A method for generating a quality sampling plan and related equipment

Abstract

The present disclosure provides a quality sampling scheme generation method and related equipment, the method comprises: obtaining sampling data of a multi-source business system and sampling information of a sampling task; based on the scene characteristics and the constraint term, the weight coefficient and the penalty coefficient of the comprehensive objective function are determined by using the parameter recommendation agent, and a multi-objective nonlinear programming model is constructed based on the weight coefficient and the penalty coefficient; based on the scale of the equipment to be sampled, the solution scheme is determined, and the multi-objective nonlinear programming model is solved according to the solution scheme to generate the sampling scheme; the sampling scheme is evaluated based on the preset evaluation standard; in response to the evaluation result not satisfying the preset condition, the constraint term or the weight coefficient is returned and the model solving is re-executed. Through the above scheme, the present disclosure can improve the intelligent level and decision scientificity of the sampling scheme compilation, and realize the optimal balance under multi-objective constraints.

Description

Technical Field

[0001] This disclosure relates to the field of quality control technology, and in particular to a method for generating a quality sampling inspection plan and related equipment. Background Technology

[0002] In the process of digital transformation of the new power system, the random inspection of the quality of materials and equipment is a key link in ensuring the safe and reliable operation of power grid equipment, and directly affects the stability of power grid operation and the resilience of the supply chain. The types of power grid materials and equipment are numerous, covering transmission line equipment, distribution equipment, electrical components and other categories. In actual sampling inspection, there are practical problems such as a large sampling base, large differences in equipment failure risk, and limited testing resources.

[0003] Traditional quality sampling methods often rely on historical experience and subjective judgments based on the importance of materials, with sampling lists being drafted manually; or they use simple sampling tools that simply generate sampling plans according to preset fixed rules. Existing sampling models are highly subjective, have low compilation efficiency, and poor adaptability to different scenarios, and cannot meet the intelligent sampling needs of large-volume, multi-category, and differentiated risk materials and equipment.

[0004] It should be noted that the above description of the technical background is only for the purpose of providing a clear and complete explanation of the technical solutions of the present invention and facilitating understanding by those skilled in the art. It should not be assumed that the above technical solutions are known to those skilled in the art simply because they have been described in the background section of this invention. Summary of the Invention

[0005] In view of this, the purpose of one or more embodiments of this disclosure is to provide a quality sampling inspection method and related equipment based on operations research optimization, so as to solve or partially solve the problems raised in the background art.

[0006] To achieve the above objectives, this disclosure provides a method for generating a quality sampling inspection plan, the method comprising:

[0007] Acquire sampling data from multi-source business systems, as well as sampling information for sampling tasks; wherein, the sampling data includes the scale of the equipment to be sampled; and the sampling information includes scenario characteristics and constraints. Based on the scenario features and the constraints, a parameter recommendation agent is used to determine the weight coefficients and penalty coefficients of the comprehensive objective function, and a multi-objective nonlinear programming model is constructed based on the weight coefficients and the penalty coefficients; wherein, the parameter recommendation agent is trained based on the sampling information of historical sampling tasks; Based on the scale of the equipment to be inspected, a solution scheme is determined, and the multi-objective nonlinear programming model is solved according to the solution scheme to generate an inspection scheme. The sampling plan is evaluated based on preset evaluation criteria; if the evaluation result does not meet the preset conditions, the constraint terms or the weight coefficients are adjusted, and the model solution is re-executed.

[0008] Based on the same inventive concept, this disclosure also provides a system for generating quality sampling inspection schemes, the system comprising: The acquisition module is configured to acquire sampling data from multi-source business systems and sampling information of sampling tasks; wherein, the sampling data includes the scale of the equipment to be sampled; and the sampling information includes scenario features and constraints. The construction module is configured to determine the weight coefficients and penalty coefficients of the comprehensive objective function using a parameter recommendation agent based on the scene features and the constraints, and to construct a multi-objective nonlinear programming model based on the weight coefficients and the penalty coefficients; wherein the parameter recommendation agent is trained based on the sampling information of historical sampling tasks. The generation module is configured to determine a solution scheme based on the scale of the equipment to be inspected in the current sampling task, and solve the multi-objective nonlinear programming model according to the solution scheme to generate a sampling scheme. The evaluation module is configured to evaluate the sampling plan based on preset evaluation criteria; in response to the evaluation result not meeting the preset conditions, return to adjust the constraint terms or the weight coefficients, and re-execute the model solution.

[0009] Based on the same inventive concept, this disclosure also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement a method for generating a quality sampling inspection scheme as described in any of the above claims.

[0010] Based on the same inventive concept, this disclosure also provides a non-transitory computer-readable storage medium storing computer instructions for causing the computer to execute any of the above-described methods for generating a quality sampling inspection scheme.

[0011] Based on the same inventive concept, this disclosure also provides a computer program product, including one or more computer programs, which, when executed by one or more processors, implement the method for generating any of the above-described quality sampling inspection schemes.

[0012] As can be seen from the above, the quality sampling inspection scheme generation method provided in this disclosure acquires sampling inspection data from multiple business systems and constructs a multi-objective nonlinear programming model. It then automatically determines the objective function weight coefficients and penalty coefficients using a parameter recommendation agent, adaptively selects a solution scheme based on the scale of the equipment to be inspected, and finally evaluates and iteratively optimizes the generated sampling inspection scheme. This solves the problems of strong subjectivity, low compilation efficiency, lack of multi-objective optimization, and single constraint handling in traditional sampling inspection methods. The technical solution of this disclosure realizes intelligent recommendation of sampling parameters, adaptive selection of solution algorithms, automatic evaluation of scheme results, and closed-loop optimization, forming an end-to-end intelligent solution from data preparation to scheme output. This effectively reduces the cost of manual intervention and improves the scientific nature and execution efficiency of the sampling inspection scheme. Attached Figure Description

[0013] To more clearly illustrate the technical solutions in one or more embodiments of this disclosure or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the accompanying drawings described below are only one or more embodiments of this disclosure. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0014] Figure 1 A flowchart illustrating a method for generating a quality sampling inspection plan provided in this embodiment of the disclosure; Figure 2 A schematic diagram illustrating the generation process of the quality sampling inspection scheme provided in this embodiment of the disclosure; Figure 3 This is a schematic diagram of the agent invocation process provided in an embodiment of the present disclosure; Figure 4 A schematic diagram of the structure of a quality sampling inspection scheme generation system provided in an embodiment of this disclosure; Figure 5 This is a schematic diagram of the hardware structure of an electronic device provided in an embodiment of this disclosure. Detailed Implementation

[0015] To make the objectives, technical solutions, and advantages of this disclosure clearer, the following detailed description is provided in conjunction with specific embodiments and the accompanying drawings.

