Path weight optimization method and system for semiconductor carrying

Abstract

The invention discloses a path weight optimization method and system for semiconductor transportation, and relates to the technical field of semiconductors, and the method comprises the following steps: 1, in a statistical period, recording transportation time consumption data after each semiconductor transportation is finished, caching the transportation time consumption data to Redis, storing starting and ending points of each semiconductor transportation to a marking area, and caching the starting and ending points to Redis; 2, starting a timed task, judging whether data exist in a mark area or not, if not, stopping updating, and if yes, reading recent x pieces of carrying time consumption data from Redis to carry out path weight calculation, and storing the updated path weight in a database until the timed task is finished; according to the path weight optimization method and system, Redis is introduced as a hotspot data cache region, so that the system load is reduced, the performance is improved, and the response speed and the real-time performance of the system are improved; and meanwhile, the path weight can be dynamically updated, and the carrying efficiency is improved.

Description

Technical Field

[0001] The present invention relates to the field of semiconductor technology, and in particular to a path weight optimization method and system for semiconductor transportation. Background Art

[0002] In the semiconductor manufacturing industry, the material handling system needs to efficiently manage and optimize the transportation paths of raw materials, wafers, chemicals, semi-finished products and finished products between production equipment and warehouses. Since the semiconductor manufacturing process is complex and sophisticated, each step of the process has strict requirements for the timely and accurate transportation of materials. Any delays or errors in the transportation process may affect production efficiency and product quality. In addition, semiconductor factories usually operate 24 hours a day, and any optimization in path selection can significantly improve overall operational efficiency.

[0003] This innovative solution aims to use the most recent transport data to update the path weight, thereby more accurately calculating the path weight value. In semiconductor material transportation, path optimization is crucial to improving efficiency and reducing costs. Traditional methods may only select the optimal path based on path distance cost and point status, but this method may not fully take into account changes in actual transportation conditions.

[0004] Prior art CN116703290 A discloses a dynamic path retrieval method in a semiconductor handling scenario. In this existing solution, a path is selected based on a fixed weight calculation formula or rule, without considering the actual performance of different paths in different time periods or different situations. Therefore, it is impossible to adaptively adjust the path preference according to the actual situation, which may result in the selected path not being optimal. Moreover, this existing solution usually only considers the current status of the path, without considering the historical handling of the path. This may ignore the long-term performance and trend of the path, resulting in the selected path being not comprehensive or accurate enough. Summary of the invention

[0005] Based on the technical problems existing in the background technology, the present invention proposes a path weight optimization method and system for semiconductor transportation, which reduces the system load and improves performance, and improves the system response speed and real-time performance; at the same time, the path weight is dynamically updated to improve the transportation efficiency.

[0006] The present invention proposes a path weight optimization method for semiconductor transportation, comprising the following steps:

[0007] Step 1: In a statistical cycle, the transport time data after each semiconductor transport is completed is recorded and cached in Redis, and the start and end points of each semiconductor transport are saved in the mark area and cached in Redis;

[0008] Step 2: When the scheduled task starts, determine whether there is data in the marked area. If not, stop updating. If so, read the latest x pieces of transport time-consuming data from Redis to calculate the path weight, and store the updated path weight in the database until the scheduled task ends.

[0009] Furthermore, in step 1, if the records of the start and end points in Redis exceed a preset threshold x, then the records in Redis are deleted starting from the earliest one until they are reduced to the threshold x.

[0010] Furthermore, the start and end points are added as incremental data to the unordered Set of Redis. When the scheduled task detects incremental data, the Set pop-up method is used to immediately remove the earliest record from the Set while extracting the data, ensuring concurrency issues of other threads.

[0011] Furthermore, after the scheduled task starts, when judging whether there is data in the marked area, the marked area is traversed and scanned, and the start and end points of the records in the marked area are used as data for Redis to perform update operations on the database.

[0012] Furthermore, in step 2, specifically:

[0013] (a1) The timing task starts, and it is determined whether there is data in the marking area. If not, it goes to (a2); if so, it goes to (a3);

[0014] (a2) Scheduled task stops;

[0015] (a3) Distribute tasks to load-balanced microservice nodes;

[0016] (a4) extracting data from the marked area in batches, and determining whether the data is extracted. If so, proceed to (a5); otherwise, proceed to (a2);

[0017] (a5) Read the most recent x pieces of transport time data from Redis according to the configuration, calculate and store the path weight value, and then go to step (a4).

