the candidate parking node set to obtain an updated parking node set in one embodiment. Figure 6 This is a schematic diagram of the electronic equipment involved in the crane repositioning control method based on partition collaboration in the embodiments of this application. Detailed Implementation

[0021] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0022] All terms used in this application (including technical and scientific terms) have the meanings commonly understood by those skilled in the art, unless otherwise defined. It should be noted that the terms used herein should be interpreted in a manner consistent with the context of this specification, and not in an idealized or overly rigid way.

[0023] For example, the terms "first," "second," etc., used in this application may be used herein to describe various elements, but these elements are not limited by these terms. These terms are only used to distinguish the first element from another element.

[0024] For example, the terms "comprising" or "including" used in this application indicate the presence of features, steps, operations and / or components, but do not exclude the presence or addition of one or more other features, steps, operations or components.

[0025] As mentioned earlier, existing idle overhead crane scheduling schemes cannot dynamically adjust the distribution density of idle cranes according to the actual needs of the wafer fab process flow, and also lack systematic and fine-grained optimization of specific parking nodes for idle overhead cranes. This results in a severe shortage of idle overhead cranes in areas with high demand for handling tasks, and a large backlog of cranes in areas with low demand, which in turn affects the overall response time of the handling needs of production equipment in the wafer fab process flow. To address this, this application proposes an overhead crane repositioning control method, electronic equipment, and medium based on zone collaboration, which can achieve a balanced number of idle overhead cranes at the zone level and a fine-grained matching of parking positions within the zone, significantly reducing the average picking time of wafer fab production equipment.

[0026] like Figure 1 As shown in the figure, this application provides a crane repositioning control method based on partitioned cooperation, the method including the following steps: Step S10: In each decision cycle, with the objective function of minimizing the total cost of scheduling idle overhead cranes among multiple logical partitions in the overhead crane system's track network, the number of idle overhead cranes that need to be scheduled between each logical partition is solved, and the first repositioning of the idle overhead cranes among the multiple logical partitions is performed based on the solution results.

[0027] In this embodiment, the decision cycle refers to the time interval during which the system executes a complete relocation control cycle. This can be a pre-set fixed time interval (e.g., every 5 minutes, every 10 minutes, every 30 minutes, etc.) or a conditional cycle dynamically triggered based on the system status. The overhead crane system refers to an automatically operating aerial transport trolley (OHT) installed on a track, used to move wafer cassettes between various production equipment in a semiconductor wafer fab. The track network refers to the physical track route diagram of the overhead crane, connecting various production equipment, buffer zones, bypass tracks, and other nodes. The overhead crane system's track network can be pre-divided into multiple logical partitions based on the process type of the production equipment (e.g., lithography area, etching area, deposition area) or physical area (e.g., different floors or workshops in the fab). The production equipment within each logical partition is called a demand node, and these demand nodes generate wafer cassette moving tasks. An idle overhead crane is an overhead crane that is not currently performing any moving tasks and is in a standby state. The total scheduling cost refers to the sum of all scheduling actions performed to eliminate the imbalance in the number of idle overhead cranes between logical partitions (excess partitions have surplus idle overhead cranes, while scarce partitions have insufficient idle overhead cranes), thereby adjusting the number of idle overhead cranes in all partitions to within the range of their respective lower and upper limits.

[0028] In some implementations, based on the comparison results between the current number of idle cranes in each logical partition of the crane system's track network and the corresponding lower and upper limits of the number of idle cranes for each logical partition, the scarce partitions and the surplus partitions can be determined. With the constraint of scheduling idle cranes from the surplus partitions to the scarce partitions, and with the objective function of minimizing the total cost of scheduling idle cranes, the number of idle cranes that need to be scheduled between each logical partition can be solved.