[0016] It should be noted that, unless otherwise defined, the technical or scientific terms used in one or more embodiments of this disclosure should have the ordinary meaning understood by one of ordinary skill in the art to which this disclosure pertains. The terms "first," "second," and similar words used in one or more embodiments of this disclosure do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Terms such as "comprising" or "including" mean that the element or object preceding the word encompasses the elements or objects listed following the word and their equivalents, without excluding other elements or objects. Terms such as "connected" or "linked" are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. Terms such as "upper," "lower," "left," and "right" are used only to indicate relative positional relationships; when the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0017] Existing quality inspection plan development primarily employs two methods. One is the traditional manual development model, where quality control personnel manually draft the inspection list based on historical inspection experience, subjective judgment of material importance, and a fixed inspection ratio standard. This model lacks standardized mathematical models and algorithmic support, relying solely on reactive adjustments to inspection strategies after failures, lacking scientific rigor and proactive predictive capabilities. The other is a rudimentary automated tool-assisted model, using basic inspection calculation tools to generate inspection plans based solely on preset fixed rules, such as mandatory inspection of high-value materials or minimum inspection quantity limits for the same batch. This model fails to comprehensively consider key constraints such as equipment failure risk levels, testing facility capacity load, and equipment distribution, nor does it address multi-objective optimization and equilibrium needs. It still relies on repeated manual parameter adjustments and plan revisions, failing to achieve end-to-end intelligent generation throughout the entire process.

[0018] As described in the background section, the relevant technologies have the following problems: I. The scientific basis of the proposed solutions is insufficient, and the accuracy of risk control is low. Current quality sampling methods generally rely on manual experience or fixed rules, which are highly subjective and have poor adaptability. This can easily lead to problems such as high-risk equipment being missed and low-risk equipment being over-sampled, making it difficult to achieve precise control of multiple factors such as failure risk, testing costs, and production capacity constraints.

[0019] Second, there is a lack of multi-objective collaboration and dynamic adaptation capabilities. Current quality sampling inspection schemes often unilaterally pursue single indicators such as sampling compliance rate, failing to achieve synergistic optimization of fault risk coverage, cost control and efficiency improvement, which can easily lead to resource waste or budget overruns. At the same time, they lack the ability to respond quickly to dynamic scenarios such as large-scale, multi-batch sampling inspection needs, emergency repairs, and supplier changes, making it difficult to flexibly adjust strategies according to changes in sampling scale and fault data.

[0020] Third, the level of intelligence is insufficient and the self-iteration capability is lacking. The current quality sampling tools do not have data learning and parameter recommendation functions, and key configurations rely on manual input; moreover, they lack closed-loop evaluation and feedback mechanisms, making it impossible to learn from historical data and intelligently coordinate multi-dimensional constraints (such as budget, capacity, and cycle), making it difficult to form an end-to-end intelligent solution.

[0021] Therefore, there is an urgent need to develop a technical solution for generating quality sampling inspection plans to meet the actual needs of risk control, multi-objective collaborative optimization, and intelligent adaptation to multiple constraints in the scenario of sampling inspection of massive materials and equipment.

[0022] The essence of developing intelligent quality inspection plans is a combinatorial optimization problem with multiple constraints and objectives. Its goal is to intelligently select inspection targets from a massive pool of equipment to be inspected, determine the inspection ratio and order, and achieve the comprehensive objectives of maximizing fault risk coverage, minimizing inspection costs, and optimizing inspection efficiency, while satisfying hard constraints such as inspection cycle, budget limits, and testing capacity, as well as soft preferences such as risk priority. Traditional manual methods relying on experience are no longer sufficient to handle these complex constraints and optimization requirements. Constructing accurate, efficient, and scalable operations research optimization models and intelligent execution systems is the core technological path to achieving the scientific and intelligent development of inspection plans.

[0023] Based on some implementations of this disclosure, a method for generating quality sampling inspection plans is provided. In this plan, considering the practical needs of digital transformation in quality sampling inspection operations, a novel "AI + solver" model is proposed. This model integrates heuristic algorithms, meta-heuristic algorithms, and optimization solvers to construct a fully intelligent sampling inspection plan development system. This addresses the problems of strong subjectivity, low efficiency, and poor adaptability inherent in traditional methods, supporting the digital and intelligent upgrading of quality control operations.

[0024] refer to Figure 1 The present disclosure discloses a method for generating a quality sampling inspection plan, which includes the following steps: Step S100: Obtain the sampling inspection data and sampling inspection information of the multi-source business system; wherein, the sampling inspection data includes the scale, basic information, fault and defect data of the equipment to be inspected, as well as the supplier's performance data and the testing agency's capacity data; the sampling inspection information includes scenario characteristics and constraints.

[0025] In some embodiments, the multi-source business system includes an equipment management system, a contract system, and a supplier system, etc.

[0026] In some embodiments, the basic information of the equipment to be inspected includes equipment identification, equipment type, fault risk level, testing cost, batch number, and region; the fault and defect data of the equipment to be inspected includes historical fault records and fault risk levels; and the capacity data of the testing institutions includes the maximum number of tests per batch and the unit testing cost for each testing institution.

[0027] In some embodiments, after obtaining the sampling data, it is preprocessed by performing operations such as data cleaning, missing value imputation, and outlier removal to ensure data quality.

[0028] In some embodiments, the scenario features include the total number of devices to be inspected, the proportion of high-risk devices, the average testing cost, and the total number of available testing institutions.

[0029] In some embodiments, the constraints include budget constraints, capacity constraints, risk constraints, cycle constraints, and preference constraints.

[0030] The budget constraint refers to the total cost of a single sampling inspection not exceeding the quarterly sampling inspection budget, which can be expressed as:

[0031] in, This represents the set of equipment to be sampled for inspection. Indicates the equipment number. ; Indicates equipment The sampling inspection status indicator variable, 1 indicates random sampling, and 0 indicates no random sampling; Indicates equipment The sampling rate is a continuous variable, with a value ranging from 0 to 1. Indicates equipment Cost per test (unit: yuan); This indicates the upper limit of the current sampling inspection task, which is usually the upper limit of the quarterly sampling inspection fee budget (unit: yuan).