[0018] Furthermore, in (a3), tasks are distributed to load-balanced microservice nodes, specifically: messages are broadcast through Redis to each microservice node that subscribes to the message, and each microservice node creates multi-threaded concurrent processing tasks through the maximum number of single-threaded processing tasks configured by the system to achieve load-balanced task distribution.

[0019] Furthermore, in step 2, outliers are removed from the x pieces of transport time data based on the standard deviation and then the average value is calculated to obtain the final path weight; the specific calculation process of the path weight is as follows:

[0020] If x<3, then the x pieces of transport time data are all considered normal values, and the arithmetic mean of all normal values ​​is taken as the final weight;

[0021] If x≥3, and the i-th transport time consumption data Y i Satisfy | Y i -μ|≤kσ, k= Then Y i It is considered to be a normal value, and the arithmetic mean of all normal values ​​is taken as the final weight, where μ represents the arithmetic mean of the x pieces of transport time data, k represents the sensitivity, and σ represents the standard deviation.

[0022] A path weight optimization system for semiconductor transport, comprising a cache module and a timing update module;

[0023] The cache module is used to record the transport time data after each semiconductor transport is completed and cache it in Redis in a statistical cycle, and save the start and end points of each semiconductor transport to the mark area and cache it in Redis;

[0024] The timing update module is used to start the timing task and determine whether there is data in the marked area. If not, the update is stopped. If so, the most recent x pieces of transport time-consuming data will be read from Redis to calculate the path weight, and the updated path weight will be stored in the database until the timing task ends.

[0025] Furthermore, the timing update module is specifically used for:

[0026] (a1) The timing task starts, and it is determined whether there is data in the marking area. If not, it goes to (a2); if so, it goes to (a3);

[0027] (a2) Scheduled task stops;

[0028] (a3) Distribute tasks to load-balanced microservice nodes;

[0029] (a4) extracting data from the marked area in batches, and determining whether the data is extracted. If so, proceed to (a5); otherwise, proceed to (a2);

[0030] (a5) Read the most recent x pieces of transport time data from Redis according to the configuration, calculate and store the path weight value, and then go to step (a4).

[0031] The advantages of the path weight optimization method and system for semiconductor transportation provided by the present invention are: (1) reducing system load and improving performance: by introducing Redis as a hot data cache area, each transportation record is cached to Redis in time, which reduces the read and write pressure of the database, especially significantly improves the performance when a large amount of data weight is calculated, and reduces the system pressure. The scheduled task combines the scanning of the mark area and the task distribution of load balancing to avoid repeated calculation of all paths, improve the calculation efficiency, and further reduce the system load; (2) improving the system response speed and real-time performance: using Redis as a hot data cache area, the transportation record can be stored in Redis in real time, and the data can be quickly read from Redis when needed, thereby improving the system response speed and real-time performance. (3) more flexible system configuration and management: the configuration item x is introduced to control the number of records in Redis, making the system behavior more flexible and controllable. The setting and adjustment of the scheduled task also makes the management of the system more convenient, and the statistical cycle and task execution frequency can be flexibly adjusted according to actual needs. (4) based on the correct historical transportation time-consuming data, the path weight is dynamically calculated to obtain a more accurate path weight, thereby improving the transportation efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 It is a schematic diagram of the process of the present invention;

[0033] Figure 2 It is a structural schematic diagram of the present invention;

[0034] Figure 3 Optimize flow charts for path weights;

[0035] Figure 4 The data processing timing diagram when it is the path weight; DETAILED DESCRIPTION

[0036] Below, the technical solution of the present invention is described in detail through specific embodiments. Many specific details are set forth in the following description to facilitate a full understanding of the present invention. However, the present invention can be implemented in many other ways different from those described herein, and those skilled in the art can make similar improvements without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific implementation disclosed below.

[0037] Existing technologies mainly rely on path distance cost and path midpoint status to select the optimal path. Path distance cost is used as a weight value to evaluate the quality of the path, while the path midpoint status includes the availability information of each point to help determine the best path.

