Equipment maintenance task priority evaluation method and system thereof

Abstract

The present application belongs to the technical field of equipment maintenance support decision, discloses a kind of equipment maintenance task priority evaluation method and system, the present application obtains maintenance task dataset and attribute evaluation set, based on scene discrimination parameter selection evaluation path;In hesitant fuzzy evaluation path, hesitant fuzzy matrix is constructed, the expected value and variance value are calculated, the weight optimization objective function is constructed based on attribute correlation degree to solve weight vector, normalization and comprehensive evaluation sorting are carried out;In analytic hierarchy process-approximate ideal solution evaluation path, hierarchical structure model and judgment matrix are constructed to calculate weight, standardized using Logistic function, determine positive and negative ideal solution, calculate closeness degree and sort, the present application realizes the unity of wartime rapid response and daily fine evaluation by the scene adaptive double-path evaluation mechanism, effectively solves the problem of insufficient description of decision ambiguity and lack of dynamic adaptation ability.

Description

Technical Field

[0001] This invention belongs to the field of equipment maintenance and support decision-making technology, and in particular relates to a method and system for evaluating the priority of equipment maintenance tasks. Background Technology

[0002] Prioritizing equipment maintenance tasks is a core component of the efficient operation of the equipment support system, directly determining the efficiency of maintenance resource allocation, the speed of equipment combat capability recovery, and the continuity of combat mission execution. With the increasing complexity of equipment technology and the diversification of mission scenarios, maintenance tasks exhibit significant characteristics such as multiple constraints, dynamism, and wide-ranging impacts. Traditional single-dimensional or static evaluation methods are no longer adequate for practical needs. Against the backdrop of the accelerating evolution of modern warfare, the challenges facing equipment maintenance support are becoming increasingly complex. It is necessary not only to consider the technical condition of the equipment itself but also to comprehensively consider multiple factors such as combat mission requirements, logistical support capabilities, and battlefield environmental conditions. This places higher demands on maintenance task prioritization methods.

[0003] Existing maintenance task prioritization techniques suffer from the following technical problems: First, traditional weighting methods have significant limitations. Existing weighting methods such as the analytic hierarchy process (AHP) and entropy weighting often overlook the ambiguity of decision-makers when evaluating maintenance tasks. Due to insufficient information and differences in decision-makers' experience, they often hesitate in judging the importance of indicators, making it difficult to provide a unique and definite evaluation result. The traditional method's requirement for a single value can easily lead to weight bias, thus affecting the reliability of the evaluation results. Especially under the high-pressure environment of wartime, commanders and maintenance experts often need to make judgments under conditions of incomplete information and time constraints. This objective existence of decision-making uncertainty requires evaluation methods to accommodate and reasonably handle the hesitant attitudes of decision-makers.

[0004] Secondly, existing evaluation methods lack the ability to adapt to dynamic changes in tasks in real time. In actual equipment maintenance and support processes, maintenance tasks are updated in real time as the combat progresses and resource status changes. Fixed weight allocation schemes cannot reflect the impact of task changes on the importance of indicators. Especially in wartime scenarios, there is a need for rapid response to dynamic adjustments in tasks, a requirement that current technologies struggle to meet. The continuous changes in equipment combat damage, the dynamic consumption and replenishment of maintenance resources, and the real-time evolution of the combat situation all demand that priority evaluation systems have the ability to quickly recalculate and dynamically update. However, existing methods are mostly static, one-off evaluations, lacking a response mechanism to changes in the task set.

[0005] Secondly, traditional evaluation methods do not adequately consider the correlations between attributes. Most existing multi-attribute decision-making methods assume that the evaluation indicators are independent, failing to fully consider the impact of the correlation between attributes on weight allocation. In equipment maintenance task evaluation, there are complex correlations between attributes such as equipment importance, repair time, resource requirements, and location. Ignoring these correlations can lead to unreasonable weight allocation, thus affecting the accuracy of priority ranking. For example, there is a positive correlation between equipment location and repair time; more remote locations usually mean longer maneuver times and more difficult support conditions. Similarly, there is a correlation between resource requirements and repair time; complex faults often require more resources and longer time simultaneously.

[0006] Furthermore, existing technologies lack differentiated evaluation strategies for different application scenarios. The requirements for evaluation methods differ significantly between wartime dynamic scenarios and routine, multi-constraint scenarios: the former requires rapid response and a concise indicator system, capable of prioritizing large batches of tasks within minutes; the latter requires refined evaluation and multi-dimensional coverage, fully considering long-term factors such as economy and sustainability. Existing single evaluation methods struggle to simultaneously meet the needs of different scenarios, failing to achieve the decision support goal of maintaining speed in wartime and accuracy in routine operations. In practical applications, decision-makers often need to handle both urgent and routine planned tasks within the same system, requiring evaluation methods to possess scenario awareness and adaptive path switching capabilities.

