Workshop task allocation method and system based on multiple dimensions

Abstract

The invention discloses a workshop task allocation method and system based on multiple dimensions, and relates to the technical field of production control management. Comprising the steps of obtaining order data and then performing preprocessing; based on the preprocessed order data, an advanced machine learning algorithm is adopted to construct a decision tree model, and the processing complexity of each order is identified; determining a task allocation scheme by using an optimization algorithm according to the identified order processing complexity and the current state of the production line; monitoring the actual operation condition of each production line in real time, starting a dynamic adjustment mechanism when the work load of a certain production line is unbalanced, and re-evaluating and adjusting task distribution; through intelligent task allocation and resource optimization, the production efficiency and customer satisfaction are improved, and meanwhile, dynamic adjustment and intelligent management of the production process are realized.

Description

Technical Field

[0001] The present invention relates to the technical field of production control management, and in particular to a multi-dimensional workshop task allocation method and system. Background Art

[0002] In recent years, with the development of artificial intelligence technology, intelligent scheduling algorithms have gradually been applied to workshop task allocation, improving the accuracy and real-time performance of task allocation. However, most of the existing intelligent scheduling algorithms only consider a single indicator and fail to fully consider the impact of multiple aspects such as production speed and product quality. Therefore, there are still certain limitations in practical applications. Summary of the invention

[0003] The purpose of the present invention is to provide a multi-dimensional workshop task allocation method and system, which improves production efficiency and customer satisfaction through intelligent task allocation and resource optimization, and realizes dynamic adjustment and intelligent management of the production process.

[0004] This application provides a multi-dimensional workshop task allocation method, comprising the following steps: Obtain order data from different channels, then use data cleaning and standardization techniques to pre-process the raw data and construct feature vectors; Based on the pre-processed order data, an advanced machine learning algorithm is used to build a decision tree model to identify the processing complexity of each order; Based on the identified order processing complexity and the current status of the production line, an optimization algorithm is used to determine the task allocation plan; Monitor the actual operation status of each production line in real time. When a production line has an unbalanced workload, activate the dynamic adjustment mechanism to re-evaluate and adjust task allocation. By collecting the operation data of the production line, the decision tree model and task allocation strategy are optimized.

[0005] Furthermore, data cleaning and standardization techniques are used to preprocess the raw data and construct feature vectors, including: Obtain order data, classify it according to the order source channel, and obtain a collection of channel-specific order data; extract key attributes from the channel-specific order data to form structured order attribute data; Data cleaning technology is used to remove duplicates, noise and outliers from order attribute data to obtain cleaned order attribute data; through data standardization, the cleaned order attribute data is converted into a unified format and representation to form standardized order attribute data; Based on the standardized order attribute data, the order feature vector is constructed as the input of the machine learning algorithm; the machine learning model is trained and optimized to perform order classification and prediction tasks based on the order feature vector; Combine the output of the machine learning model with business rules to automatically generate order processing decisions and form an intelligent order management solution.

[0006] Furthermore, an advanced machine learning algorithm is used to build a decision tree model, including: According to the preprocessed order data, the feature vector of each order is obtained, and the feature vector includes various information required for order processing; The decision tree algorithm is used to train the order feature vector to obtain a decision tree model, which determines the processing complexity based on the order features. For new orders, extract their feature vectors and input them into the decision tree model. The model makes judgments along the branch conditions of the decision tree based on the dimensional values ​​of the feature vectors to obtain the processing complexity of the order. When the processing complexity of an order exceeds the preset threshold, the order is marked as a high-complexity order and requires special processing; For orders with processing complexity below the threshold, they will directly enter the normal processing process; Dynamically adjust the allocation of production resources according to the processing complexity of the order.

[0007] Furthermore, the order feature vector is trained using a decision tree algorithm to obtain a decision tree model. The model determines the processing complexity based on the order features, specifically including: By analyzing the various features in the order data set, a tree model is constructed. The processing complexity of the order is predicted through binary selection based on the input feature vector. During the training process, the segmentation points representing the data features are selected to form the nodes of the tree. Then, according to the decision path of the nodes, the data is divided into different areas, and each area corresponds to a prediction result. The decision tree model can make judgments along the branch conditions of the tree based on the feature vector of the new order.

