Adaptive task allocation method, system and device based on operation and maintenance data analysis

Abstract

The application discloses a self-adaptive task allocation method, system and device based on operation and maintenance data analysis, belongs to the technical field of intelligent operation and maintenance scheduling, and comprises the following steps: firstly, multi-dimensional operation and maintenance data of equipment, personnel, tasks and environment are collected in real time, standardized processing is carried out, and a multi-objective optimization adaptive model is constructed; then, an adaptive crossover mutation coefficient optimization genetic algorithm is used to solve an optimal allocation scheme, operation and maintenance scene characteristic changes are monitored in real time, model weight parameters are adaptively corrected, scheduling strategies are dynamically iteratively optimized, and finally, task issuing and log storage are completed. The application can realize intelligent and dynamic precise scheduling of the whole operation and maintenance task process, effectively improve operation and maintenance resource utilization and task completion quality, shorten equipment fault disposal time, reduce equipment downtime and operation and maintenance cost, adapt to various complex industrial operation and maintenance scenes, and has high intelligentization and engineering application value.

Description

Technical Field

[0001] This invention relates to the field of intelligent operation and maintenance scheduling technology, and in particular to an adaptive task allocation method, system and device based on operation and maintenance data analysis. Background Technology

[0002] In fields such as intelligent industrial manufacturing, internet data centers, and power grid maintenance, equipment operating continuously for extended periods can experience wear, malfunctions, and performance degradation. Routine inspections, emergency repairs, equipment debugging, and hardware replacements are essential to ensure stable equipment operation. The proper allocation of maintenance tasks directly determines maintenance efficiency, equipment stability, resource utilization, and cost control effectiveness.

[0003] Currently, the mainstream methods for allocating operation and maintenance tasks in the industry are mainly divided into two categories: manual experience-based allocation and fixed-rule intelligent allocation. Among them, manual experience-based allocation relies on the professional experience of operation and maintenance management personnel and makes rough allocations based on the number of operation and maintenance personnel and the number of tasks. It has a strong degree of subjectivity and arbitrariness and cannot take into account multiple factors such as personnel skill matching, equipment operating status, task urgency and difficulty, and on-site operation and maintenance environment. It is prone to problems such as personnel load imbalance, mismatch of high-difficulty tasks with low-skilled personnel, delayed processing of emergency tasks, and idle and wasted operation and maintenance resources. The accuracy and efficiency of operation and maintenance scheduling are extremely low, and it cannot be adapted to large-scale, high-frequency modern operation and maintenance scenarios.

[0004] The fixed-rule intelligent allocation method allocates tasks based on preset static rules, which offers a certain degree of standardization compared to manual allocation. However, it still has significant technical shortcomings: First, the data collection dimension is limited, simply collecting basic data such as the number of tasks and personnel on-duty status, ignoring core influencing parameters such as equipment operating wear and tear, fault level, personnel performance, on-site environment, and resource inventory. This fails to comprehensively represent the real state of the operation and maintenance scenario, resulting in insufficient allocation basis. Second, the model parameters are statically fixed, with the weights and constraint parameters of the allocation model being preset fixed values. This makes it impossible to adapt to dynamic changes in personnel load and equipment failures during operation and maintenance. The scheduling scheme suffers from severe lag due to dynamic changes such as sudden failures, fluctuations in environmental parameters, and temporary additions of tasks. Third, the optimization algorithm has poor versatility. Traditional genetic algorithms use fixed crossover and mutation coefficients, resulting in slow iterative convergence and a tendency to get trapped in local optima. They cannot find the globally optimal task allocation combination, leading to uneven task time consumption, low resource utilization, and unstable on-time completion rates. Fourth, there is a lack of dynamic correction and emergency scheduling mechanisms. When problems such as task backlog, insufficient resources, and declining performance occur during the execution of operation and maintenance tasks, the allocation scheme cannot be automatically optimized and can only be adjusted manually, resulting in low intelligence.

[0005] Most existing operation and maintenance scheduling solutions focus on a single optimization objective, pursuing only task completion efficiency or personnel load balancing, failing to consider multi-dimensional indicators such as resource utilization and task completion quality. Furthermore, they generally suffer from static scheduling, poor adaptability, and insufficient dynamic response capabilities. In complex industrial operation and maintenance scenarios, these shortcomings directly lead to chaotic operation and maintenance task scheduling, untimely equipment failure handling, wasted operation and maintenance resources, and increased equipment downtime losses, severely impacting the continuity and stability of industrial production and failing to meet the high-precision, dynamic, and high-efficiency scheduling requirements of modern intelligent operation and maintenance. Summary of the Invention

[0006] The purpose of this invention is to provide an adaptive task allocation method, system, and device based on operation and maintenance data analysis. It aims to solve the technical problems of existing operation and maintenance task allocation schemes, such as the inability to dynamically adapt to changes in operation and maintenance scenarios, low resource utilization, and unbalanced task scheduling. It realizes multi-dimensional perception, intelligent solution, dynamic correction, and precise scheduling of operation and maintenance tasks, and comprehensively improves the intelligence, precision, and adaptability of operation and maintenance scheduling.