[0032] The capacity constraint refers to the fact that the number of samples taken by a testing institution in a single batch cannot exceed its capacity limit, which can be expressed as:

[0033] in, This represents the set of equipment to be sampled for inspection. Indicates the equipment number. ; This represents a collection of testing institutions. For the organization number, ; Indicates equipment The sampling inspection status indicator variable, 1 indicates random sampling, and 0 indicates no random sampling; Indicates equipment The sampling rate is a continuous variable. Its value ranges from 0 to 1; Indicates equipment Whether to assign to testing institutions , 1 indicates allocation, 0 indicates no allocation; Indicates testing agency Maximum testing capacity per batch (unit: units / batch).

[0034] The risk constraint refers to the requirement that the total number of samples inspected for each risk level is not less than the minimum inspection ratio requirement for that level; whereby the inspection ratio for high-risk equipment is greater than or equal to a preset threshold, and the inspection ratio for medium-risk and low-risk equipment is greater than or equal to the minimum inspection ratio, can be expressed as:

[0035] in, This represents a set of equipment risk levels, including high, medium, and low levels. ; Indicates risk level The corresponding minimum sampling rate ( ); Indicates equipment The failure risk coefficients are: high risk = 3, medium risk = 2, and low risk = 1. Indicates equipment The sampling inspection status indicator variable, 1 indicates random sampling, and 0 indicates no random sampling; Indicates equipment The sampling rate is a continuous variable. The value ranges from 0 to 1.

[0036] In some embodiments, in addition to satisfying risk constraints, it is also necessary to further satisfy allocation constraints between equipment and testing institutions. Sampling equipment must be uniquely assigned to a specific testing institution, while unsampling equipment is not assigned to any testing institution. This can be represented as:

[0037] in, Indicates equipment Whether to assign to testing institutions , 1 indicates allocation, 0 indicates no allocation; Indicates equipment The sampling inspection status indicator variable, 1 indicates random sampling, and 0 indicates no random sampling; This represents the set of equipment to be sampled for inspection. ; This represents a collection of testing institutions. For the organization number, .

[0038] The period constraint refers to the requirement that the execution time of any testing institution's sampling plan does not exceed the preset sampling period, which can be expressed as:

[0039] in, This represents the set of equipment to be sampled for inspection. ; This represents a collection of testing institutions. ; Indicates the preset sampling period (unit: days); This represents the set of equipment to be sampled for inspection. ; Indicates equipment At the testing agency The sampling inspection execution time (unit: days). .

[0040] The aforementioned preference constraint (soft constraint) refers to the introduction of a penalty coefficient to ensure that equipment from the same batch and region is preferentially allocated to the same testing institution. ( (This can be adjusted according to business needs), and the cost of violating this preference will be included in the objective function to balance business preferences with other optimization objectives.

[0041] In some embodiments, standardized constraints and budget control mechanisms are set to avoid budget overruns and compliance risks caused by manual operations.

[0042] S200. Based on the scene features and the constraints, a parameter recommendation agent is used to determine the weight coefficients and penalty coefficients of the comprehensive objective function, and a multi-objective nonlinear programming model is constructed based on the weight coefficients and the penalty coefficients. The parameter recommendation agent is trained based on sampling information from historical sampling tasks.

[0043] In some embodiments, the parameter recommendation agent is a deep learning-based environment perception and decision-making engine. First, training samples are constructed from a large number of historical sampling cases. These samples are all taken from historical sampling cases that have been confirmed by experts or have achieved the best actual performance. The "scene feature-optimal parameter" mapping pairs are extracted from each group of cases; wherein, the optimal parameters include the objective function weight coefficients. , and and the penalty coefficient of preference constraints .

[0044] Based on the scene features and constraints in the sampling information, the weight coefficients of the comprehensive objective function are determined using the trained parameter intelligent recommender. , and and penalty coefficient .

[0045] In some embodiments, a deep neural network is constructed, with the scene feature vector of historical sampling tasks as input and the optimal parameter set as output. Supervised learning is performed using historical labeled samples, so that the neural network fits the nonlinear mapping relationship between different sampling scenarios and the optimal parameters, learns the inherent rules of weight allocation and penalty coefficient setting, and completes the training of the parameter recommendation agent.

[0046] Specifically, with the optimization directions of maximizing fault risk coverage, minimizing detection costs, and minimizing sampling cycle deviation, a multi-objective nonlinear programming model is constructed using the weighted summation method.

[0047] Fault risk coverage The value ranges from [0,1]. The smaller the value, the more comprehensive the sampling coverage of high-risk equipment and the better the risk prevention and control effect. Testing cost control rate The value ranges from [0,1] (if the optimization is effective). The smaller the value, the closer the sampling cost is to the upper limit of the budget or lower than the budget, and the better the cost control effect. Its value is ≥0. The smaller the value, the closer the sampling execution time is to the preset cycle, and the better the sampling efficiency.

[0048] A multi-objective nonlinear programming model is determined based on the weighting coefficients and the penalty parameters.

[0049] In some embodiments, after the model training is completed, the scene features of the current sampling task are input into the trained deep neural network, i.e., the parameter recommendation agent, and the inference output is the weight coefficient of the objective function adapted to the current task. , and and penalty coefficient Then, the weighting coefficients and penalty coefficients are substituted into the pre-constructed multi-objective optimization framework, and combined with the constraint terms in the sampling information, to obtain the multi-objective nonlinear programming model for the current sampling task. The expression for the comprehensive objective function of the multi-objective nonlinear programming model is as follows:

[0050] in, This represents the minimized value of the comprehensive objective function; 、 The weight coefficients for each optimization objective correspond to fault risk coverage, detection cost control rate, and sampling cycle deviation, respectively. This represents the set of equipment to be sampled for inspection. Indicates the equipment number. ; This represents a collection of testing institutions. For the organization number, ; For equipment The failure risk coefficients are: high risk = 3, medium risk = 2, and low risk = 1. Indicates equipment The sampling inspection status indicator variable, 1 indicates random sampling, and 0 indicates no random sampling; Indicates equipment The sampling rate is a continuous variable, with a value ranging from 0 to 1. Indicates equipment Cost per test (unit: yuan); This indicates the current budget ceiling for the sampling inspection task, which is usually the quarterly budget ceiling for sampling inspection fees (unit: yuan). Indicates equipment At the testing agency The sampling inspection execution time (unit: days). ; This indicates the preset sampling period (unit: days).

[0051] The expression for the overall objective function after introducing the penalty coefficient is as follows:

[0052] in, This represents the overall objective function after introducing preference penalties; Represent the overall objective function; This represents the preference penalty coefficient, used to balance business preferences with optimization objectives such as risk, cost, and cycle time. This represents the set of equipment to be sampled for inspection; This represents a collection of testing institutions; Indicates equipment The batch to be sampled This represents the set of batches sampled for inspection. ; Indicates equipment The area where it is located Represents a set of regions. ; Indicates equipment To the testing agency The allocation indicator variable, A value of 1 indicates allocation, and a value of 0 indicates no allocation; This indicates a preference violation indicator. When equipment from the same batch and region is assigned to different testing institutions, this item takes a value of 1; otherwise, it takes a value of 0.