[0008] Furthermore, an optimization algorithm is used to determine the task allocation plan, including: Obtain the processing complexity information of the order and use it as one of the input parameters of the optimization algorithm; monitor the operating status of the production line in real time and use it as another set of input parameters of the optimization algorithm; build an optimization model based on the order complexity and production line status; use a heuristic algorithm to search for task allocation solutions to obtain the processing sequence and time schedule of each order in each process; when it is impossible to converge to a satisfactory solution in a short time, introduce a local search strategy to explore again in the current vicinity; convert the optimization results into an executable production plan and send it to the workshop management system and equipment control system; continuously monitor the production execution, compare the optimization plan with the actual output, and when deviations are found, trigger the algorithm to recalculate and adjust the task allocation.

[0009] Furthermore, the heuristic algorithm is used to search for a task allocation scheme, and the task allocation scheme is expressed as: ,in, is a decision variable vector, representing the task allocation scheme, is the objective function, which represents the performance index to be optimized. is an inequality constraint function, which represents various resource constraints. is an equality constraint function that represents a specific production requirement or equilibrium condition that must be met. represents the task allocation scheme, Used to introduce constraints.

[0010] Furthermore, a dynamic adjustment mechanism is initiated to re-evaluate and adjust task allocation, including: Obtain the real-time operation data of each production line, and analyze and process the data in real time; By analyzing the production line operation data, calculate the workload index of each production line to determine whether there is load imbalance; When it is found that the load index of a production line is greatly different from that of other production lines and exceeds the preset threshold range, the dynamic adjustment mechanism is triggered; Based on the real-time status and historical data of the production line, a machine learning algorithm is used to establish a load balancing optimization model to evaluate the effectiveness of different task allocation schemes; Input the current task allocation situation of the production line into the optimization model, and obtain the task allocation plan through model calculation and analysis; According to the optimized task allocation plan, dynamically adjust the task allocation of the production line, transfer tasks from the high-load production line to the low-load production line, and balance the workload of each production line; Continuously monitor the operating status of the production line, and make dynamic adjustments and optimizations based on real-time data and feedback from optimization models.

[0011] Furthermore, through model calculation and analysis, a task allocation scheme is obtained, and load balancing optimization is performed through a load strategy, which is expressed as: ,in Indicates The time allocated for each task, Indicates in all Find the load value of the load in the computing resource, is the time interval or polling period for task allocation, is the assigned time or start time of the first task.

[0012] Furthermore, the decision tree model and task allocation strategy are optimized, including: By deploying sensors and data acquisition equipment on the production line, the data of the production line is continuously collected to build a production line operation database; the collected production line operation data is preprocessed to obtain a training data set suitable for the machine learning algorithm; according to the characteristics of the production line and the optimization goals, the machine learning algorithm is selected to build a production line optimization model; the machine learning model is trained using the training data set, and the model hyperparameters are optimized through cross-validation and grid search; the trained machine learning model is deployed to the control system of the production line, and the production efficiency and quality are predicted based on the real-time collected production line operation data, and the task allocation strategy and operation parameters of the production line are dynamically adjusted; the operation status and optimization effect of the production line are continuously monitored, new operation data is collected, and the machine learning model is regularly retrained and optimized; the efficiency and quality comparison before and after the production line optimization is displayed through visualization tools.

[0013] The present invention provides a multi-dimensional workshop task allocation system, which is used to implement a multi-dimensional workshop task allocation method, specifically comprising: The data collection module obtains order data from different channels, classifies them according to the order source channels, obtains channel-specific order data sets, extracts key attributes, and forms structured order attribute data. The machine learning model training module selects and trains suitable machine learning algorithms, builds optimization models, trains models using preprocessed data sets, and optimizes model hyperparameters using cross-validation and grid search; The task allocation optimization module automatically generates order processing decisions based on the output results of the optimization model and business rules, and then dynamically adjusts the task allocation of the production line to perform load balancing and rational use of resources; The monitoring and visualization module is used to continuously monitor the operating status and optimization effect of the production line, collect new operating data, and regularly retrain and optimize the machine learning model. It then uses visualization tools to display the efficiency and quality comparison before and after the production line optimization.