[0007] To achieve the above objectives, this invention provides an adaptive task allocation method based on operation and maintenance data analysis, comprising the following steps: S1. Real-time collection of multi-dimensional raw operation and maintenance data of the target operation and maintenance scenario, and preprocessing to obtain a standardized operation and maintenance status feature set; S2. Based on the preset task allocation evaluation index, construct a multi-objective optimization and adaptation model; S3. Introduce an adaptive crossover mutation coefficient optimization genetic algorithm, combine it with a standardized operation and maintenance status feature set, iteratively solve the optimal task allocation combination, and generate an initial adaptive task allocation scheme. S4. Collect dynamic operation and maintenance data in real time during task execution, dynamically update the operation and maintenance status feature set, adaptively adjust the model weight parameters based on feature changes, and iteratively optimize the task allocation scheme. S5. The optimized task allocation scheme is sent to the terminal execution unit for operation and maintenance task scheduling, and the allocation log data is stored synchronously.

[0008] Preferably, in S1, the collected multidimensional raw operation and maintenance data includes equipment operating status data, operation and maintenance personnel load data, task attribute data to be executed, and operation and maintenance environment data; Equipment operating status data includes equipment base load, equipment runtime, fault alarms, hardware wear parameters, and equipment operation and maintenance permissions; Operations and maintenance personnel load data includes current task load, personnel skills, work status, historical operations and maintenance performance, and resource availability; The attribute data for tasks to be executed includes urgency, task difficulty, task type, constraint parameters, and historical correlation parameters; Operation and maintenance environment data includes physical environment, network transmission, cluster busy, on-site resources, and external constraint parameters.

[0009] Preferably, in S1, the collected multidimensional raw operation and maintenance data undergoes data cleaning, interpolation, and completion. Anomaly removal is performed by normalizing extreme values ​​to generate feature vectors of equipment operating status. Operations and maintenance personnel load characteristic vector Feature vector of attributes of the task to be executed and operation and maintenance environment feature vector The standardized operation and maintenance status feature set is obtained by merging. Its expression is: ; The normalization formula for a single feature vector is as follows: ; In the formula, These are the normalized eigenvalues. These are the original eigenvalues. The minimum value of the characteristic. This represents the maximum value of the characteristic.

[0010] Preferably, in S2, the preset task allocation evaluation indicators include the balance of task time consumption. Operation and maintenance resource utilization rate and on-time completion rate of tasks ; Task time balance The calculation formula is: ; In the formula, The total number of tasks to be assigned; For the first The actual time taken for each task; The average time taken for all tasks; Operation and maintenance resource utilization The calculation formula is: ; In the formula, This represents the amount of operational and maintenance resources already used. This represents the total available operational resources. On-time completion rate of tasks The calculation formula is: ; In the formula, To ensure the timely completion of the required tasks; The total number of tasks assigned.

[0011] Preferably, in S3, the formula for calculating the crossover coefficient in the adaptive crossover mutation coefficient optimization genetic algorithm is: ; In the formula, These are the maximum and minimum preset cross coefficient values, respectively. This represents the current iteration number; This represents the maximum number of iterations. The formula for calculating the coefficient of variation is: ; In the formula, These are the preset maximum and minimum values ​​of the coefficient of variation, respectively.

[0012] Preferably, in S4, the characteristic change amount The calculation formula is: ; In the formula, for A standardized set of operational and maintenance status features at any given time; for A standardized set of operational and maintenance status features at any given time; This is a 2-norm operation used to quantify the overall bias of the feature set; when Greater than the preset threshold At that time, the model weights are adaptively updated, and the weight update formula is: ; In the formula, for The updated model metric weight vector at each time step; for Original model weight vector at time step; Adjust the learning rate to the preset weights; This is the preset optimal weighted benchmark vector.

[0013] This invention provides an adaptive task allocation system based on operation and maintenance data analysis, used to implement the above-mentioned adaptive task allocation method, including: The data acquisition and preprocessing module is used to collect multi-dimensional raw operation and maintenance data in real time, and construct a standardized operation and maintenance status feature set after preprocessing. The model building module is used to build a multi-objective optimization and adaptation model based on preset task allocation evaluation indicators, and to define the model operation parameters and constraints. The intelligent task allocation solution module is used to iteratively solve the optimal task allocation scheme through an adaptive crossover and mutation coefficient optimization genetic algorithm, and output the initial scheduling strategy. The dynamic adaptive optimization module is used to monitor changes in operation and maintenance data in real time, calculate the amount of feature change, and adaptively adjust the model weights and task allocation scheme. The scheduling execution storage module is used to issue task allocation schemes and record and store log data of the entire operation and maintenance scheduling process.

[0014] Preferably, the data acquisition and preprocessing module includes a data acquisition unit, a data cleaning unit, and a normalization processing unit; The data acquisition unit is used to collect data from equipment sensors, personnel terminals, and the task management backend. The data cleaning unit is used to remove outliers, missing values, and redundant data. The normalization processing unit is used to standardize and transform feature data.