[0029] Specifically, for each logical partition i∈N={1,2,…,n}, where n represents the number of logical partitions and N represents the set of logical partitions, the following parameters are defined: : The number of idle overhead cranes in the i-th logical partition (which can be obtained in real time through sensors); The minimum number of idle overhead cranes that the i-th logical partition should maintain. ,in, This represents the average number of transport requests per unit time for this logical partition. As a lower limit empirical coefficient (e.g., 0.1), this parameter ensures the basic responsiveness of idle overhead cranes in logical partitions; The maximum number of idle cranes allowed to remain in the i-th logical partition. ,in The total number of overhead cranes in the system. This parameter is an upper limit empirical coefficient (e.g., 0.3) to prevent logical partitions from becoming overcrowded.

[0030] Therefore, the shortage of available vehicles in each logical partition can be expressed as: The surplus of idle vehicles can be expressed as: .like If logical partition i is in a state of overhead crane shortage, then If so, then there is an oversupply of overhead cranes.

[0031] Step S10, which performs the first relocation of idle overhead cranes among multiple logical partitions, aims to: schedule idle overhead cranes from surplus partitions to scarce partitions, so that after scheduling, the number of idle overhead cranes in all logical partitions falls within […]. Within the interval, minimize the total scheduling cost of idle cranes between different logical partitions.

[0032] The objective function is to minimize the sum of the unit cost of scheduling a single idle crane from one logical partition to another multiplied by the number of idle cranes scheduled from one logical partition to another.

[0033] in, This represents the unit cost of scheduling an idle overhead crane from logical partition i to logical partition j, which can be set as the shortest path distance or average travel time between the center points of the two partitions. This represents the number of idle cranes scheduled from logical partition i to logical partition j, and is a decision variable.

[0034] Based on the above objective function, the following constraints can also be constructed: The first constraint is that the total number of idle cranes transferred from the excess partition does not exceed the surplus of idle cranes corresponding to the excess partition. ,in This represents the total number of idle overhead cranes transferred from logical partition i. The second constraint is that the total number of idle overhead cranes transferred into the scarce partition does not exceed the corresponding shortage of idle overhead cranes in that scarce partition. The number of idle cranes in each logical partition after scheduling all fall within [ The interval contains the third constraint condition, i.e. ; .in, This represents the original number of idle overhead cranes within logical partition j. This represents the number of idle cranes transferred from any logical partition to logical partition j. This represents the number of idle cranes retrieved from logical partition j.

[0035] The linear programming problem constructed based on the above decision variables, constraints, and objective function can be solved in milliseconds using the minimum cost flow algorithm, thus yielding the number of idle cranes scheduled from logical partition i to logical partition j. After scheduling is completed, the overhead crane control system issues a first repositioning movement command across zones to the relevant idle overhead cranes, moving these idle overhead cranes from their original logical zones to any parking node in the corresponding target logical zone. Parking nodes can be loading / unloading points for production equipment, dedicated buffer zones, bypass rails, or waiting areas near switches in the overhead crane system's track network.

[0036] It should be noted that after the first relocation between logical partitions is completed, the number of idle cranes in each logical partition i will be updated. for: .in, This represents the number of idle cranes transferred from any logical partition to logical partition i. This represents the total number of idle overhead cranes retrieved from logical partition i.

[0037] Step S20: For each of the plurality of logical partitions, determine the target parking node set based on the minimum weighted distance cost value between the demand node set and the idle crane set within each logical partition.

[0038] In this embodiment, the demand node set refers to the set of all production equipment (i.e., wafer process equipment) within a partition that has material handling requests. Each demand node will generate a handling task with a certain probability. The idle crane set refers to the set of all idle cranes within the partition. The target parking node set refers to the set of optimal locations for idle cranes to park, obtained by solving the minimum weighted distance cost value between the demand node set and the idle crane set. The weighted distance cost is an indicator used to evaluate the quality of a parking node set.

[0039] This embodiment takes into account that after the first relocation of idle overhead cranes between logical partitions is completed in step S10, the idle overhead cranes in each logical region may be distributed on different demand nodes or other arbitrary parking nodes within the current partition. At this time, the location of the idle overhead cranes in each logical partition is not optimal. Therefore, in step S20, heuristic relocation is performed on each logical partition separately to move the idle overhead cranes in each logical partition to the optimal set of parking nodes (i.e., the target set of parking nodes) to minimize the minimum weighted distance cost between the demand node set and the idle overhead crane set.