[0014] The beneficial effects of the present invention are: By adopting advanced machine learning algorithms and heuristic algorithms, this method can accurately identify the processing complexity of each order and dynamically adjust task allocation according to the real-time production status. This intelligent scheduling mechanism enables the production line to reasonably allocate production resources with the goal of minimizing production time and maximizing resource utilization. For high-complexity orders, more resources are allocated to ensure quality and progress; for low-complexity orders, resource allocation is reduced to avoid waste. This not only improves the overall efficiency of the production line, but also ensures the rational use of resources, reduces idleness and overload, thereby achieving cost savings and optimization of production processes. By combining the output results of the machine learning model with business rules, this method can automatically generate order processing decisions, such as priority sorting, resource allocation, and risk warning, forming an intelligent order management solution. This automated and intelligent processing method greatly reduces human errors and delays, and improves the accuracy and real-time performance of order processing. At the same time, by continuously monitoring the operating status of the production line and performing dynamic optimization, this method can quickly respond to changes in the production process, adjust production plans in a timely manner, and ensure that orders are completed on time, which not only improves customer satisfaction, but also enhances the company's market competitiveness. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] For better understanding and implementation, the technical solution of the present application is described in detail below with reference to the accompanying drawings.

[0016] Figure 1 A flowchart of a multi-dimensional workshop task allocation method provided in Example 1 of the present application; Figure 2 A schematic diagram of a flow chart of preprocessing raw data and constructing feature vectors in a multi-dimensional workshop task allocation method provided in Example 1 of the present application; Figure 3 A structural diagram of a multi-dimensional workshop task allocation system provided in Example 2 of the present application. DETAILED DESCRIPTION

[0017] In order to further explain the technical means and effects taken by the present invention to achieve the predetermined invention purpose, exemplary embodiments will be described in detail here, and examples thereof are shown in the accompanying drawings. When the following description refers to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The implementations described in the following exemplary embodiments do not represent all implementations consistent with the present application. Instead, they are only examples of methods and systems consistent with some aspects of the present application as detailed in the attached claims.

[0018] The terms used in this application are for the purpose of describing specific embodiments only and are not intended to limit this application. The singular forms of "a", "said" and "the" used in this application and the appended claims are also intended to include plural forms unless the context clearly indicates other meanings. It should also be understood that the term "and / or" used herein refers to and includes any or all possible combinations of one or more associated listed items.

[0019] The specific implementation methods, features and effects of the present invention are described in detail below in conjunction with the accompanying drawings and preferred embodiments.

[0020] Example 1 See also Figure 1-Figure 2 This embodiment provides a multi-dimensional workshop task allocation method, comprising the following steps: S1. Obtain order data from different channels, including but not limited to order quantity, product type, delivery deadline, etc.; then use data cleaning and standardization technology to pre-process the raw data and construct feature vectors; Furthermore, data cleaning and standardization techniques are used to preprocess the raw data and construct feature vectors, including: S11. Obtain order data, classify them according to the order source channels, and obtain a set of channel-specific order data; extract key attributes such as order quantity, product type, and delivery deadline from the channel-specific order data to form structured order attribute data; S12. Using data cleaning technology to remove duplication, remove noise, and process outliers on the order attribute data to obtain cleaned order attribute data; through data standardization processing, converting the cleaned order attribute data into a unified format and representation to form standardized order attribute data; S13. Construct order feature vectors based on standardized order attribute data as input to the machine learning algorithm; train and optimize machine learning models, such as decision trees, support vector machines, neural networks, etc., to perform tasks such as order classification and prediction based on order feature vectors; S14. Combine the output results of the machine learning model with business rules to automatically generate order processing decisions, such as priority sorting, resource allocation, risk warning, etc., to form an intelligent order management solution.