[0053] In some embodiments, the multi-objective nonlinear programming model aims to determine the optimal sampling list, sampling ratio, and organizational allocation scheme. Under the premise of satisfying hard constraints and soft constraints related to business preferences, it achieves comprehensive optimization of multiple objectives. Simultaneously, a parameter recommendation agent is used to recommend various parameters, thus solidifying historical sampling experience and business preferences into the model's automatic decision-making logic. This replaces manual trial-and-error parameter tuning, enabling autonomous optimization and iterative development of the sampling plan.

[0054] In some embodiments, equipment failure risks exhibit significant differences in actual sampling operations. The probability and impact of failure vary among different devices, and using a uniform sampling ratio could easily lead to an imbalance in risk control. The model addresses this by configuring sampling priorities differently, prioritizing coverage of high-risk equipment to maximize failure risk coverage. Simultaneously, the scale of equipment to be sampled varies considerably, ranging from hundreds to tens of thousands of items per sampling. Therefore, the solution strategy needs to be dynamically adjusted according to the scale of the sampling task: small-scale tasks seek precise optimal solutions, while large-scale tasks pursue efficient near-optimal solutions, ensuring a balance between efficiency and quality in solution development. By reducing the idle time of testing institutions and equipment transportation waiting time, the efficiency of sampling completion is improved, achieving optimal sampling efficiency.

[0055] In addition, the capacity of testing institutions (such as the number of tests per day and the number of testing equipment) and testing costs (such as labor costs and consumable costs) are all constrained by the quarterly budget. Within the budget and capacity range, it is necessary to achieve the optimal allocation of testing resources, control the total cost of overall sampling while meeting the risk coverage requirements, and minimize the testing cost.

[0056] In some embodiments, due to the complex nonlinear interactions between factors such as failure risk, sampling scale, and capacity budget, this disclosure adopts nonlinear programming (NLP) as the core modeling framework, and combines heuristic and metaheuristic algorithms to adapt to the efficient solution requirements of different sampling scale scenarios.

[0057] In some embodiments, this disclosure comprehensively considers core factors such as equipment failure risk, sampling base, testing capacity, budget, and regional distribution. Combining hard constraints (budget, capacity, cycle) and soft preferences (sampling within the same batch / region), a multi-objective nonlinear programming mathematical model is constructed. Through a weighted objective function, a comprehensive optimization of risk coverage, cost control, and efficiency improvement is achieved, accurately depicting the complex logic of the sampling business.

[0058] S300. Based on the scale of the equipment to be inspected, determine the solution scheme, and solve the multi-objective nonlinear programming model according to the solution scheme to generate the inspection scheme.

[0059] In some embodiments, a scale-adaptive solution strategy is adopted, which dynamically selects the solution method according to the number of devices to be sampled, taking into account both the solution accuracy and execution efficiency.

[0060] In some embodiments, determining a solution scheme based on the scale of equipment to be inspected in the current sampling task, and solving the multi-objective nonlinear programming model according to the solution scheme to generate a sampling plan, includes: Small-scale task: Number of devices to be sampled ≤ preset threshold When the number of devices to be inspected is less than or equal to a preset threshold, the optimization solver is invoked to perform a warm_start solution, which uses the initial feasible solution to accelerate convergence and outputs the globally optimal solution.

[0061] Large-scale task: Number of devices to be inspected > preset threshold In response to the number of devices to be sampled exceeding the preset threshold, a combination of heuristic and metaheuristic algorithms is used to solve the problem. The specific steps are as follows: The method of solving the problem by combining heuristic algorithms and metaheuristic algorithms includes: Using heuristic algorithms, based on the basic information of the equipment to be inspected (such as key features like equipment risk level, batch, and region), the core equipment to be inspected is quickly screened and a preliminary allocation scheme is determined, generating an initial solution that meets the constraints. Meta-heuristic algorithms (such as genetic algorithms and particle swarm optimization algorithms) are used to iteratively optimize the initial solution, dynamically adjust the sampling ratio and the allocation strategy of testing institutions, and gradually approach the multi-objective comprehensive optimal solution until the preset convergence conditions are met.

[0062] The preset threshold can be 100 items.

[0063] Based on the solution scheme, the multi-objective nonlinear programming model is solved to obtain candidate sampling schemes; the candidate sampling schemes are then verified to ensure they satisfy all the constraints, namely budget constraints, capacity constraints, risk constraints, cycle constraints, and preference constraints.

[0064] If the verification passes, the optimal sampling ratio, equipment allocation relationship and execution sequence are fixed, and the candidate sampling plan is output as the sampling plan; if the verification fails, the process returns to iterative optimization until a compliant and feasible sampling plan is output.

[0065] S400. Evaluate the sampling plan based on the preset evaluation criteria; if the evaluation result does not meet the preset conditions, return to adjust the constraint parameters or the objective function weight coefficients, and solve again.

[0066] In some embodiments, evaluating the sampling plan based on preset evaluation criteria includes: After the solution is completed, the system automatically generates an executable sampling plan and conducts a quantitative evaluation of the sampling plan based on a multi-dimensional evaluation index system to verify the scientificity, compliance and effectiveness of the plan.

[0067] The evaluation indicators of the multi-dimensional evaluation indicator system include: Risk coverage, specifically the sampling coverage rate of high / medium risk equipment, reflects the precision of risk prevention and control; the higher the value, the more targeted the risk prevention and control. The calculation formula is: .

[0068] Cost control rate, the ratio of the actual cost of the sampling plan to the total budget, reflects the effectiveness of cost control. A value ≤ 1 indicates that the cost did not exceed the budget; the smaller the value, the better the cost control effect. The calculation formula is: .

[0069] Sampling efficiency is the ratio of the estimated completion time of the plan to the preset sampling cycle. This indicator reflects execution efficiency; a value ≤ 1 indicates that the plan can be completed within the cycle, and the smaller the value, the higher the execution efficiency. The calculation formula is: .

[0070] Compliance rate, the percentage of solutions that meet all hard constraints, reflects the compliance of the solution. This value must be 100%; otherwise, the constraint parameters or objective function weights must be adjusted, and the model must be solved again. The calculation formula is: .

[0071] The fault detection rate, based on historical fault probability data, is the ability of a sampling inspection plan to detect potential faults. This indicator reflects the plan's ability to detect known risk points and is used to assist in evaluating the plan's risk control effectiveness. The calculation formula is: .