[0040] In some implementations, such as Figure 2 As shown, step S20 includes: Step S21: For the candidate parking node set within each logical partition, select an initial parking node set from the candidate parking node set according to the selection probability value of each candidate parking node.

[0041] In this embodiment, the candidate parking node set refers to the set of all available locations for idle overhead cranes to park within a partition, including but not limited to: the demand node itself (equipment loading / unloading point), dedicated buffer zones, bypass rails, and waiting areas near switches. The selection probability value refers to the likelihood that each candidate parking node will be selected into the initial parking node set. This probability can be calculated based on the demand coverage of each candidate parking node and the shortest path distance between it and the already selected initial parking nodes. The initial parking node set refers to the set of parking nodes selected by the selection probability value that meets a predetermined number. The predetermined number can be the same as the number of idle overhead cranes within the current logical partition. .

[0042] Specifically, such as Figure 3 As shown, step S21, selecting an initial parking node set from the candidate parking node set based on the selection probability value of each candidate parking node, includes: Step S211: Select the demand node with the highest task occurrence probability weight in the demand node set as the first initial parking node.

[0043] That is, the probability value of selecting the first initial parking node. This equals the maximum task occurrence probability weight of the demand node. Wherein, the task occurrence probability weight... This refers to the demand node. Normalized value of the frequency of generating transport requests. Task occurrence probability weight for each demand node. It can be based on each demand node The data was calculated from historical data transfer requests.

[0044] For example, within the length of a sliding statistics window (e.g., the past 24 hours), each demand node within the current logical partition of the window is recorded. Number of transfer requests used as the starting point for transfer requests ,So Where H represents the set of demand nodes, This represents the total number of transport requests for the entire set of demand nodes.

[0045] It is important to note the probability weight of each task occurrence at each requirement node. It is dynamically updated over time and can be updated periodically based on each decision cycle or several decision cycles. For example, an exponentially weighted moving average can be used to update the weight value, thereby reflecting the changes in the probability of task occurrence at each demand node in real time, giving greater influence to recent observations, and allowing the relocation strategy of this embodiment to drift naturally with changes in demand distribution.

[0046] Step S212: Calculate the demand coverage of each candidate parking node and the shortest path distance between it and the currently selected initial parking node in sequence, and use the product of the demand coverage and the shortest path distance as the weighted distance.

[0047] Among them, demand coverage Used to measure candidate parking nodes The ability to cover surrounding demand nodes. Defined as a candidate parking node. The probability weight of task occurrence for all demand nodes within a certain distance range The sum, i.e., demand coverage The purpose of demand coverage is to prioritize candidate parking nodes that cover more demand nodes with a higher probability of occurrence. In the formula above, Represents the Gaussian kernel. For demand nodes To candidate parking node The shortest path distance, The Gaussian kernel bandwidth parameter is a preset parameter, and its value is positively correlated with the average coverage radius of candidate parking nodes in the current partition. For example, it can be 0.2 to 0.8 times the average shortest path length between a candidate parking node and a demand node. Demand coverage depends not only on the distance between the candidate parking node and the demand node, but also on the number of demand nodes covered by it and their weights, and the coverage intensity decreases exponentially with distance.

[0048] The shortest path distance between each candidate parking node and the currently selected initial parking node. , ,in Let s represent the set of currently selected initial parking nodes, and let s represent any initial parking node among the currently selected initial parking nodes. Represents each candidate parking node The shortest path distance between each candidate parking node and any of the currently selected initial parking nodes, and the shortest path distance between each candidate parking node and the currently selected initial parking node. That is, for each candidate parking node The minimum of the shortest path distances between the candidate parking node and any of the currently selected initial parking nodes (s). This value indicates the distance between the candidate parking node and the currently selected initial parking node.

[0049] Weighted distance It is the required coverage of each candidate parking node. and shortest path distance The product of, i.e. .