[0021] Specifically, we collect and classify order data from different channels, and then extract key attributes such as order quantity, product type and delivery period to form structured order attribute data. Then, we use data cleaning technology to remove duplication, noise and outliers, and convert the data into a unified format through standardization processing to construct feature vectors for machine learning algorithms. By training and optimizing machine learning models such as decision trees, support vector machines, and neural networks, we can classify and predict orders based on these feature vectors. Finally, we combine the output of the model with business rules to automatically generate order processing decisions and realize functions such as priority sorting, resource allocation and risk warning, thereby forming a set of intelligent order management solutions to improve production efficiency and customer satisfaction.

[0022] S2. Based on the pre-processed order data, an advanced machine learning algorithm is used to build a decision tree model to identify the processing complexity of each order; Furthermore, an advanced machine learning algorithm is used to build a decision tree model, including: According to the preprocessed order data, the feature vector of each order is obtained, and the feature vector includes various information required for order processing; The decision tree algorithm is used to train the order feature vector to obtain a decision tree model, which can determine the processing complexity based on the order features; For new orders, extract their feature vectors and input them into the decision tree model. The model makes judgments along the branch conditions of the decision tree based on the values ​​of each dimension of the feature vector, and finally obtains the processing complexity of the order. If the processing complexity of an order exceeds a preset threshold, the order is marked as a high-complexity order that requires special processing; For orders with processing complexity below the threshold, they will directly enter the normal processing process; Dynamically adjust the allocation of production resources according to the processing complexity of the order, allocate more resources to high-complexity orders, and allocate fewer resources to low-complexity orders, so as to achieve rational use of resources; By predicting the complexity of order processing and dynamically allocating resources, we can improve production efficiency, shorten order processing time and enhance customer satisfaction.

[0023] Specifically, by using advanced machine learning algorithms to build a decision tree model, the processing complexity of each order can be identified based on the feature vectors in the preprocessed order data, such as the various information required for order processing. This model can evaluate the feature vectors of new orders and make judgments along the branch conditions of the decision tree, thereby accurately predicting the processing complexity of each order. Based on this prediction result, orders with complexity exceeding the preset threshold can be marked as high-complexity orders and given special treatment, while orders below the threshold can be included in the normal processing process. In addition, the allocation of production resources can be dynamically adjusted according to the processing complexity of the order, ensuring that high-complexity orders obtain more resources, while low-complexity orders are allocated fewer resources, thereby achieving rational use of resources. This method not only improves production efficiency and shortens order processing time, but also improves customer satisfaction, providing an intelligent solution for workshop task allocation.

[0024] Furthermore, the order feature vector is trained using a decision tree algorithm to obtain a decision tree model. The model can determine the processing complexity based on the order features, including: By analyzing the various features in the order data set (such as the required processes, processing time, resource consumption, etc.), a tree model is constructed. The processing complexity of the order is predicted through binary selection based on the input feature vector (that is, the specific information of each order). During the training process, the segmentation points representing the data features are selected to form the nodes of the tree. Then, according to the decision path of the nodes, the data is divided into different areas, and each area corresponds to a prediction result. The decision tree model can make judgments along the branch conditions of the tree based on the feature vector of the new order, and quickly and accurately give an evaluation of the processing complexity of the order.

[0025] S3. Based on the identified order processing complexity and the current status of the production line, such as capacity utilization and production speed, an optimization algorithm is used to determine the optimal task allocation plan; Furthermore, an optimization algorithm is used to determine the optimal task allocation solution, including: Obtain the processing complexity information of the order, including the required processes, processing time, resource consumption, etc., and use it as one of the input parameters of the optimization algorithm; Real-time monitoring of the operating status of the production line, collecting data such as capacity utilization and production speed as another set of input parameters for the optimization algorithm; According to the order complexity and production line status, an optimization model is constructed, and the objective function is set to minimize production time or maximize resource utilization, while considering constraints such as equipment and personnel; Use heuristic algorithms, such as genetic algorithms or ant colony algorithms, to search for the optimal task allocation solution and obtain the processing sequence and time schedule for each order in each process; If the algorithm cannot converge to a satisfactory solution in a short time, a local search strategy is introduced to conduct in-depth exploration near the current optimal solution to further improve the quality of the solution; Convert the optimization results into executable production plans and send them to the workshop management system and equipment control system to guide the actual production process; Continuously monitor production execution and compare the optimization plan with actual output. If significant deviations are found, the algorithm will be triggered to recalculate and adjust task allocation to achieve dynamic optimization.