[0015] The present invention also provides an adaptive task allocation device based on operation and maintenance data analysis, which is equipped with the above system and includes a front-end monitoring terminal, an operation and maintenance execution terminal, a data storage server and a network transmission module; The front-end monitoring terminal is used to display operation and maintenance status data, task allocation results, and scheduling optimization logs; The operation and maintenance execution terminal is used to receive assigned operation and maintenance tasks and provide feedback on task execution progress and abnormal data. Data storage servers are used to store raw operation and maintenance data, standardized feature sets, model parameters, allocation logs, and task execution data; The network transmission module is used to enable real-time data interaction between various modules.

[0016] Preferably, it also includes a built-in anomaly warning unit. When the detected task backlog exceeds a preset threshold, resource utilization is lower than a preset lower limit, or task completion rate is lower than a standard value, the anomaly warning unit issues an early warning and initiates an emergency task reallocation strategy.

[0017] Therefore, the present invention employs the above-mentioned adaptive task allocation method, system, and device based on operation and maintenance data analysis, and the beneficial effects are as follows: (1) This invention collects multi-dimensional operation and maintenance parameters in full coverage, constructs an accurate feature set through standardized processing, and comprehensively represents the operation and maintenance scenario status. It solves the problem of insufficient data dimensions and inaccurate representation from the source, and greatly improves the accuracy and rationality of task allocation.

[0018] (2) The present invention adopts an adaptive cross-mutation coefficient optimization genetic algorithm to avoid the defects of local optima. The multi-objective model achieves the optimal balance, utilization rate and completion rate in a coordinated manner. The dynamic weight correction adapts to the changes in the scenario. The measured task balance and resource utilization rate are significantly improved.

[0019] (3) The adaptive task allocation system constructed in this invention has a built-in abnormal early warning and emergency reassignment mechanism, which can quickly resolve scheduling abnormalities, improve operation and maintenance stability, and reduce equipment downtime losses and operation and maintenance costs.

[0020] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0021] Figure 1 This is an overall flowchart of an embodiment of the adaptive task allocation method based on operation and maintenance data analysis of the present invention; Figure 2 This is a schematic diagram of the structure of an embodiment of the adaptive task allocation device based on operation and maintenance data analysis of the present invention; Figure 3 This is an overall structural block diagram of an embodiment of the adaptive task allocation system based on operation and maintenance data analysis of the present invention; Figure 4 This is a schematic diagram showing the comparison of comprehensive performance indicators of different operation and maintenance task allocation schemes in embodiments of the adaptive task allocation method, system and device based on operation and maintenance data analysis of the present invention. Figure 5 This is a schematic diagram showing the performance comparison between the traditional genetic algorithm and the adaptive optimization algorithm of this invention in an embodiment of the adaptive task allocation method, system and device based on operation and maintenance data analysis. Detailed Implementation

[0022] The technical solution of the present invention will be further described below with reference to the accompanying drawings and embodiments.

[0023] Unless otherwise defined, the technical or scientific terms used in this invention shall have the ordinary meaning understood by one of ordinary skill in the art to which this invention pertains. The terms "first," "second," and similar terms used in this invention 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.

[0024] like Figure 1 As shown, the adaptive task allocation method based on operation and maintenance data analysis includes the following steps: S1. Establish a full-domain operation and maintenance data collection link. Collect multi-dimensional raw operation and maintenance data of the target operation and maintenance scenario in real time through equipment sensors, personnel operation and maintenance terminals, task management backend and environmental monitoring equipment. After preprocessing, obtain a standardized operation and maintenance status feature set.

[0025] The multidimensional raw operation and maintenance data collected by this invention includes equipment operating status data, operation and maintenance personnel load data, task attribute data to be executed, and operation and maintenance environment data.

[0026] Equipment operating status data includes equipment base load, equipment runtime, fault alarms, hardware wear parameters, and equipment operation and maintenance permissions, which are used to characterize the health of equipment operation and operation and maintenance needs.

[0027] The equipment's basic load includes CPU utilization, memory usage, hardware load, unit output power, and server bandwidth utilization. Equipment runtime includes continuous runtime for the day, total cumulative runtime, and the interval since the last maintenance. Fault alarms include the number of real-time fault alarms, historical fault frequency, fault severity level, fault recurrence probability, and equipment aging coefficient. Hardware wear and tear includes component wear, remaining service life, and equipment downtime risk. Equipment maintenance permissions include the equipment maintenance team identifier and the equipment operation qualification threshold level.

[0028] Operations and maintenance personnel load data includes current task load, personnel skills, work status, historical operations and maintenance performance, and resource holdings, which are used to characterize the workload and operational capabilities of operations and maintenance personnel.

[0029] The current workload includes the number of accepted but unfinished tasks, the cumulative estimated working hours for each type of task, and the remaining working hours for the day. Personnel skills include various maintenance skill levels, fault repair proficiency, and equipment operation qualifications. Work status includes on-duty or off-duty status, overtime hours worked that day, and temporary leave status. Historical maintenance performance includes historical task on-time completion rate, historical maintenance error count, and average maintenance processing speed. Resource availability includes the availability of maintenance tools and access to testing equipment.