[0007] Existing equipment maintenance task priority evaluation technologies suffer from problems such as insufficient characterization of decision fuzziness, lack of dynamic adaptation capability, insufficient consideration of attribute correlation, and poor scenario adaptability. There is an urgent need for a new evaluation method that can take into account decision reliability, real-time responsiveness, and scenario adaptability.

[0008] Based on the above analysis, the problems and shortcomings of the existing technology are as follows:

[0009] Existing equipment maintenance task priority evaluation technologies suffer from problems such as insufficient characterization of decision fuzziness, lack of dynamic adaptation capability, insufficient consideration of attribute correlation, and poor scenario adaptability. Summary of the Invention

[0010] To address the problems existing in the prior art, this invention provides a method and system for prioritizing equipment maintenance tasks.

[0011] This invention is implemented as follows: A method for prioritizing equipment maintenance tasks includes:

[0012] Step 1, Data Acquisition Steps;

[0013] Obtain the maintenance task dataset and attribute evaluation set, and determine the evaluation path based on the scenario discrimination parameters; when the scenario discrimination parameters are less than the preset scenario threshold, execute the hesitant fuzzy evaluation path:

[0014] Step 2, hesitant fuzzy modeling steps;

[0015] A hesitant fuzzy matrix is ​​constructed based on the evaluation data of multiple decision-makers, and the expected value and variance value of each hesitant fuzzy element are calculated.

[0016] Step 3, dynamic weight calculation steps;

[0017] Construct a correlation matrix, and build a weight optimization objective function based on hesitant fuzzy meta-features, attribute dispersion, and attribute correlation. Solve the attribute weight vector using the Lagrange multiplier method.

[0018] Step 4: Comprehensive evaluation and ranking steps;

[0019] The attribute values ​​are normalized to obtain a normalized attribute matrix. A comprehensive evaluation score is calculated based on the normalized attribute matrix and the attribute weight vector. The scores are then sorted in descending order to generate a priority ranking sequence. When the scene discrimination parameter is greater than or equal to a preset scene threshold, the analytic hierarchy process (AHP) – approximation of the ideal solution evaluation path is executed.

[0020] Step 5, Hierarchical Analysis weight calculation steps;

[0021] Construct a hierarchical structure model and a judgment matrix, calculate the weights of the criterion layer and the index layer, and verify the effectiveness of the judgment matrix through the consistency ratio value;

[0022] Step 6, sorting steps to approximate the ideal solution;

[0023] A standardized decision matrix is ​​obtained by standardizing the attribute data using a nonlinear value function, and positive and negative ideal solution vectors are determined. The weighted Euclidean distance from each task to the positive and negative ideal solution vectors is calculated. Based on the weighted Euclidean distance, the proximity value is calculated, and a priority sorting sequence is generated by arranging the proximity values ​​in descending order.

[0024] Furthermore, the attribute evaluation set includes equipment importance attribute, equipment repair time attribute, equipment resource demand attribute, and equipment location attribute; the equipment importance attribute is a benefit-type attribute with a value range of 0 to 1; the equipment repair time attribute, the equipment resource demand attribute, and the equipment location attribute are all cost-type attributes.

[0025] The scenario discrimination parameter is the average remaining time to complete the task, and the scenario threshold ranges from 24h to 72h.

[0026] Furthermore, the correlation matrix is ​​a symmetric matrix, with its elements ranging from 0 to 1, and the diagonal elements having a value of 1. When there is a strong correlation between two attributes, the corresponding element value ranges from 0.6 to 0.9; when there is a moderate correlation between two attributes, the corresponding element value ranges from 0.3 to 0.6; and when there is a weak correlation between two attributes, the corresponding element value ranges from 0 to 0.3.

[0027] Furthermore, the expected value of the hesitant fuzzy element is equal to the arithmetic mean of the evaluation scores of each decision-maker; the variance of the hesitant fuzzy element is equal to the arithmetic mean of the squares of the differences between the evaluation scores of each decision-maker and the expected value.

[0028] The weight optimization objective function includes three weighted components. The first component is calculated based on the ratio of the expected value to the variance value of the hesitant fuzzy element. The second component is calculated based on the sum of the Euclidean distances of each task on the attribute. The third component is calculated based on the sum of the complements of the correlation between the attribute and other attributes. The sum of the balance coefficients of the three components is 1.