[0026] Specifically, by using optimization algorithms, the optimal task allocation scheme can be determined based on the processing complexity information of each order and the real-time status of the production line, such as capacity utilization and production speed. This process involves building an optimization model that aims to minimize production time or maximize resource utilization, and takes into account constraints such as equipment and personnel. Using heuristic algorithms such as genetic algorithms or ant colony algorithms, the optimal task allocation scheme can be searched to determine the processing sequence and time schedule of each order in each process. If the algorithm fails to converge to a satisfactory solution quickly, a local search strategy will be introduced to improve the quality of the scheme. The optimization results are then converted into specific production plans to guide the actual production process, and dynamic optimization is achieved by continuously monitoring production execution and comparing the optimization scheme with actual output. This method can ensure maximum production efficiency and optimal resource allocation, while shortening order processing time, improving customer satisfaction, and realizing intelligent management of the production process.

[0027] Furthermore, the heuristic algorithm is used to search for the optimal task allocation solution, and the optimal task allocation solution is expressed as: ,in, is a decision variable vector, representing the task allocation scheme, is the objective function, which represents the performance indicator to be optimized, such as minimizing production time or maximizing resource utilization. is an inequality constraint function, which represents various resource constraints. is an equality constraint function that represents a specific production requirement or equilibrium condition that must be met. represents the optimal task allocation solution, Used to introduce constraints.

[0028] Specifically, the use of heuristic algorithms, such as genetic algorithms or ant colony algorithms, can find the optimal task allocation plan, that is, optimize performance indicators (such as minimizing production time or maximizing resource utilization) while satisfying resource constraints such as equipment and personnel and specific production requirements or balance conditions, thereby achieving improved production efficiency, reasonable allocation of resources, improved customer satisfaction, and dynamic adjustment and intelligent management of the production process.

[0029] S4. Monitor the actual operation of each production line in real time. When a production line has an unbalanced workload, start the dynamic adjustment mechanism to re-evaluate and adjust the task allocation; Furthermore, a dynamic adjustment mechanism is initiated to re-evaluate and adjust task allocation, including: Obtain real-time operation data of each production line, including equipment status, production tasks, output and other information, and perform real-time analysis and processing of the data; By analyzing the production line operation data, calculate the workload indicators of each production line, such as equipment utilization rate, task completion rate, etc., to determine whether there is load imbalance; If it is found that the load index of a production line is significantly different from that of other production lines and exceeds the preset threshold range, the dynamic adjustment mechanism will be triggered; Based on the real-time status and historical data of the production line, machine learning algorithms such as decision trees and support vector machines are used to establish a load balancing optimization model to evaluate the effects of different task allocation schemes. Input the current task allocation of the production line into the optimization model, and obtain a better task allocation plan through model calculation and analysis to achieve load balancing of the production line; According to the optimized task allocation plan, dynamically adjust the task allocation of the production line, transfer tasks from the high-load production line to the low-load production line, and balance the workload of each production line; Continuously monitor the operating status of the production line, and make dynamic adjustments and optimizations based on real-time data and feedback from optimization models to ensure load balance and efficient operation of the production line.

[0030] Specifically, by real-time monitoring of the operating data of each production line, including equipment status, production tasks, and output, the workload indicators of the production line, such as equipment utilization and task completion rate, can be dynamically evaluated. Once load imbalance is found, that is, the load indicator of a production line exceeds the preset threshold, the dynamic adjustment mechanism will be triggered; the load balancing optimization model established by machine learning algorithms can evaluate the effects of different task allocation schemes based on the real-time status and historical data of the production line, and calculate a better task allocation scheme to achieve load balancing; then, according to this optimization scheme, dynamically adjust the task allocation, transfer tasks from high-load production lines to lower-load production lines, and balance the workload. At the same time, the operating status of the production line is continuously monitored, and dynamic adjustments and optimizations are continuously made based on real-time data and model feedback to ensure that the production line always maintains load balance and efficient operation; this process not only improves production efficiency, but also enhances the adaptability and flexibility of the production line to changes.