[0030] The attribute data of the tasks to be executed includes urgency, task difficulty, task type, constraint parameters, and historical correlation parameters, which are used to characterize the execution requirements and constraints of the operation and maintenance tasks.

[0031] The urgency level includes the task urgency grade, estimated downtime loss due to failure, and customer repair timeframe. Task difficulty includes the number of operation steps, required skill levels, estimated completion time, and the need for multi-person collaboration. Task types include routine inspections, emergency repairs, equipment debugging, hardware replacement, and system upgrades. Constraint parameters include the task execution time window, maximum allowable delay, and required supporting equipment resources. Historical correlation parameters include the historical average time for similar tasks and the historical recurrence probability of similar tasks.

[0032] Operation and maintenance environment data includes physical environment, network transmission, cluster busy, on-site resources, and external constraint parameters, which are used to characterize the external environmental constraints of operation and maintenance operations.

[0033] The physical environment includes temperature, humidity, dust concentration, and equipment heat dissipation conditions in the workshop or computer room. Network transmission includes on-site network latency, bandwidth packet loss rate, and remote maintenance channel connectivity. Cluster congestion includes the number of simultaneous failures in the overall equipment cluster and the total number of global maintenance tasks. On-site resources include the number of on-site backup maintenance tools and spare parts inventory. External constraints include plant power outage plans, workshop production scheduling windows, and restricted on-site operation periods.

[0034] Data cleaning, interpolation, and completion are performed on the collected multidimensional raw operation and maintenance data. Anomaly removal is performed by normalizing extreme values ​​to generate feature vectors of equipment operating status. Operations and maintenance personnel load characteristic vector Feature vector of attributes of the task to be executed and operation and maintenance environment feature vector The four categories of standardized feature vectors are merged to obtain a standardized operation and maintenance status feature set that is dimensionally unified, data compliant, and time-series corresponding. Standardized Operation and Maintenance Status Feature Set Its expression is: .

[0035] The normalization formula for a single feature vector is as follows: ; In the formula, These are the normalized eigenvalues, with values ​​ranging from [0,1]. These are the original eigenvalues. The minimum value of the characteristic. This represents the maximum value of the characteristic.

[0036] S2. Based on the preset task allocation evaluation index, construct a multi-objective optimization and adaptation model; Based on the core business requirements of operation and maintenance task allocation, the selected preset task allocation evaluation indicators include task time balance. Operation and maintenance resource utilization rate and on-time completion rate of tasks .

[0037] Task time balance This value is used to measure the uniformity of the workload among maintenance personnel. The smaller the value, the more balanced the load distribution. The calculation formula is: ; In the formula, The total number of tasks to be assigned; For the first The actual time taken for each task; The average time taken for all tasks.

[0038] Operation and maintenance resource utilization This value is used to measure the effective utilization of global operation and maintenance resources. A higher value indicates more efficient resource utilization. The calculation formula is as follows: ; In the formula, This refers to the amount of operational and maintenance resources already in use, including the total amount of operational and maintenance personnel, equipment, and tools. This represents the total available operational resources that the system can schedule.

[0039] On-time completion rate of tasks This value is used to measure the timeliness and accuracy of operation and maintenance task scheduling. A higher value indicates a more reasonable scheduling scheme. The calculation formula is as follows: ; In the formula, To ensure the timely completion of the required tasks; The total number of tasks assigned.

[0040] S3. To address the shortcomings of traditional genetic algorithms, such as slow convergence and susceptibility to local optima due to fixed crossover and mutation coefficients, this invention introduces an adaptive crossover and mutation coefficient optimization genetic algorithm. Using the standardized operational status feature set constructed in S1 as model input, the algorithm initializes operational parameters such as the genetic population, iteration count, and convergence threshold. By adaptively adjusting the crossover and mutation coefficients, the algorithm continuously iteratively updates the task allocation combinations corresponding to individuals in the population, constantly filtering for the optimal solution. When the algorithm reaches the maximum number of iterations or when the optimal solution of the population remains unchanged for multiple consecutive generations, the algorithm is considered to have converged, and the globally optimal initial adaptive task allocation scheme is output.

[0041] In the adaptive crossover mutation coefficient optimization genetic algorithm of this invention, the adaptive crossover coefficient decreases linearly with the number of iterations. The crossover coefficient is larger in the early stages, improving population diversity, and smaller in the later stages, stabilizing the optimal solution. Its calculation formula is as follows: ; In the formula, These are the maximum and minimum preset cross coefficient values, respectively. This represents the current iteration number; This represents the maximum number of iterations.

[0042] The coefficient of variation increases linearly with the number of iterations. The coefficient of variation is smaller in the early stages to avoid destroying the optimal solution, and larger in the later stages to escape local optima. Its calculation formula is as follows: ; In the formula, These are the preset maximum and minimum values ​​of the coefficient of variation, respectively.

[0043] S4. Throughout the entire lifecycle of the operation and maintenance task, collect dynamic operation and maintenance data in real time during the task execution process, update the operation and maintenance status feature set synchronously and dynamically, calculate the change in operation and maintenance status features between the current time and the previous time using the time-series difference algorithm, and compare the change in features with the preset sensitivity threshold.