[0038] This embodiment introduces the historical transport conditions of the path. Specifically, by recording the latest x transport conditions of each path (i.e., each pair of starting and ending points), the usage frequency, efficiency, and other information of the path can be understood. Then, the weight value of the path distance cost is updated according to these transport conditions. In this way, when selecting the optimal path, not only the current path cost and status are considered, but also the historical usage of the path is considered, so as to more comprehensively evaluate the pros and cons of the path.

[0039] 1. Considering historical transport conditions: This embodiment introduces the historical transport conditions of the path, which makes the path selection more comprehensive, considering not only the current situation but also the past usage. This can help avoid the problem of congestion or reduced efficiency caused by frequent use of certain paths.

[0040] 2. Weight value update: The weight value of this embodiment is dynamically updated and adjusted according to the historical transportation situation of the path. This means that the system can adapt to the ever-changing working environment and needs, so as to more flexibly select the optimal path.

[0041] 3. Comprehensive consideration: This embodiment comprehensively considers the path distance cost and historical transport conditions, rather than relying solely on current status information. This can better balance the long-term and short-term benefits of the path and improve the accuracy and efficiency of path selection. The following specifically describes the path weight optimization for semiconductor transport.

[0042] like Figures 1 to 4 As shown, the present invention proposes a path weight optimization method for semiconductor transportation, comprising the following steps:

[0043] Step 1: In a statistical cycle, the transport time data after each semiconductor transport is completed is recorded and cached in Redis, and the start and end points of each semiconductor transport are saved in the mark area and cached in Redis;

[0044] In this embodiment, Redis is introduced as a hot data cache: after each transfer, the transfer time of the data is cached in Redis in time; if the records of the start and end points in Redis exceed the preset threshold x, the earliest data is deleted in Redis until it is reduced to the threshold x; each transfer record is stored in Redis, which can reduce the read and write pressure of the database. In particular, when calculating the weight of data, especially the weight of a large amount of data, the performance is significantly improved and the system pressure is reduced.

[0045] At the same time, when this implementation introduces Redis as a hot data cache, it is necessary to set the Redis configuration items: configure Redis in the system and set relevant parameters, such as the Redis server address, port number, etc. Data cache after the move: After each move, the system should capture relevant data (such as the time taken to move, etc.) and store these data in Redis. This can be done by calling the relevant Redis client library at the end of the move. Data cleaning exceeding the threshold: Regularly check whether the number of records at each starting and ending point in Redis exceeds the preset threshold x. If exceeded, delete the earliest record to keep the number of records below the threshold. This operation can be completed by setting a scheduled task in the system.

[0046] At the same time, data cache marks are introduced in this embodiment: after each transport is completed, the start and end points are recorded in the mark area; the mark area records the start and end points that have been updated after the previous statistical cycle, which can be a specific database table or a memory cache area, used to record the updated content that needs to be calculated in the next statistical cycle; therefore, these start and end points indicate that the path has updated content; these marked start and end points are used for calculation in the next statistical cycle.

[0047] Specifically, the Set unordered set structure in Redis is used to ensure that data is not repeated. The start and end points are added to the Set unordered set as incremental data. When the scheduled task detects incremental data, the Set pop-up method is used to immediately remove it from the Set while extracting the data, ensuring concurrency issues with other threads. Moreover, if new data still comes in during the extraction process, it can still be extracted by the scheduled task and added to the calculation, making the data more real-time.

[0048] Step 2: When the scheduled task starts, determine whether there is data in the marked area. If not, stop updating. If so, read the latest x pieces of transport time-consuming data from Redis to calculate the path weight, and store the updated path weight in the database until the scheduled task ends.

[0049] Scheduled task setting: (b1) Set up a scheduled task in the system to count and update the weight of the path (each starting and ending point). This task can be scheduled by the system scheduler to ensure scheduled execution. (b2) Scanning of the mark area: The scheduled task first scans the mark area to check whether there are records that need to be updated. Only the start and end points with records in the mark area will be updated from Redis to the database. Otherwise, the path will not be updated. This can be done by querying the database table of the mark area or reading the mark area data in memory. (b3) Update of path weight: For the start and end points that need to be updated, read the relevant data from Redis and calculate the path weight. If the data update is completed, the updated weight will be stored in the database.