[0031] Furthermore, through model calculation and analysis, a better task allocation scheme is obtained, and load balancing optimization is performed through the minimum load strategy, which is expressed as: ,in Indicates The time allocated for each task, Indicates in all Find the load value of the resource with the lowest load among the computing resources. is the time interval or polling period for task allocation, is the assigned time or start time of the first task.

[0032] For each task, find the computing resource with the lowest load, then multiply this minimum load by the time interval, and finally add the start time to determine the task allocation time. This method helps to achieve load balancing and ensure that tasks are evenly distributed to all computing resources.

[0033] S5. Continuously optimize the decision tree model and task allocation strategy by collecting production line operation data.

[0034] Furthermore, the decision tree model and task allocation strategy are optimized, including: By deploying sensors and data acquisition equipment on the production line, we continuously collect data such as production line operation parameters, equipment status, product quality, etc., and build a production line operation database; Preprocess the collected production line operation data, including data cleaning, feature extraction, data standardization, etc., to obtain a training data set suitable for machine learning algorithms; According to the characteristics of the production line and the optimization objectives, select appropriate machine learning algorithms, such as decision trees, random forests, support vector machines, etc., to build a production line optimization model; Use the training data set to train the machine learning model, optimize the model hyperparameters through cross-validation and grid search, and obtain the production line optimization model with the best performance; Deploy the trained machine learning model to the control system of the production line, predict production efficiency and quality based on the real-time collected production line operation data, and dynamically adjust the task allocation strategy and operation parameters of the production line; Continuously monitor the operating status and optimization effect of the production line, collect new operating data, regularly retrain and optimize the machine learning model, and continuously improve the efficiency and quality of the production line; Visual tools are used to display the efficiency and quality comparison before and after production line optimization, as well as the effect differences of different task allocation strategies, to provide decision support for production managers and continuously improve production line operation management.

[0035] Specifically, by deploying sensors and data acquisition equipment on the production line, we continuously collect data such as operating parameters, equipment status, and product quality, and build a production line operation database. After preprocessing, these data form a data set suitable for machine learning algorithm training. Select a suitable machine learning algorithm, build and train the production line optimization model, optimize the model hyperparameters through methods such as cross-validation and grid search to obtain the model with the best performance, and deploy these models to the control system, which can predict production efficiency and quality based on real-time data, and dynamically adjust task allocation strategies and operating parameters. At the same time, we continuously monitor the operating status and optimization effects, and regularly retrain and optimize the model to improve the efficiency and quality of the production line. In addition, through visualization tools, the comparison before and after optimization and the difference in the effects of different strategies are displayed to provide decision support for production managers and achieve continuous improvement in production line operation management. This process not only improves production efficiency and product quality, but also enhances the adaptability and intelligence level of the production line, providing strong data support for production decisions.

[0036] Example 2 See also Figure 3 This embodiment provides a multi-dimensional workshop task allocation system, which is used to implement a multi-dimensional workshop task allocation method, specifically including: The data collection module obtains order data from different channels, classifies them according to the order source channels, obtains channel-specific order data sets, extracts key attributes, and forms structured order attribute data. The machine learning model training module selects and trains suitable machine learning algorithms, builds optimization models, trains models using preprocessed data sets, and optimizes model hyperparameters using methods such as cross-validation and grid search to achieve optimal performance; The task allocation optimization module automatically generates order processing decisions, such as priority sorting, resource allocation, and risk warning, based on the output results of the optimization model and business rules, and then dynamically adjusts the task allocation of the production line to achieve load balancing and rational use of resources; The monitoring and visualization module is used to continuously monitor the operating status and optimization effect of the production line, collect new operating data, and regularly retrain and optimize the machine learning model. It then uses visualization tools to display the efficiency and quality comparison before and after the production line optimization, as well as the effect differences of different task allocation strategies, to provide decision support for production managers and help them continuously improve the operation management of the production line.