[0050] Step S213: Determine the selection probability value of each candidate parking node based on the normalized result of the weighted distance of each candidate parking node.

[0051] In this embodiment, a probability value is selected. ,in This represents the weighted sum of distances to candidate parking nodes u, excluding the currently selected initial parking node. V represents the set of candidate parking nodes. This represents the set of currently selected initial parking nodes.

[0052] Step S214: Select the remaining initial parking nodes from the candidate parking node set in descending order of the selection probability values ​​to form an initial parking node set.

[0053] In this embodiment, a probability value is selected. The larger the value, the higher the priority for the candidate parking node to be selected into the initial parking node set in the current decision. Selection probability value respectively with and They are positively correlated.

[0054] For example, if a candidate parking node is far from all currently selected initial parking nodes ( A relatively large number indicates that the surrounding area of ​​this candidate parking node is currently not covered by any parking nodes and should be prioritized for selection. This expands the spatial coverage of the initial parking node set, preventing multiple idle overhead cranes from clustering in the same small area. This allows the cranes to respond quickly from different directions when multiple demand nodes generate handling requests simultaneously, reducing the average pickup distance. If the candidate parking node is beneficial for serving demand nodes with a high probability of task occurrence (…), it should be prioritized. A larger value indicates that the crane parked at this candidate parking node can handle the transportation tasks of multiple demand nodes with a shorter travel distance, and should be given priority selection, thereby significantly reducing the average response time of high-probability requests.

[0055] However, if a candidate parking node has a high demand coverage ( (Larger), but very close to the currently selected initial parking node ( If the value is very small, it means that the area where the candidate parking node is located has already been covered by the currently selected initial parking node, and the marginal benefit of adding another initial parking node is low; conversely, if the demand coverage of a candidate parking node is not high ( Smaller), but far from all selected initial parking nodes ( If the value is extremely high, it indicates that the area where the candidate parking node is located is a new service area, which can avoid the excessive concentration of idle overhead cranes.

[0056] Therefore, choose the probability value The top-ranked nodes ensure that the initial parking nodes in the initial parking node set can cover the demand nodes with a high probability of task occurrence, while also ensuring the discreteness of the distribution of idle cranes.

[0057] Step S22: Calculate the weighted distance cost between the set of demand nodes in each logical partition and the initial set of parking nodes.

[0058] In this embodiment, the weighted distance cost is a comprehensive indicator calculated by weighted summation, taking into account the task occurrence probability weights of all demand nodes within each logical partition and the shortest path distance from each demand node to the initial parking node set. Mathematically, it can be expressed as follows: multiply the task occurrence probability weight of each demand node by the shortest path distance from that demand node to the nearest initial parking node in the initial parking node set, and then sum the products of all demand nodes. The sum obtained is the weighted distance cost corresponding to the initial parking node set.

[0059] In some implementations, such as Figure 4 As shown, step S22 includes: Step S221: For each set of demand nodes in each logical partition, calculate the shortest path distance from each demand node in the set of demand nodes to the initial set of parking nodes.

[0060] Specifically, for each demand node Calculate the distance from it to the initial parking node set. Each initial parking node The path distances are calculated, and the minimum value among these distances is denoted as . This means that if multiple overhead cranes are parked at different initial parking nodes in the initial parking node set, when a demand node generates a handling request, the system will dispatch the available overhead crane closest to that demand node. Therefore, the shortest path distance for actual pickup is this minimum value.

[0061] Step S222: The product of the shortest path distance from each demand node to the initial set of parking nodes and the task occurrence probability weight of each demand node is summed to obtain the weighted distance cost.

[0062] In this embodiment, weighted distance cost ,in, This indicates the calculation of each demand node. With the initial set of parking nodes The shortest path distance between them.

[0063] Step S23: Replace the initial parking node in the initial parking node set with the remaining parking nodes in the candidate parking node set to obtain an updated parking node set. Repeatedly calculate the weighted distance cost between the demand node set and the updated parking node set until the minimum weighted distance cost value is reached.