[0044] Among them, the change in operation and maintenance status characteristics The calculation formula is: ; In the formula, for A standardized set of operational and maintenance status features at any given time; for A standardized set of operational and maintenance status features at any given time; This is a 2-norm operation used to quantify the overall bias of a feature set.

[0045] When the change in operation and maintenance status characteristics Greater than the preset threshold When a significant change in the operational scenario is detected, the model weights are adaptively updated, and the weight parameters of each indicator in the multi-objective optimization adaptation model are dynamically adjusted. The weight update formula is as follows: ; In the formula, for The updated model metric weight vector at each time step; for Original model weight vector at time step; Adjust the learning rate to the preset weights; This is the preset optimal weighted benchmark vector.

[0046] Based on the updated model parameters, the calculation is iterated again to optimize the task allocation scheme in real time, so as to realize the allocation strategy dynamically and adaptively adjusts according to the operation and maintenance scenario; when the feature change is less than or equal to the preset threshold, the current allocation scheme is maintained to ensure scheduling stability.

[0047] S5. The optimal task allocation scheme after iterative optimization is sent to the terminal execution unit through the network link for operation and maintenance task scheduling, generating standardized operation and maintenance scheduling logs. The original operation and maintenance data, standardized feature sets, model weight parameters, task allocation schemes of various versions, and task execution results are classified, archived, and encrypted for storage, so as to realize the traceability, review, and iterative optimization of the entire operation and maintenance scheduling process.

[0048] like Figure 3 As shown, the adaptive task allocation system based on operation and maintenance data analysis of the present invention is used to implement the above-mentioned adaptive task allocation method, including: The data acquisition and preprocessing module serves as the system's data entry point, used to collect multi-dimensional raw operation and maintenance data in real time, such as equipment, personnel, and task environment data. After preprocessing, it constructs a standardized operation and maintenance status feature set. The data acquisition and preprocessing module includes a data acquisition unit, a data cleaning unit, and a normalization processing unit.

[0049] The data acquisition unit is used to collect data from equipment sensors, personnel terminals, and the task management backend. It collects four types of raw data in real time: equipment operation, personnel load, task attributes, and operation and maintenance environment through encrypted network protocols. It supports millisecond-level data acquisition and time-series marking to ensure data real-time performance and time-series consistency.

[0050] The data cleaning unit is used to automatically identify and remove outliers, interpolate and repair missing data, delete duplicate and invalid redundant data, and complete the compliance purification of raw data.

[0051] The normalization processing unit is used to standardize and transform feature data, generate standardized feature vectors with consistent value ranges, and merge them to construct a structured and standardized operation and maintenance status feature set.

[0052] The model building module is used to build a multi-objective optimization and adaptation model based on preset task allocation evaluation indicators, and to define the model operation parameters and constraints.

[0053] The intelligent task allocation solution module is used to iteratively solve the optimal task allocation scheme through an adaptive crossover and mutation coefficient optimization genetic algorithm, and outputs the initial scheduling strategy after the convergence condition is met.

[0054] The dynamic adaptive optimization module is used to monitor changes in operation and maintenance data in real time, calculate the amount of feature change, determine the scenario fluctuation state based on the change threshold, and adaptively adjust the model weights and task allocation scheme to realize the scheduling strategy to dynamically adapt to changes in the operation and maintenance scenario.

[0055] The scheduling execution and storage module is used to issue task allocation schemes, manage the implementation process of operation and maintenance task scheduling, and record and store the log data of the entire operation and maintenance scheduling process.

[0056] like Figure 2 As shown, the adaptive task allocation device based on operation and maintenance data analysis of the present invention is equipped with the above-mentioned adaptive task allocation system and includes a front-end monitoring terminal, an operation and maintenance execution terminal, a data storage server and a network transmission module.

[0057] The front-end monitoring terminal uses a visual interactive interface to display real-time operation and maintenance status data, task allocation results, and scheduling optimization logs. It supports manual viewing, parameter fine-tuning, and solution intervention, realizing visualized management and control of the entire operation and maintenance scheduling process.

[0058] The operation and maintenance execution terminal is deployed in various operation and maintenance positions to receive assigned operation and maintenance tasks and provide feedback on task execution progress and abnormal data.

[0059] The data storage server adopts a distributed encrypted storage architecture to classify and store raw operation and maintenance data, standardized feature sets, model parameters, allocation logs, and task execution data.

[0060] The network transmission module adopts a dual-link redundant transmission mechanism, which can transmit encrypted data via wired or wireless means to achieve real-time data interaction between modules, ensuring stable data transmission, low latency, and high security.

[0061] The adaptive task allocation system also includes a built-in anomaly warning unit. When the system detects that the backlog of tasks exceeds a preset threshold, the resource utilization rate is lower than a preset lower limit, or the task completion rate is lower than a standard value, the anomaly warning unit issues an alert and initiates an emergency task reallocation strategy.