[0072] The weighted scoring model is used to calculate the comprehensive score of each evaluation indicator to obtain the quantitative evaluation result of the current sampling plan. If the comprehensive evaluation result does not meet the preset conditions, the iterative optimization process is triggered, the parameters of the constraint terms or the weight coefficients of the comprehensive objective function are adjusted, and the model solution steps are re-executed until a sampling plan that meets all evaluation criteria is generated.

[0073] In addition, the optimization logic and rating indicators of the sampling plan can be presented through visualization methods such as radar charts and bar charts, making the decision-making process traceable and quantifiable, and improving the transparency of quality control operations.

[0074] In some embodiments, this disclosure relies on a multi-agent collaborative orchestration mechanism to automate the entire process of data integration, parameter recommendation, model solving, and result evaluation. This reduces the traditional manual compilation cycle of sampling plans from several days to hours or even minutes, significantly decreasing the degree of human intervention in data preparation, parameter tuning, and plan iteration. Specifically, the parameter recommendation agent can continuously learn iteratively from historical sampling cases, gradually improving the accuracy of parameter recommendations. Simultaneously, the solution strategy can dynamically and adaptively adjust with equipment scale and business data, adapting to diverse business scenarios such as emergency sampling and batch changes, thereby continuously optimizing and upgrading the intelligent capabilities of sampling plan compilation.

[0075] This disclosure constructs a nonlinear programming model that integrates multiple factors such as equipment failure risk, sampling base, testing agency capacity, quarterly budget, and sampling cycle to ensure the scientific validity and rationality of the sampling plan in terms of risk coverage, cost control, and efficiency improvement. It combines heuristic algorithms, meta-heuristic algorithms, and optimization solvers to solve the model, achieving efficient solutions for large-scale sampling problems and accurate solutions for small-scale problems. Furthermore, it establishes a multi-dimensional evaluation system including indicators such as risk coverage, cost control rate, sampling efficiency, and compliance rate to comprehensively assess the effectiveness and rationality of the sampling plan, providing a basis for plan optimization. Through the "intelligent agent orchestration + operations research optimization" end-to-end intelligent sampling plan of this disclosure, it effectively solves the problems of strong subjectivity, low compilation efficiency, poor dynamic adaptability, lack of multi-objective optimization, and single constraint handling in traditional sampling methods. It achieves intelligent recommendation of sampling parameters, adaptive algorithm selection, and automatic evaluation and optimization of results, forming an end-to-end intelligent solution from data preparation to plan output, reducing the cost of manual intervention.

[0076] It is understandable that this method can be executed by any device, equipment, platform, or cluster of devices with computing and processing capabilities.

[0077] It should be noted that the methods of one or more embodiments of this disclosure can be executed by a single device, such as a computer or server. The methods of this embodiment can also be applied in a distributed scenario, where multiple devices cooperate to complete the process. In such a distributed scenario, one of these devices may execute only one or more steps of the methods of one or more embodiments of this disclosure, and the multiple devices will interact with each other to complete the method described.

[0078] It should be noted that the above description pertains to specific embodiments of this disclosure. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps described in the claims may be performed in a different order than those shown in the embodiments and may still achieve the desired results. Furthermore, the processes depicted in the drawings do not necessarily require a specific or sequential order to achieve the desired results. In some embodiments, multitasking and parallel processing are possible or may be advantageous.

[0079] Figure 2 This is a schematic diagram illustrating the generation process of the quality sampling inspection plan provided in this embodiment of the disclosure. Figure 2 As shown, the process for generating the quality sampling inspection plan includes the following steps: First, through relevant data integration services, the raw data required for generating quality sampling inspection plans is obtained from multiple data sources. These data sources include: equipment management system, supplier information system, contract management system, historical sampling inspection plan database, and other external business systems. Specifically, equipment fault and defect hazard information, defect and fault records, and maintenance orders are obtained from the equipment management system to acquire equipment fault analysis data; supplier performance data is obtained from the supplier information system; contract information is obtained from the contract management system; historical sampling inspection plan information is obtained from the historical sampling inspection plan database; and business constraint data such as quarterly sampling inspection fee budget, testing agency capacity, preset sampling cycle, sampling ratio, and sampling base are obtained from other external systems.

[0080] Secondly, the raw data obtained above is uniformly connected to the sampling inspection platform for data preprocessing. The preprocessing includes data cleaning, format standardization, outlier removal, and missing value completion, forming standardized input parameters for subsequent model construction and solving.

[0081] Then, based on the preprocessed data, optimization objectives and constraints are determined. The optimization objectives include: maximizing fault risk coverage, minimizing testing costs, and minimizing sampling cycle deviation; the constraints include: quarterly sampling fee budget constraints, testing agency capacity constraints, sampling base constraints, sampling cycle constraints, and business preference constraints.

[0082] Next, the solution strategy is adaptively selected based on the scale of the equipment to be inspected: When the number of equipment to be inspected does not exceed 100, the optimization solver is called and the warm_start method is used to perform accurate solution. The initial feasible solution is used to accelerate convergence and output the global optimal solution. When the number of devices to be inspected exceeds 100, a hybrid solution strategy combining heuristic and metaheuristic algorithms is adopted: first, an initial solution is generated through a heuristic algorithm, and then a metaheuristic algorithm is used for iterative optimization to gradually approach the approximate optimal solution.

[0083] After the model is solved, the solution is verified to check whether the output scheme meets all constraints. If the verification passes, the final sampling scheme is generated, and the output includes: sampling probability, sampling base, sampling cycle arrangement, etc. If the verification fails, the strategy parameters are adjusted, and the model is rebuilt and solved until a compliant and feasible sampling scheme is output.

[0084] Finally, the generated sampling plan is pushed to the front end of the sampling platform, allowing business personnel to view, confirm, modify, and finally issue the sampling strategy for execution. It also supports the recommendation and analysis functions of sampling strategies, forming an end-to-end closed-loop process from data integration to solution output.

[0085] Figure 3 This is a schematic diagram of the agent invocation process provided in an embodiment of this disclosure. Figure 3 As shown, the agent invocation process includes the following steps: First, the user sends a request to a Large Model (LLM). LLMs integrate an agent with multiple cognitive and reasoning capabilities, including intelligent question answering, semantic understanding, causal inference, and chain reasoning.

[0086] Secondly, LLMs invoke the intelligent agent to perform semantic parsing, intent understanding, and causal logical chain reasoning on the user's input request, generating intelligent recommendation parameters that are adapted to the current sampling task scenario.