[0050] Load-balanced task distribution: Based on the number of start and end points marked in the cycle and the number of system microservices, load-balanced task distribution is achieved. This can be accomplished by designing a task scheduler in the system to dynamically assign tasks to available microservice nodes based on the system load. Specifically, Redis broadcasts messages to each microservice node that subscribes to the message, and each service node creates multi-threaded concurrent processing tasks through the maximum number of single-threaded processing tasks configured by the system. Overall, tasks are processed through multiple service nodes and multiple threads under each node.

[0051] This embodiment introduces Redis as a hot data cache area, data cache markers, and enables scheduled tasks. This technical solution improves the accuracy and efficiency of path weight calculation, reduces system pressure, and improves system performance.

[0052] The prior art usually only considers the path distance cost, and this embodiment introduces the latest x pieces of transport time consumption data as the basis for weight calculation, which can more accurately reflect the actual situation of the path, thereby improving the accuracy of path weight calculation.

[0053] Due to the uncertainty in semiconductor transportation, there will always be some times when the transportation time is too long. If these data, which are temporarily called abnormal data, are added to the statistics, it will inevitably not objectively reflect the actual time consumption of the path. When the transportation speed and transportation distance remain unchanged, the random value of the transportation time roughly conforms to the normal distribution. For this reason, the standard deviation-based method is used to remove abnormal values ​​from the x transportation time data and then calculate the arithmetic mean to obtain the final path weight. The specific formula is as follows:

[0054] If x<3, then the x pieces of transport time data are all considered normal values, and the arithmetic mean of all normal values ​​is taken as the final weight;

[0055] If x≥3, and the i-th transport time consumption data Y i Satisfy | Y i -μ|≤kσ, k= Then Y i It is considered to be a normal value, and the arithmetic mean of all normal values ​​is taken as the final weight, where μ represents the arithmetic mean of the x pieces of transport time data, k represents the sensitivity, and σ represents the standard deviation.

[0056] Therefore, the weight updating process of the semiconductor transport method of this embodiment is as follows: Figure 3 As shown:

[0057] (a1) The timing task starts, and it is determined whether there is data in the marking area. If not, it goes to (a2); if so, it goes to (a3);

[0058] (a2) Scheduled task stops;

[0059] (a3) Distribute tasks to load-balanced microservice nodes;

[0060] (a4) extracting data from the marked area in batches, and determining whether the data is extracted. If so, proceed to (a5); otherwise, proceed to (a2);

[0061] (a5) Read the most recent x pieces of transport time data from Redis according to the configuration, calculate and store the path weight value, and then go to step (a4).

[0062] The main technical point of this embodiment is that multiple factors are taken into account in the process of calculating the path weight value, including the path distance cost and the time consumption of the latest x transportations. This method comprehensively considers the actual situation and historical data of the path, so that the weight value of the path can be evaluated more accurately. By updating the weight according to the path distance cost and the time consumption of the latest x transportations, this embodiment solves the problem of ignoring the path historical data in the traditional path weight calculation method. It can more accurately reflect the actual situation of the path, thereby improving the accuracy and efficiency of path planning. The technical effect of this embodiment is that it can provide a more accurate weight value for the path planning system, so that the system can better cope with the path planning needs in various complex situations. It can improve the efficiency and performance of logistics systems, traffic management systems and other fields, and bring significant advantages and benefits to practical applications.

[0063] Therefore, this embodiment has the following advantages: (1) Reducing system load and improving performance: By introducing Redis as a hot data cache, each transport record is cached to Redis in time, which reduces the read and write pressure of the database, especially when calculating the weight of a large amount of data, the performance is significantly improved, and the system pressure is reduced. The scheduled task combines the scanning of the mark area and the task distribution of load balancing to avoid repeated calculations of all paths, improve the calculation efficiency, and further reduce the system load; (2) Improving system response speed and real-time performance: Using Redis as a hot data cache, the transport record can be stored in Redis in real time, and data can be quickly read from Redis when needed, thereby improving the response speed and real-time performance of the system. (3) More flexible system configuration and management: The configuration item x is introduced to control the number of records in Redis, making the behavior of the system more flexible and controllable. The setting and adjustment of scheduled tasks also makes the management of the system more convenient, and the statistical cycle and task execution frequency can be flexibly adjusted according to actual needs. (4) Dynamically calculate the path weight based on the correct historical transport time-consuming data to obtain a more accurate path weight, thereby improving the transport efficiency.