[0062] Simultaneously, an emergency task redistribution strategy is automatically activated: the inefficient task allocation process is paused, and data on the load of maintenance personnel, equipment resources, and environmental status across the entire domain are retrieved again. The emergency allocation plan is quickly iterated and optimized, and the priority of backlogged tasks and timeout risk tasks is reordered. Priority is given to matching maintenance personnel with high availability, high skills, and high performance, so as to quickly resolve the problems of task backlog, resource waste, and inefficient scheduling, and ensure the stable and efficient operation of the maintenance scheduling system.

[0063] Example 1: To verify the practical application performance and technical advantages of the present invention, a comparative experiment was conducted in a smart operation and maintenance scenario of an industrial intelligent manufacturing workshop. The scenario included 28 production equipment and 12 operation and maintenance personnel. The daily operation and maintenance tasks included equipment inspection, fault repair, hardware debugging, spare parts replacement, etc. There were an average of 80 to 120 operation and maintenance tasks to be assigned per day. The scenario was characterized by complex task types, large differences in personnel workload, dynamic fluctuations in equipment status, and a changeable operation and maintenance environment.

[0064] This embodiment sets up three control schemes: the traditional manual experience allocation scheme in control example 1, the traditional fixed rule intelligent allocation scheme in control example 2, and the adaptive task allocation scheme of the present invention. All three experiments were conducted under the same equipment operating conditions, the same total amount of maintenance tasks, and the same personnel configuration. The tests were conducted continuously for 30 working days, and the average value of the core indicators of daily maintenance scheduling was collected to complete the performance comparison and verification.

[0065] 1. Experimental environment: The industrial intelligent manufacturing production workshop includes 28 CNC machining equipment, intelligent testing equipment, and server clusters; there are 12 maintenance personnel with varying skill levels, years of experience, and workloads; and an average of 95 maintenance tasks per day, covering various types of tasks such as emergency fault tasks, routine inspection tasks, and equipment debugging tasks.

[0066] The core parameters of the algorithm in this invention are set as follows: the maximum value of the adaptive crossover coefficient is 0.9 and the minimum value is 0.4; the maximum value of the adaptive coefficient of variation is 0.1 and the minimum value is 0.01; the maximum number of iterations of the algorithm is 100; the preset threshold for feature change is 0.08; the weight correction learning rate is 0.15; and the model weight baseline vector is the industry-optimal standard parameter.

[0067] 2. Comparison with the scheme settings: Comparative Example 1: Traditional manual experience-based task allocation involves operations and maintenance personnel roughly assigning tasks based on their subjective experience and the availability of staff, without standardized data references or model optimization logic.

[0068] Comparative Example 2: Traditional fixed-rule intelligent allocation uses a static allocation model and a fixed-coefficient genetic algorithm. It only collects basic data such as the number of tasks and the number of personnel on duty. The model weights and algorithm parameters remain fixed and there is no dynamic correction mechanism.

[0069] Example: The present invention employs a multi-dimensional data acquisition, adaptive optimization genetic algorithm, and dynamic weight correction adaptive task allocation scheme, executing the entire process steps and equipped with an adaptive task allocation system and device.

[0070] The experiment selected three core evaluation indicators: task time balance, operation and maintenance resource utilization, and task on-time completion rate. Task scheduling time, task mismatch rate, and average equipment failure handling time were also used as auxiliary verification indicators. Task time balance measures the uniformity of task load among operation and maintenance personnel; a lower value indicates a more balanced load distribution, with the optimal value approaching 0. Operation and maintenance resource utilization measures the effective utilization of global operation and maintenance resources; a higher value indicates more efficient resource utilization, with an optimal standard of 100%. Task on-time completion rate measures the timeliness and accuracy of operation and maintenance task scheduling; a higher value indicates a more reasonable scheduling scheme, with an optimal standard of 100%. Task scheduling time refers to the time required for a single task scheduling; a lower value indicates a faster scheduling response. Task mismatch rate characterizes the degree of mismatch between tasks and operation and maintenance personnel and resources; a lower value indicates higher matching accuracy. Average failure handling time refers to the overall time spent handling equipment failures; a lower value indicates higher operation and maintenance efficiency. The average test data for the three schemes are shown in Table 1 below. Figure 4 As shown.

[0071] Table 1. Comparison of Comprehensive Performance Indicators of Different Operation and Maintenance Task Allocation Schemes

[0072] To verify the superiority of the adaptive crossover mutation coefficient optimization genetic algorithm of this invention, the iterative convergence speed, optimal solution accuracy, and local optimum escape ability of the traditional fixed parameter genetic algorithm (Comparative Example 2) and the optimization algorithm of this invention were compared. The test results are shown in Table 2 below. Figure 5 As shown.

[0073] Table 2 Performance Comparison of Traditional Genetic Algorithm and Adaptive Optimization Algorithm of the Present Invention

[0074] To further verify the dynamic weight correction and adaptive optimization capabilities of this invention, dynamic changes in operation and maintenance scenarios were simulated, including sudden equipment failures, temporary addition of operation and maintenance tasks, temporary absence of personnel, and surge in cluster load. The performance of the three schemes under dynamic scenarios was tested, and the test results are shown in Table 3 below.