[0064] In some implementations, such as Figure 5 As shown, step S23 includes: Step S231: Traverse the remaining parking nodes in the candidate parking node set and the initial parking nodes in the initial parking node set. Replace an initial parking node with a remaining parking node to obtain an updated parking node set. Calculate the weighted distance cost between the demand node set and the updated parking node set.

[0065] Specifically, let the set of updated parking nodes be... The weighted distance cost between the demand node set and the updated parking node set is... .

[0066] Step S232: If the calculated weighted distance cost is less than the weighted distance cost corresponding to the initial parking node set, then the updated parking node set is used as the new initial parking node set.

[0067] Specifically, if Then update the set of parking nodes. As a new initial set of parking nodes .

[0068] Repeat step S231 until the minimum weighted distance cost value is reached, that is, the traversal of the remaining parking nodes in the candidate parking node set and the initial parking nodes in the initial parking node set is completed. At this time, the weighted distance cost corresponding to the latest initial parking node set is the minimum weighted distance cost value.

[0069] Step S24: The updated parking node set corresponding to the minimum weighted distance cost value is taken as the target parking node set.

[0070] In this embodiment, the target parking node set fully considers the task occurrence probability weight of each demand node, and avoids parking nodes clustering or coverage blind spots, thus forming a concentrated coverage of high-probability demand nodes and a reasonable balance of low-probability demand nodes within the partition.

[0071] Step S30: Perform a second relocation on the set of idle overhead cranes according to the set of target parking nodes, so as to relocate the idle overhead cranes in the set of idle overhead cranes to the target parking nodes in the set of target parking nodes.

[0072] In some implementations, the shortest travel distance from the current position (actual parking position before the second relocation) of each idle crane in the idle crane set to each target parking node in the target parking node set is calculated sequentially; and each idle crane is assigned a target parking node in the target parking node set in ascending order of the shortest travel distance, with only one idle crane assigned to each target parking node.

[0073] It should be noted that steps S10 to S30 above are executed cyclically in each decision cycle, forming a continuously optimized closed-loop control. By executing steps S10 to S30 in each decision cycle, a continuously optimized closed-loop control is formed. At the beginning of each decision cycle, the overhead crane control system obtains the actual number of idle overhead cranes in each logical partition and the actual position of each overhead crane, and sequentially performs the first relocation and the second relocation. This allows the number and position distribution of idle overhead cranes to continuously track the dynamic changes in the wafer fab process flow, significantly reducing the average pick-up time of production equipment and improving the overall response efficiency of the overhead crane system.

[0074] The crane relocation control method based on partitioned coordination proposed in this application, on the one hand, adopts a partitioned coordination crane relocation strategy to achieve dynamic balance optimization of the number and position of idle cranes among logical partitions. This ensures that partitions with high handling needs obtain sufficient idle crane resources, while partitions with low handling needs avoid crane backlog, and achieves crane resource balancing among partitions with minimal scheduling cost. This avoids unnecessary large-scale empty runs that waste energy and track resources, and improves the overall resource utilization and system throughput of the crane system.

[0075] On the other hand, by optimizing the parking node positions within the partition through micro-relocation within the partition, idle overhead cranes are concentrated at target parking nodes with high task occurrence probability and good coverage, which greatly reduces the average pickup distance and can significantly reduce the average pickup time of production equipment in the wafer fab process flow, thereby improving the overall production efficiency of the wafer fab process flow.

[0076] Therefore, the embodiments of this application achieve a balance between global scheduling efficiency and local precision from two levels: macro-level quantity balance and micro-level position optimization. This collaboratively reduces the overall response time of the overhead crane system to the handling requests of wafer fab production equipment, improves the utilization efficiency of overhead crane resources, avoids unnecessary long-distance empty runs, and ultimately improves the overall production efficiency of the wafer fab process.

[0077] In one embodiment, a computer storage medium is provided that stores executable instructions that, when executed by a processor, cause the processor to perform the steps in the above method embodiments.