[0087] Finally, the intelligent recommendation parameters are output to the operations research and optimization algorithm module as input to the operations research and optimization algorithm, which is used for the construction and solution of subsequent sampling schemes.

[0088] Based on the same inventive concept, corresponding to any of the methods in the above embodiments, this disclosure also provides a system for generating quality sampling inspection schemes. For example... Figure 4 As shown, the above system includes: The acquisition module 401 is configured to acquire sampling data from a multi-source business system and sampling information of the sampling task; wherein, the sampling data includes the scale of the equipment to be sampled; and the sampling information includes scenario features and constraints. The construction module 402 is configured to determine the weight coefficients and penalty coefficients of the comprehensive objective function using a parameter recommendation agent based on the scene features and the constraints, and to construct a multi-objective nonlinear programming model based on the weight coefficients and the penalty coefficients; wherein the parameter recommendation agent is trained based on the sampling information of historical sampling tasks. The generation module 403 is configured to determine a solution scheme based on the scale of the equipment to be inspected in the current sampling task, and solve the multi-objective nonlinear programming model according to the solution scheme to generate a sampling scheme. Evaluation module 404 is configured to evaluate the sampling plan based on preset evaluation criteria; in response to the evaluation result not meeting the preset conditions, return to adjust the constraint terms or the weight coefficients, and re-execute the model solution.

[0089] In some embodiments, the acquisition module 401 further includes: The sampling inspection data also includes fault and defect data of the equipment to be inspected, supplier performance data, and testing agency capacity data. The constraints include budget constraints, capacity constraints, risk constraints, cycle constraints, and preference constraints.

[0090] In some embodiments, the building module 402 further includes: The expression for the comprehensive objective function of the multi-objective nonlinear programming model is as follows:

[0091] in, This represents the minimized value of the comprehensive objective function; 、 The weight coefficients for each optimization objective correspond to fault risk coverage, detection cost control rate, and sampling cycle deviation, respectively. This represents the set of equipment to be sampled for inspection. Indicates the equipment number. ; This represents a collection of testing institutions. For the organization number, ; For equipment The failure risk coefficients are: high risk = 3, medium risk = 2, and low risk = 1. Indicates equipment The sampling inspection status indicator variable, 1 indicates random sampling, and 0 indicates no random sampling; Indicates equipment The sampling rate is a continuous variable, with a value ranging from 0 to 1. Indicates equipment The cost per test; This indicates the current budget ceiling for the sampling inspection task, which is usually the quarterly budget ceiling for sampling inspection fees; Indicates equipment At the testing agency The time for random inspection execution, ; Indicates the preset sampling period; The expression for the overall objective function after introducing the penalty coefficient is as follows:

[0092] in, This represents the overall objective function after introducing preference penalties; Represent the overall objective function; This represents the preference penalty coefficient, used to balance business preferences with optimization objectives such as risk, cost, and cycle time. This represents the set of equipment to be sampled for inspection; This represents a collection of testing institutions; Indicates equipment The batch to be sampled This represents the set of batches sampled for inspection. ; Indicates equipment The area where it is located Represents a set of regions. ; Indicates equipment To the testing agency The allocation indicator variable, A value of 1 indicates allocation, and a value of 0 indicates no allocation; This indicates a preference violation indicator. When equipment from the same batch and region is assigned to different testing institutions, this item takes a value of 1; otherwise, it takes a value of 0.

[0093] In some embodiments, the generation module 403 further includes: When the number of devices to be inspected is less than or equal to a preset threshold, the optimization solver is invoked to perform a warm_start solution and output the solution. In response to the number of devices to be inspected exceeding the preset threshold, a solution is obtained by combining heuristic algorithms and metaheuristic algorithms, and the solution is output. In some embodiments, the generation module 403 further includes: The method of solving the problem by combining heuristic algorithms and metaheuristic algorithms includes: Using a heuristic algorithm, based on the basic information of the equipment to be inspected, the equipment to be inspected is screened and a preliminary allocation scheme is determined, generating an initial solution; A metaheuristic algorithm is used to iteratively optimize the initial solution and output a solution scheme. Based on the solution scheme, the multi-objective nonlinear programming model is solved to obtain a candidate sampling scheme; Verify whether the candidate sampling scheme satisfies all the constraints, namely budget constraints, capacity constraints, risk constraints, cycle constraints, and preference constraints; If the verification passes, the candidate sampling plan will be output as the sampling plan. If the verification fails, return to iterate and optimize again until the sampling plan is output.

[0094] In some embodiments, the evaluation module 404 further includes: The sampling plan is quantitatively evaluated based on a multi-dimensional evaluation index system; wherein the evaluation indexes of the multi-dimensional evaluation index system include risk coverage, cost control rate, sampling efficiency, compliance rate, and fault detection rate. The comprehensive score of each evaluation indicator is calculated using a weighted scoring model to obtain the evaluation result of the sampling plan; if the evaluation result does not meet the preset conditions, an iterative optimization process is triggered until a sampling plan that meets all evaluation indicators is generated.

[0095] For ease of description, the above system is described by dividing it into various modules based on their functions. Of course, when implementing one or more embodiments of this disclosure, the functions of each module can be implemented in one or more software and / or hardware.

[0096] The system described above is used to implement the corresponding methods in the foregoing embodiments and has the beneficial effects of the corresponding method embodiments, which will not be repeated here.

[0097] Figure 5 This embodiment illustrates a more specific hardware structure of an electronic device, which may include a processor 1010, a memory 1020, an input / output interface 1030, a communication interface 1040, and a bus 1050. The processor 1010, memory 1020, input / output interface 1030, and communication interface 1040 are interconnected internally via the bus 1050.

[0098] The processor 1010 can be implemented using a general-purpose CPU (Central Processing Unit), microprocessor, application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of this disclosure.

[0099] The memory 1020 can be implemented in the form of ROM (Read Only Memory), RAM (Random Access Memory), static storage device, dynamic storage device, etc. The memory 1020 can store the operating system and other applications. When the technical solutions provided in the embodiments of this disclosure are implemented by software or firmware, the relevant program code is stored in the memory 1020 and is called and executed by the processor 1010.

[0100] The input / output interface 1030 is used to connect input / output modules to realize information input and output. Input / output modules can be configured as components within the device (not shown in the figure) or externally connected to the device to provide corresponding functions. Input devices may include keyboards, mice, touchscreens, microphones, various sensors, etc., while output devices may include displays, speakers, vibrators, indicator lights, etc.

[0101] The communication interface 1040 is used to connect a communication module (not shown in the figure) to enable communication between this device and other devices. The communication module can communicate via wired means (such as USB, Ethernet cable, etc.) or wireless means (such as mobile network, WIFI, Bluetooth, etc.).