[0075] Table 3. Comparison of Adaptive Performance of Different Allocation Schemes in Dynamic Operation and Maintenance Scenarios

[0076] Based on the above three sets of comparative experimental data, the multiple technical advantages of the present invention can be clearly verified, as follows: This invention comprehensively covers the influencing factors of operation and maintenance scenarios by collecting and standardizing data across four dimensions: equipment, personnel, tasks, and environment. The task time balance is reduced by 80.1% compared to manual allocation schemes and by 64.7% compared to traditional fixed rule schemes. It completely solves the problems of unbalanced personnel load and task mismatch in traditional schemes, with a task mismatch rate as low as 1.82% and significantly improved scheduling accuracy.

[0077] This invention employs an adaptive crossover mutation coefficient optimization genetic algorithm, which, compared to the traditional fixed parameter algorithm, reduces the number of convergence iterations by 51.4%, significantly lowers the probability of local optimum traps, and improves the accuracy of finding the global optimum by 15.47%. It effectively solves the technical defects of traditional algorithms, such as slow convergence and easy trapping in local optima, and greatly improves the optimality of task allocation schemes.

[0078] This invention constructs a multi-objective optimization and adaptation model that takes into account multiple objectives such as load balancing, resource utilization, and task timeliness. The utilization rate of operation and maintenance resources is improved by 29.83 percentage points compared with manual allocation and by 16.56 percentage points compared with traditional fixed rule schemes. The on-time completion rate of tasks reaches 97.86%, and the average fault handling time is reduced by more than 50%, which significantly improves the efficiency of operation and maintenance work and reduces equipment downtime losses.

[0079] Faced with sudden changes in operational scenarios, traditional manual solutions and fixed-rule solutions cannot achieve autonomous adaptation, easily leading to problems such as task backlog, resource waste, and task timeouts. This invention can quantify the changes in scenario features in real time, adaptively adjust model weight parameters, and complete solution iteration and optimization in seconds. At the same time, it relies on an anomaly early warning unit to quickly initiate emergency redistribution strategies. Even in complex and dynamic scenarios, it can maintain a task on-time completion rate of over 94%, and its scenario adaptability and operational stability far exceed those of traditional solutions.

[0080] This invention automates the entire process of operation and maintenance data collection, model calculation, solution solving, dynamic optimization, and log storage. A single scheduling takes only 6.5 seconds, which is more than 70% more efficient than traditional scheduling methods. At the same time, the entire process log is archived and stored, making operation and maintenance scheduling traceable, replayable, and iteratively optimized, adapting to the needs of large-scale, high-frequency modern intelligent operation and maintenance scenarios.

[0081] In summary, the adaptive task allocation method, system, and device based on operation and maintenance data analysis of this invention completely solves the technical pain points of existing manual allocation, such as strong subjectivity, static and rigid traditional intelligent allocation, poor dynamic adaptability, low resource utilization, and uneven scheduling, through the coordinated cooperation of multi-dimensional global data standardization modeling, adaptive optimization genetic algorithm, dynamic weight correction mechanism, and abnormal emergency scheduling mechanism.

[0082] Therefore, the present invention adopts the above-mentioned adaptive task allocation method, system and device based on operation and maintenance data analysis, which solves the pain points of traditional operation and maintenance allocation, such as strong subjectivity, poor adaptability, uneven scheduling and low resource utilization. It realizes intelligent operation and maintenance scheduling throughout the entire process, effectively improves the accuracy and efficiency of operation and maintenance scheduling, reduces operation and maintenance and equipment downtime losses, and has extremely high application and promotion value in industrial intelligent operation and maintenance.

[0083] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the technical solutions of the present invention, and these modifications or equivalent substitutions cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. An adaptive task allocation method based on operation and maintenance data analysis, characterized in that, Includes the following steps: S1. Real-time collection of multi-dimensional raw operation and maintenance data of the target operation and maintenance scenario, and preprocessing to obtain a standardized operation and maintenance status feature set; S2. Based on the preset task allocation evaluation index, construct a multi-objective optimization and adaptation model; S3. Introduce an adaptive crossover mutation coefficient optimization genetic algorithm, combine it with a standardized operation and maintenance status feature set, iteratively solve the optimal task allocation combination, and generate an initial adaptive task allocation scheme. S4. Collect dynamic operation and maintenance data in real time during task execution, dynamically update the operation and maintenance status feature set, adaptively adjust the model weight parameters based on feature changes, and iteratively optimize the task allocation scheme. S5. The optimized task allocation scheme is sent to the terminal execution unit for operation and maintenance task scheduling, and the allocation log data is stored synchronously.

2. The adaptive task allocation method based on operation and maintenance data analysis according to claim 1, characterized in that, In S1, the collected multidimensional raw operation and maintenance data includes equipment operating status data, operation and maintenance personnel load data, task attribute data to be executed, and operation and maintenance environment data. Equipment operating status data includes equipment base load, equipment runtime, fault alarms, hardware wear parameters, and equipment operation and maintenance permissions; Operations and maintenance personnel load data includes current task load, personnel skills, work status, historical operations and maintenance performance, and resource availability; The attribute data for tasks to be executed includes urgency, task difficulty, task type, constraint parameters, and historical correlation parameters; Operation and maintenance environment data includes physical environment, network transmission, cluster busy, on-site resources, and external constraint parameters.