[0078] In one embodiment, an electronic device is also provided, including one or more processors; and a memory storing one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors perform the steps in the above method embodiments.

[0079] In one embodiment, such as Figure 6 The diagram illustrates the structure of an electronic device used to implement an embodiment of this application. The electronic device includes a central processing unit (CPU) 101, which can perform various appropriate actions and processes based on a program stored in a read-only memory (ROM) 102 or a program loaded from a storage portion 108 into a random access memory (RAM) 103. The RAM 103 also stores various programs and data required for the operation of the electronic device. The CPU 101, ROM 102, and RAM 103 are interconnected via a bus 104. An input / output (I / O) interface 105 is also connected to the bus 104.

[0080] The following components are connected to I / O interface 105: an input section 106 including a keyboard, mouse, etc.; an output section 107 including a cathode ray tube (CRT), liquid crystal display (LCD), etc., and speakers, etc.; a storage section 108 including a hard disk, etc.; and a communication section 109 including a network interface card such as a LAN card, modem, etc. The communication section 109 performs communication processing via a network such as the Internet. A drive 110 is also connected to I / O interface 105 as needed. A removable medium 111, such as a disk, optical disk, magneto-optical disk, semiconductor memory, etc., is installed on drive 110 as needed so that computer programs read from it can be installed into storage section 108 as needed.

[0081] Specifically, according to embodiments of this application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of this application include a computer program product comprising a computer-readable medium carrying instructions that, in such embodiments, can be downloaded and installed from a network via communication section 109, and / or installed from removable medium 111. When the instructions are executed by central processing unit (CPU) 101, the various method steps described in this application are performed.

[0082] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of this application.

[0083] Furthermore, those skilled in the art will understand that although some embodiments herein include certain features included in other embodiments but not others, combinations of features from different embodiments are intended to be within the scope of this application and form different embodiments. For example, any of the embodiments or implementations claimed above can be used in any combination. The information disclosed in this background section is intended only to enhance the understanding of the general background of this application and should not be construed as an admission or in any way implying that such information constitutes prior art known to those skilled in the art.

Claims

1. A crane repositioning control method based on partitioned cooperation, characterized in that, The method includes: In each decision cycle, with the objective function of minimizing the total cost of scheduling idle overhead cranes among multiple logical partitions in the overhead crane system's track network, the number of idle overhead cranes that need to be scheduled among each logical partition is solved, and the first repositioning of the idle overhead cranes among the multiple logical partitions is performed based on the solution results. For each of the plurality of logical partitions, the target parking node set is determined based on the minimum weighted distance cost between the set of demand nodes and the set of idle overhead cranes within each logical partition; A second relocation is performed on the set of available overhead cranes based on the set of target parking nodes, so as to relocate the available overhead cranes in the set of available overhead cranes to the target parking nodes in the set of target parking nodes.

2. The crane repositioning control method based on partitioned collaboration according to claim 1, characterized in that, The objective function is to minimize the total cost of scheduling idle overhead cranes among multiple logical partitions in the overhead crane system's track network. The solution involves determining the number of idle overhead cranes that need to be scheduled between each logical partition, including: Based on the comparison results between the current number of idle cranes in each logical partition of the crane system's track network and the lower limit and upper limit of the number of idle cranes corresponding to each logical partition, the scarce partitions in the state of crane shortage and the surplus partitions in the state of crane surplus are determined. With the constraint of scheduling idle overhead cranes from the surplus partition to the shortage partition, and with the objective function of minimizing the total cost of scheduling idle overhead cranes, the number of idle overhead cranes that need to be scheduled between each logical partition is solved.

3. The crane repositioning control method based on partitioned collaboration according to claim 2, characterized in that, The constraint of scheduling idle overhead cranes from the surplus partition to the shortage partition includes: The first constraint is that the total number of idle overhead cranes transferred from the surplus zone does not exceed the surplus amount of idle overhead cranes corresponding to the surplus zone, and the second constraint is that the total number of idle overhead cranes transferred into the shortage zone does not exceed the shortage amount of idle overhead cranes corresponding to the shortage zone. The objective function of minimizing the total cost of scheduling idle overhead cranes includes minimizing the sum of the unit cost of scheduling a single idle overhead crane from one logical partition to another by multiplying it by the number of idle overhead cranes scheduled from one logical partition to another.