[0102] Bus 1050 includes a pathway for transmitting information between various components of the device, such as processor 1010, memory 1020, input / output interface 1030, and communication interface 1040.

[0103] It should be noted that although the above-described device only shows the processor 1010, memory 1020, input / output interface 1030, communication interface 1040, and bus 1050, in specific implementations, the device may also include other components necessary for normal operation. Furthermore, those skilled in the art will understand that the above-described device may only include the components necessary for implementing the embodiments of this disclosure, and not necessarily all the components shown in the figures.

[0104] The electronic devices described above are used to implement the corresponding methods in the foregoing embodiments and have the beneficial effects of the corresponding method embodiments, which will not be repeated here.

[0105] The computer-readable medium of this embodiment includes permanent and non-permanent, removable and non-removable media, and information storage can be implemented by any method or technology. Information can be computer-readable instructions, data structures, program modules, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic magnetic disk storage or other magnetic storage devices, or any other non-transfer medium that can be used to store information accessible by a computing device.

[0106] Based on the same inventive concept, corresponding to the method for generating a quality sampling inspection scheme in any of the above embodiments, this disclosure also provides a computer program product, which includes one or more computer programs. In some embodiments, the one or more computer programs are executable by one or more processors to cause the one or more processors to execute the method for generating a quality sampling inspection scheme. Corresponding to the execution entity for each step in each embodiment of the method for generating a quality sampling inspection scheme, the processor executing the corresponding step may belong to the corresponding execution entity. The computer program product of the above embodiments is used to cause the processor to execute the method for generating a quality sampling inspection scheme as described in any of the above embodiments, and has the beneficial effects of the corresponding method embodiments, which will not be repeated here.

[0107] Those skilled in the art should understand that the discussion of any of the above embodiments is merely exemplary and is not intended to imply that the scope of this disclosure (including the claims) is limited to these examples; within the framework of this disclosure, the technical features of the above embodiments or different embodiments can also be combined, the steps can be implemented in any order, and there are many other variations of different aspects of one or more embodiments of this disclosure as described above, which are not provided in detail for the sake of brevity.

[0108] Additionally, to simplify the description and discussion, and to avoid obscuring one or more embodiments of this disclosure, the provided drawings may or may not show well-known power / ground connections to integrated circuit (IC) chips and other components. Furthermore, the apparatus may be shown in block diagram form to avoid obscuring one or more embodiments of this disclosure, and this also takes into account the fact that the details of implementation of these block diagram apparatuses are highly dependent on the platform on which one or more embodiments of this disclosure will be implemented (i.e., these details should be fully understood by those skilled in the art). While specific details (e.g., circuitry) are set forth to describe exemplary embodiments of this disclosure, it will be apparent to those skilled in the art that one or more embodiments of this disclosure may be implemented without these specific details or with variations thereof. Therefore, these descriptions should be considered illustrative rather than restrictive.

[0109] Although this disclosure has been described in conjunction with specific embodiments thereof, many substitutions, modifications, and variations of these embodiments will be apparent to those skilled in the art from the foregoing description. For example, other memory architectures (e.g., dynamic RAM (DRAM)) may be used with the embodiments discussed.

[0110] This disclosure includes one or more embodiments intended to cover all such substitutions, modifications, and variations falling within the broad scope of the appended claims. Therefore, any omissions, modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of one or more embodiments of this disclosure should be included within the scope of protection of this disclosure.

Claims

1. A method for generating a quality sampling inspection plan, characterized in that, The method includes: Acquire sampling data from multi-source business systems, as well as sampling information for sampling tasks; wherein, the sampling data includes the scale of the equipment to be sampled; and the sampling information includes scenario characteristics and constraints. Based on the scenario features and the constraints, a parameter recommendation agent is used to determine the weight coefficients and penalty coefficients of the comprehensive objective function, and a multi-objective nonlinear programming model is constructed based on the weight coefficients and the penalty coefficients; wherein, the parameter recommendation agent is trained based on the sampling information of historical sampling tasks; Based on the scale of the equipment to be inspected, a solution scheme is determined, and the multi-objective nonlinear programming model is solved according to the solution scheme to generate an inspection scheme. The sampling plan is evaluated based on preset evaluation criteria; if the evaluation result does not meet the preset conditions, the constraint terms or the weight coefficients are adjusted, and the model solution is re-executed.

2. The method according to claim 1, characterized in that, The sampling inspection data also includes basic information of the equipment to be inspected, fault and defect data, supplier performance data, and testing agency capacity data; The constraints include budget constraints, capacity constraints, risk constraints, cycle constraints, and preference constraints.

3. The method according to claim 1, characterized in that, The expression for the comprehensive objective function of the multi-objective nonlinear programming model is as follows: in, This represents the minimized value of the comprehensive objective function; 、 The weight coefficients for each optimization objective correspond to fault risk coverage, detection cost control rate, and sampling cycle deviation, respectively. This represents the set of equipment to be sampled for inspection. Indicates the equipment number. ; This represents a collection of testing institutions. For the organization number, ; For equipment The failure risk coefficients are: high risk = 3, medium risk = 2, and low risk = 1. Indicates equipment The sampling inspection status indicator variable, 1 indicates random sampling, and 0 indicates no random sampling; Indicates equipment The sampling rate is a continuous variable, with a value ranging from 0 to 1. Indicates equipment The cost per test; This indicates the current budget ceiling for the sampling inspection task, which is usually the quarterly budget ceiling for sampling inspection fees; Indicates equipment At the testing agency The time for random inspection execution, ; Indicates the preset sampling period; The expression for the overall objective function after introducing the penalty coefficient is as follows: in, This represents the overall objective function after introducing preference penalties; Represent the overall objective function; This represents the preference penalty coefficient, used to balance business preferences with optimization objectives such as risk, cost, and cycle time. This represents the set of equipment to be sampled for inspection; This represents a collection of testing institutions; Indicates equipment The batch to be sampled This represents the set of batches sampled for inspection. ; Indicates equipment The area where it is located Represents a set of regions. ; Indicates equipment To the testing agency The allocation indicator variable, A value of 1 indicates allocation, and a value of 0 indicates no allocation; This indicates a preference violation indicator. When equipment from the same batch and region is assigned to different testing institutions, this item takes a value of 1; otherwise, it takes a value of 0.

4. The method according to claim 1, characterized in that, The process of determining a solution based on the scale of the equipment to be inspected includes: When the number of devices to be inspected is less than or equal to a preset threshold, the optimization solver is invoked to perform a warm_start solution and output the solution. In response to the number of devices to be sampled exceeding the preset threshold, a solution is obtained by combining heuristic algorithms and metaheuristic algorithms, and the solution is output.