[0037] The above description is only a preferred embodiment of the present invention and does not limit the present invention in any form. Although the present invention has been disclosed as a preferred embodiment as above, it is not used to limit the present invention. Any technical personnel in this field can make some changes or modify the technical contents disclosed above into equivalent embodiments without departing from the scope of the technical solution of the present invention. However, any brief modifications, equivalent changes and modifications made to the above embodiments based on the technical essence of the present invention without departing from the content of the technical solution of the present invention are still within the scope of the technical solution of the present invention.

Claims

1. A multi-dimensional workshop task allocation method, characterized by: The steps include: Obtain order data from different channels, then use data cleaning and standardization techniques to pre-process the raw data and construct feature vectors; Based on the pre-processed order data, an advanced machine learning algorithm is used to build a decision tree model to identify the processing complexity of each order; Based on the identified order processing complexity and the current status of the production line, an optimization algorithm is used to determine the task allocation plan; Monitor the actual operation status of each production line in real time. When a production line has an unbalanced workload, activate the dynamic adjustment mechanism to re-evaluate and adjust task allocation. The dynamic adjustment mechanism is started to re-evaluate and adjust the task allocation, including: obtaining real-time operation data of each production line, and performing real-time analysis and processing on the data; By analyzing the production line operation data, the workload indicators of each production line are calculated to determine whether there is a load imbalance. When it is found that the load indicator of a production line is greatly different from that of other production lines and exceeds the preset threshold range, the dynamic adjustment mechanism is triggered. According to the real-time status and historical data of the production line, a machine learning algorithm is used to establish a load balancing optimization model to evaluate the effects of different task allocation schemes. The current task allocation situation of the production line is input into the optimization model, and the task allocation plan is obtained through model calculation and analysis. According to the optimized task allocation plan, the task allocation of the production line is dynamically adjusted to transfer tasks from high-load production lines to low-load production lines to balance the workload of each production line. The operating status of the production line is continuously monitored, and dynamic adjustments and optimizations are continuously made based on real-time data and feedback from the optimization model.

2. The multi-dimensional workshop task allocation method according to claim 1 is characterized in that: Data cleaning and standardization techniques are used to preprocess the raw data and construct feature vectors, including: Obtain order data, classify it according to the order source channel, and obtain a collection of channel-specific order data; extract key attributes from the channel-specific order data to form structured order attribute data; Using data cleaning technology, the order attribute data is deduplicated, de-noised and processed for outliers to obtain cleaned order attribute data; through data standardization, the cleaned order attribute data is converted into a unified format and representation to form standardized order attribute data; Based on the standardized order attribute data, the order feature vector is constructed as the input of the machine learning algorithm; the machine learning model is trained and optimized to perform order classification and prediction tasks based on the order feature vector; Combine the output of the machine learning model with business rules to automatically generate order processing decisions and form an intelligent order management solution.

3. The multi-dimensional workshop task allocation method according to claim 1, characterized in that: Use advanced machine learning algorithms to build decision tree models, including: According to the preprocessed order data, the feature vector of each order is obtained, and the feature vector includes various information required for order processing; The decision tree algorithm is used to train the order feature vector to obtain a decision tree model, which determines the processing complexity based on the order features. For new orders, extract their feature vectors and input them into the decision tree model. The model makes judgments along the branch conditions of the decision tree based on the dimensional values ​​of the feature vectors to obtain the processing complexity of the order. When the processing complexity of an order exceeds the preset threshold, the order is marked as a high-complexity order and requires special processing; For orders with processing complexity below the threshold, they will directly enter the normal processing process; Dynamically adjust the allocation of production resources according to the processing complexity of the order.