4. The crane repositioning control method based on partitioned collaboration according to claim 1, characterized in that, The step of determining the target parking node set based on the minimum weighted distance cost value between the demand node set and the idle crane set within each logical partition includes: For each logical partition, an initial set of parking nodes is selected from the set of candidate parking nodes based on the selection probability value of each candidate parking node. Calculate the weighted distance cost between the set of demand nodes within each logical partition and the initial set of parking nodes; Replace the initial parking node in the initial parking node set with the remaining parking nodes in the candidate parking node set to obtain an updated parking node set. Repeat the calculation of the weighted distance cost between the demand node set and the updated parking node set until the minimum weighted distance cost value is reached. The updated parking node set corresponding to the minimum weighted distance cost value is taken as the target parking node set.

5. The crane repositioning control method based on partitioned collaboration according to claim 4, characterized in that, The step of selecting an initial set of parking nodes from the set of candidate parking nodes based on the selection probability value of each candidate parking node includes: The demand node with the highest task occurrence probability weight in the set of demand nodes is selected as the first initial parking node. Calculate the demand coverage and the shortest path distance between each candidate parking node and the currently selected initial parking node in sequence, and use the product of the demand coverage and the shortest path distance as the weighted distance; The selection probability value of each candidate parking node is determined based on the normalized result of the weighted distance of each candidate parking node; The remaining initial parking nodes are selected sequentially from the candidate parking node set in descending order of the selection probability values ​​to form the initial parking node set.

6. The crane repositioning control method based on partitioned collaboration according to claim 4, characterized in that, The calculation of the weighted distance cost between the set of demand nodes within each logical partition and the initial set of parking nodes includes: For each set of demand nodes within a logical partition, calculate the shortest path distance from each demand node in the set of demand nodes to the initial set of parking nodes. The weighted distance cost is obtained by summing the products of the shortest path distance from each demand node to the initial set of parking nodes and the task occurrence probability weight of each demand node.

7. The crane repositioning control method based on partitioned collaboration according to claim 4, characterized in that, The step of replacing the initial parking nodes in the initial parking node set with the remaining parking nodes in the candidate parking node set to obtain an updated parking node set, and repeatedly calculating the weighted distance cost between the demand node set and the updated parking node set until the minimum weighted distance cost value is reached, includes: Traverse the remaining parking nodes in the candidate parking node set and the initial parking nodes in the initial parking node set. Replace an initial parking node with a remaining parking node to obtain an updated parking node set. Calculate the weighted distance cost between the demand node set and the updated parking node set. If the calculated weighted distance cost is less than the weighted distance cost corresponding to the initial parking node set, then the updated parking node set is used as the new initial parking node set. Repeat the process of traversing the remaining parking nodes in the candidate parking node set and the initial parking node in the initial parking node set, replacing an initial parking node with a remaining parking node to obtain an updated parking node set, until the minimum weighted distance cost value is reached.

8. The crane repositioning control method based on partitioned collaboration according to any one of claims 1 to 7, characterized in that, The second relocation of the set of idle overhead cranes based on the set of target parking nodes includes: Calculate the shortest travel distance from the current position of each idle crane in the idle crane set to each target parking node in the target parking node set in turn; According to the ascending order of the shortest travel distance, each idle crane is assigned a target parking node from the target parking node set, and each target parking node is assigned only one idle crane.

9. An electronic device, characterized in that, include: One or more processors; Memory, used to store one or more programs. When the one or more programs are executed by the one or more processors, the one or more processors perform the crane repositioning control method based on partition collaboration as described in any one of claims 1 to 8.

10. A computer storage medium, characterized in that, The storage medium stores executable instructions, which, when executed by a processor, cause the processor to perform the crane repositioning control method based on partition collaboration as described in any one of claims 1 to 8.

Images