3. The adaptive task allocation method based on operation and maintenance data analysis according to claim 2, characterized in that, In S1, the collected multidimensional raw operation and maintenance data is cleaned, interpolated, and completed. Anomaly removal is performed by normalizing extreme values ​​to generate feature vectors of equipment operating status. Operations and maintenance personnel load characteristic vector Feature vector of attributes of the task to be executed and operation and maintenance environment feature vector The standardized operation and maintenance status feature set is obtained by merging. Its expression is: ； The normalization formula for a single feature vector is as follows: ； In the formula, These are the normalized eigenvalues. These are the original eigenvalues. The minimum value of the characteristic. This represents the maximum value of the characteristic.

4. The adaptive task allocation method based on operation and maintenance data analysis according to claim 3, characterized in that, In S2, the preset task allocation evaluation indicators include the balance of task time consumption. Operation and maintenance resource utilization rate and on-time completion rate of tasks ; Task time balance The calculation formula is: ； In the formula, The total number of tasks to be assigned; For the first The actual time taken for each task; The average time taken for all tasks; Operation and maintenance resource utilization The calculation formula is: ； In the formula, This represents the amount of operational and maintenance resources already used. This represents the total available operational resources. On-time completion rate of tasks The calculation formula is: ； In the formula, To ensure the timely completion of the required tasks; The total number of tasks assigned.

5. The adaptive task allocation method based on operation and maintenance data analysis according to claim 4, characterized in that, In S3, the formula for calculating the crossover coefficient in the adaptive crossover mutation coefficient optimization genetic algorithm is: ； In the formula, These are the maximum and minimum preset cross coefficient values, respectively. This represents the current iteration number; This represents the maximum number of iterations. The formula for calculating the coefficient of variation is: ； In the formula, These are the preset maximum and minimum values ​​of the coefficient of variation, respectively.

6. The adaptive task allocation method based on operation and maintenance data analysis according to claim 5, characterized in that, In S4, the characteristic change amount The calculation formula is: ； In the formula, for A standardized set of operational and maintenance status features at any given time; for A standardized set of operational and maintenance status features at any given time; This is a 2-norm operation used to quantify the overall bias of the feature set; when Greater than the preset threshold At that time, the model weights are adaptively updated, and the weight update formula is: ； In the formula, for The updated model metric weight vector at each time step; for Original model weight vector at time step; Adjust the learning rate to the preset weights; This is the preset optimal weighted benchmark vector.

7. An adaptive task allocation system based on operation and maintenance data analysis, used to implement the adaptive task allocation method according to any one of claims 1-6, characterized in that, include: The data acquisition and preprocessing module is used to collect multi-dimensional raw operation and maintenance data in real time, and construct a standardized operation and maintenance status feature set after preprocessing. The model building module is used to build a multi-objective optimization and adaptation model based on preset task allocation evaluation indicators, and to define the model operation parameters and constraints. The intelligent task allocation solution module is used to iteratively solve the optimal task allocation scheme through an adaptive crossover and mutation coefficient optimization genetic algorithm, and output the initial scheduling strategy. The dynamic adaptive optimization module is used to monitor changes in operation and maintenance data in real time, calculate the amount of feature change, and adaptively adjust the model weights and task allocation scheme. The scheduling execution storage module is used to issue task allocation schemes and record and store log data of the entire operation and maintenance scheduling process.

8. The adaptive task allocation system based on operation and maintenance data analysis according to claim 7, characterized in that, The data acquisition and preprocessing module includes a data acquisition unit, a data cleaning unit, and a normalization processing unit; The data acquisition unit is used to collect data from equipment sensors, personnel terminals, and the task management backend. The data cleaning unit is used to remove outliers, missing values, and redundant data. The normalization processing unit is used to standardize and transform feature data.

9. An adaptive task allocation device based on operation and maintenance data analysis, equipped with the system according to any one of claims 7-8, characterized in that, It includes a front-end monitoring terminal, an operation and maintenance execution terminal, a data storage server, and a network transmission module; The front-end monitoring terminal is used to display operation and maintenance status data, task allocation results, and scheduling optimization logs; The operation and maintenance execution terminal is used to receive assigned operation and maintenance tasks and provide feedback on task execution progress and abnormal data. Data storage servers are used to store raw operation and maintenance data, standardized feature sets, model parameters, allocation logs, and task execution data; The network transmission module is used to enable real-time data interaction between various modules.

10. The adaptive task allocation device based on operation and maintenance data analysis according to claim 9, characterized in that, It also includes a built-in anomaly warning unit. When the detected task backlog exceeds a preset threshold, resource utilization is lower than a preset lower limit, or task completion rate is lower than a standard value, the anomaly warning unit issues an alert and initiates an emergency task reallocation strategy.

Images