[0064] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A path weight optimization method for semiconductor transport, characterized in that: The steps include: Step 1: In a statistical cycle, the transport time data after each semiconductor transport is completed is recorded and cached in Redis, and the start and end points of each semiconductor transport are saved in the mark area and cached in Redis; Step 2: When the scheduled task starts, determine whether there is data in the marked area. If not, stop updating. If so, read the latest x pieces of transport time-consuming data from Redis to calculate the path weight, and store the updated path weight in the database until the scheduled task ends.

2. The path weight optimization method for semiconductor transport according to claim 1, characterized in that: In step 1, if the records of the start and end points in Redis exceed a preset threshold x, then the records in Redis are deleted starting from the earliest one until they are reduced to the threshold x.

3. The path weight optimization method for semiconductor transport according to claim 2, characterized in that: The start and end points are added as incremental data to the unordered Set of Redis. When the scheduled task detects incremental data, the Set pop-up method is used to immediately remove the earliest record from the Set while extracting the data, ensuring concurrency issues of other threads.

4. The path weight optimization method for semiconductor transport according to claim 1, characterized in that: After the scheduled task starts, when judging whether there is data in the marked area, the marked area is traversed and scanned, and the start and end points of the records in the marked area are used as the data for Redis to perform update operations on the database.

5. The path weight optimization method for semiconductor transport according to claim 1, characterized in that: In step 2, specifically: (a1) The timing task starts, and it is determined whether there is data in the marking area. If not, it goes to (a2); if so, it goes to (a3); (a2) Scheduled task stops; (a3) Distribute tasks to load-balanced microservice nodes; (a4) extracting data from the marked area in batches, and determining whether the data is extracted. If so, proceed to (a5); otherwise, proceed to (a2); (a5) Read the most recent x pieces of transport time data from Redis according to the configuration, calculate and store the path weight value, and then go to step (a4).

6. The path weight optimization method for semiconductor transport according to claim 5, characterized in that: In (a3), tasks are distributed to load-balanced microservice nodes, specifically: Redis broadcasts messages to each microservice node that subscribes to the message, and each microservice node creates multi-threaded concurrent processing tasks through the maximum number of single-threaded processing tasks configured by the system to achieve load-balanced task distribution.

7. The path weight optimization method for semiconductor transport according to claim 1, characterized in that: In step 2, outliers are removed from the x pieces of transport time data based on the standard deviation, and then the average value is calculated to obtain the final path weight. The specific calculation process of the path weight is as follows: If x<3, then the x pieces of transport time data are all considered normal values, and the arithmetic mean of all normal values ​​is taken as the final weight; If x≥3, and the i-th transport time consumption data Y i Satisfy | Y i -μ|≤kσ, Then Y i It is considered to be a normal value, and the arithmetic mean of all normal values ​​is taken as the final weight, where μ represents the arithmetic mean of the x pieces of transport time data, k represents the sensitivity, and σ represents the standard deviation.

8. A path weight optimization system for semiconductor transport, characterized in that: Including cache module and timed update module; The cache module is used to record the transport time data after each semiconductor transport is completed and cache it in Redis in a statistical cycle, and save the start and end points of each semiconductor transport to the mark area and cache it in Redis; The timing update module is used to start the timing task and determine whether there is data in the marked area. If not, the update is stopped. If so, the most recent x pieces of transport time-consuming data will be read from Redis to calculate the path weight, and the updated path weight will be stored in the database until the timing task ends.

9. The path weight optimization system for semiconductor transport according to claim 1, characterized in that: The timing update module is specifically used for: (a1) The timing task starts, and it is determined whether there is data in the marking area. If not, it goes to (a2); if so, it goes to (a3); (a2) Scheduled task stops; (a3) Distribute tasks to load-balanced microservice nodes; (a4) extracting data from the marked area in batches, and determining whether the data is extracted. If so, proceed to (a5); otherwise, proceed to (a2); (a5) Read the most recent x pieces of transport time data from Redis according to the configuration, calculate and store the path weight value, and then go to step (a4).

Images