5. The method according to claim 4, characterized in that, The method of solving the problem by combining heuristic algorithms and metaheuristic algorithms includes: Using a heuristic algorithm, based on the basic information of the equipment to be inspected, the equipment to be inspected is screened and a preliminary allocation scheme is determined, generating an initial solution; A metaheuristic algorithm is used to iteratively optimize the initial solution and output a solution scheme.

6. The method according to claim 4, characterized in that, The step of solving the multi-objective nonlinear programming model according to the solution scheme to generate a sampling plan includes: Based on the solution scheme, the multi-objective nonlinear programming model is solved to obtain a candidate sampling scheme; Verify whether the candidate sampling plan satisfies the constraints; If the verification passes, the candidate sampling plan will be output as the sampling plan. If the verification fails, return to iterate and optimize again until the sampling plan is output.

7. The method according to claim 6, characterized in that, The evaluation of the sampling plan based on preset evaluation criteria includes: The sampling plan is quantitatively evaluated based on a multi-dimensional evaluation index system; wherein the evaluation indexes of the multi-dimensional evaluation index system include risk coverage, cost control rate, sampling efficiency, compliance rate, and fault detection rate. The comprehensive score of each evaluation indicator is calculated using a weighted scoring model to obtain the evaluation result of the sampling plan; if the evaluation result does not meet the preset conditions, an iterative optimization process is triggered until a sampling plan that meets all evaluation indicators is generated.

8. A system for generating quality sampling inspection plans, characterized in that, The system includes: The acquisition module is configured to acquire sampling data from multi-source business systems and sampling information of sampling tasks; wherein, the sampling data includes the scale of the equipment to be sampled; and the sampling information includes scenario features and constraints. The construction module is configured to determine the weight coefficients and penalty coefficients of the comprehensive objective function using a parameter recommendation agent based on the scene features and the constraints, and to construct a multi-objective nonlinear programming model based on the weight coefficients and the penalty coefficients; wherein the parameter recommendation agent is trained based on the sampling information of historical sampling tasks. The generation module is configured to determine a solution scheme based on the scale of the equipment to be inspected in the current sampling task, and solve the multi-objective nonlinear programming model according to the solution scheme to generate a sampling scheme. The evaluation module is configured to evaluate the sampling plan based on preset evaluation criteria; in response to the evaluation result not meeting the preset conditions, return to adjust the constraint terms or the weight coefficients, and re-execute the model solution.

9. The system according to claim 8, characterized in that, The acquisition module further includes: The sampling inspection data also includes basic information of the equipment to be inspected, fault and defect data, supplier performance data, and testing agency capacity data; The constraints include budget constraints, capacity constraints, risk constraints, cycle constraints, and preference constraints.

10. The system according to claim 8, characterized in that, The building module also includes: The expression for the comprehensive objective function of the multi-objective nonlinear programming model is as follows: in, This represents the minimized value of the comprehensive objective function; 、 The weight coefficients for each optimization objective correspond to fault risk coverage, detection cost control rate, and sampling cycle deviation, respectively. This represents the set of equipment to be sampled for inspection. Indicates the equipment number. ; This represents a collection of testing institutions. For the organization number, ; For equipment The failure risk coefficients are: high risk = 3, medium risk = 2, and low risk = 1. Indicates equipment The sampling inspection status indicator variable, 1 indicates random sampling, and 0 indicates no random sampling; Indicates equipment The sampling rate is a continuous variable, with a value ranging from 0 to 1. Indicates equipment The cost per test; This indicates the current budget ceiling for the sampling inspection task, which is usually the quarterly budget ceiling for sampling inspection fees; Indicates equipment At the testing agency The time for random inspection execution, ; Indicates the preset sampling period; The expression for the overall objective function after introducing the penalty coefficient is as follows: in, This represents the overall objective function after introducing preference penalties; Represent the overall objective function; This represents the preference penalty coefficient, used to balance business preferences with optimization objectives such as risk, cost, and cycle time. This represents the set of equipment to be sampled for inspection; This represents a collection of testing institutions; Indicates equipment The batch to be sampled This represents the set of batches sampled for inspection. ; Indicates equipment The area where it is located Represents a set of regions. ; Indicates equipment To the testing agency The allocation indicator variable, A value of 1 indicates allocation, and a value of 0 indicates no allocation; This indicates a preference violation indicator. When equipment from the same batch and region is assigned to different testing institutions, this item takes a value of 1; otherwise, it takes a value of 0.

11. The system according to claim 8, characterized in that, The generation module further includes: When the number of devices to be inspected is less than or equal to a preset threshold, the optimization solver is invoked to perform a warm_start solution and output the solution. In response to the number of devices to be sampled exceeding the preset threshold, a solution is obtained by combining heuristic algorithms and metaheuristic algorithms, and the solution is output.

12. The system according to claim 8, characterized in that, The generation module further includes: The method of solving the problem by combining heuristic algorithms and metaheuristic algorithms includes: Using a heuristic algorithm, based on the basic information of the equipment to be inspected, the equipment to be inspected is screened and a preliminary allocation scheme is determined, generating an initial solution; A metaheuristic algorithm is used to iteratively optimize the initial solution and output a solution scheme. Based on the solution scheme, the multi-objective nonlinear programming model is solved to obtain a candidate sampling scheme; Verify whether the candidate sampling plan satisfies the constraints; If the verification passes, the candidate sampling plan will be output as the sampling plan. If the verification fails, return to iterate and optimize again until the sampling plan is output.

13. The system according to claim 8, characterized in that, The evaluation module further includes: quantitatively evaluating the sampling plan based on a multi-dimensional evaluation index system; wherein the evaluation indicators of the multi-dimensional evaluation index system include risk coverage, cost control rate, sampling efficiency, compliance rate, and fault detection rate. The comprehensive score of each evaluation indicator is calculated using a weighted scoring model to obtain the evaluation result of the sampling plan; if the evaluation result does not meet the preset conditions, an iterative optimization process is triggered until a sampling plan that meets all evaluation indicators is generated.

14. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executed by the processor, characterized in that, When the processor executes the computer program, it implements the method as described in any one of claims 1 to 7.

15. A non-transitory computer-readable storage medium, characterized in that, The non-transitory computer-readable storage medium stores computer instructions for causing the computer to perform the method according to any one of claims 1 to 7.

16. A computer program product, characterized in that, It includes one or more computer programs that, when executed by one or more processors, implement the method as described in any one of claims 1 to 7.

Images