4. The multi-dimensional workshop task allocation method according to claim 3 is characterized in that: The decision tree algorithm is used to train the order feature vector to obtain a decision tree model. The model determines the processing complexity based on the order features, including: By analyzing the various features in the order data set, a tree model is constructed. The processing complexity of the order is predicted through binary selection based on the input feature vector. During the training process, the segmentation points representing the data features are selected to form the nodes of the tree. Then, according to the decision path of the nodes, the data is divided into different areas, and each area corresponds to a prediction result. The decision tree model can make judgments along the branch conditions of the tree based on the feature vector of the new order.

5. The multi-dimensional workshop task allocation method according to claim 1 is characterized in that: Use optimization algorithms to determine the task allocation plan, including: Obtain the processing complexity information of the order and use it as one of the input parameters of the optimization algorithm; monitor the operating status of the production line in real time and use it as another set of input parameters of the optimization algorithm; build an optimization model based on the order complexity and production line status; use a heuristic algorithm to search for task allocation solutions to obtain the processing sequence and time schedule of each order in each process; when it is impossible to converge to a satisfactory solution in a short time, introduce a local search strategy to explore again in the current vicinity; convert the optimization results into an executable production plan and send it to the workshop management system and equipment control system; continuously monitor the production execution, compare the optimization plan with the actual output, and when deviations are found, trigger the algorithm to recalculate and adjust the task allocation.

6. The multi-dimensional workshop task allocation method according to claim 5 is characterized in that: The heuristic algorithm is used to search for a task allocation scheme, and the task allocation scheme is expressed as: ,in, is a decision variable vector, representing the task allocation scheme, is the objective function, which represents the performance index to be optimized. is an inequality constraint function, which represents various resource constraints. is an equality constraint function that represents a specific production requirement or equilibrium condition that must be met. represents the task allocation scheme, Used to introduce constraints.

7. The multi-dimensional workshop task allocation method according to claim 1 is characterized in that: Through model calculation and analysis, the task allocation scheme is obtained, and load balancing optimization is performed through the load strategy. The load strategy is expressed as: ,in Indicates The time allocated for each task, Indicates in all Find the load value of the load in the computing resource, is the time interval or polling period for task allocation, is the assigned time or start time of the first task.

8. The multi-dimensional workshop task allocation method according to claim 1 is characterized in that: Also includes: By collecting the operation data of the production line, the decision tree model and task allocation strategy are optimized.

9. The multi-dimensional workshop task allocation method according to claim 8, characterized in that: Optimize the decision tree model and task allocation strategy, including: By deploying sensors and data acquisition equipment on the production line, the data of the production line is continuously collected to build a production line operation database; the collected production line operation data is preprocessed to obtain a training data set suitable for the machine learning algorithm; according to the characteristics of the production line and the optimization goals, the machine learning algorithm is selected to build a production line optimization model; the machine learning model is trained using the training data set, and the model hyperparameters are optimized through cross-validation and grid search; the trained machine learning model is deployed to the control system of the production line, and the production efficiency and quality are predicted based on the real-time collected production line operation data, and the task allocation strategy and operation parameters of the production line are dynamically adjusted; the operating status and optimization effect of the production line are continuously monitored, new operation data is collected, and the machine learning model is regularly retrained and optimized; the efficiency and quality comparison before and after the production line optimization is displayed through visualization tools.

10. A multi-dimensional shop task allocation system, used to implement a multi-dimensional shop task allocation method according to any one of claims 1 to 9, characterized in that: include: The data collection module obtains order data from different channels, classifies them according to the order source channels, obtains channel-specific order data sets, extracts key attributes, and forms structured order attribute data. The machine learning model training module selects and trains suitable machine learning algorithms, builds optimization models, trains models using preprocessed data sets, and optimizes model hyperparameters using cross-validation and grid search; The task allocation optimization module automatically generates order processing decisions based on the output results of the optimization model and business rules, and then dynamically adjusts the task allocation of the production line to perform load balancing and rational use of resources; The monitoring and visualization module is used to continuously monitor the operating status and optimization effect of the production line, collect new operating data, and regularly retrain and optimize the machine learning model. It then uses visualization tools to display the efficiency and quality comparison before and after the production line optimization.

Images