[0029] Furthermore, the nonlinear value function is a Logistic function; for positive indicators, the standardized value is equal to the Logistic function value; for negative indicators, the standardized value is equal to 1 minus the Logistic function value; the parameters of the Logistic function include the indicator's intermediate threshold and the indicator's dispersion parameter.

[0030] The hierarchical model includes a target layer, a criterion layer, and an indicator layer. The criterion layer includes the urgency criterion, equipment importance criterion, maintenance resource demand criterion, and fault severity criterion. The indicator layer includes the remaining completion time indicator, the urgency level indicator, the strategic value of the equipment indicator, the scope of the fault impact indicator, the maintenance man-hour indicator, the spare parts preparation time indicator, the fault deterioration risk indicator, and the economic loss estimation indicator.

[0031] Furthermore, the consistency ratio is equal to the ratio of the consistency index to the random consistency index; the consistency index is equal to the difference between the largest eigenvalue of the judgment matrix and the number of indicators divided by the number of indicators minus 1; when the consistency ratio is less than 0.1, the judgment matrix is ​​deemed to be of good consistency.

[0032] Another object of the present invention is to provide an equipment maintenance task priority evaluation system comprising:

[0033] The data acquisition module is used to acquire maintenance task datasets and attribute evaluation sets, and determine the evaluation path based on scenario discrimination parameters;

[0034] The scene discrimination module is used to compare the scene discrimination parameters with the preset scene thresholds. When the scene discrimination parameters are less than the scene thresholds, the hesitant fuzzy evaluation path is triggered. When the scene discrimination parameters are greater than or equal to the scene thresholds, the hierarchical analysis-approximation ideal solution evaluation path is triggered.

[0035] The hesitant fuzzy modeling module is used to construct a hesitant fuzzy matrix based on evaluation data from multiple decision-makers and to calculate the expected value and variance value of each hesitant fuzzy element.

[0036] The dynamic weight calculation module is used to construct the correlation matrix and build a weight optimization objective function based on hesitant fuzzy element features, attribute dispersion, and attribute correlation. The attribute weight vector is obtained by solving the Lagrange multiplier method.

[0037] The comprehensive evaluation and ranking module is used to normalize the attribute values ​​to obtain a normalized attribute matrix, calculate the comprehensive evaluation score based on the normalized attribute matrix and the attribute weight vector, and generate a priority ranking sequence by arranging the comprehensive evaluation scores in descending order.

[0038] The hierarchical analysis weight module is used to construct a hierarchical structure model and a judgment matrix, calculate the weights of the criterion layer and the index layer, and verify the effectiveness of the judgment matrix through the consistency ratio value.

[0039] The approximation ideal solution sorting module is used to standardize attribute data using a nonlinear value function to obtain a standardized decision matrix, determine the positive ideal solution vector and the negative ideal solution vector, calculate the weighted Euclidean distance value and the proximity value, and generate a priority sorting sequence by arranging them in descending order according to the proximity value.

[0040] Another object of the present invention is to provide a computer device including a memory and a processor, the memory storing a computer program, which, when executed by the processor, causes the processor to perform the steps of the equipment maintenance task priority evaluation method.

[0041] Another object of the present invention is to provide a computer-readable storage medium storing a computer program, which, when executed by a processor, causes the processor to perform the steps of the equipment maintenance task priority evaluation method.

[0042] Another objective of this invention is to provide an information data processing terminal for implementing the equipment maintenance task priority evaluation system.

[0043] Based on the above technical solutions and the technical problems solved, please analyze the advantages and positive effects of the technical solution to be protected by this invention from the following aspects:

[0044] First, addressing the technical problems existing in the prior art and the difficulty in solving them, this paper closely analyzes, in conjunction with the technical solution to be protected by this invention and the results and data obtained during the research and development process, how the technical solution of this invention solves the technical problems, and the inventive technical effects brought about by solving these problems. The specific description is as follows:

[0045] By constructing a hesitant fuzzy matrix to characterize the fuzzy attitudes of multiple decision-makers, the problem of weight bias caused by the forced single assignment in traditional methods is effectively solved. By introducing an attribute correlation matrix and constructing a weight optimization objective function, dynamic adjustment of attribute weights is realized, improving the responsiveness of evaluation results to task changes. Through a scenario-adaptive dual-path evaluation mechanism, differentiated evaluation strategies for wartime and daily scenarios are realized, balancing the speed and accuracy of evaluation. Data standardization using a Logistic nonlinear value function eliminates the impact of dimensional differences on evaluation results, improving the scientificity and reliability of the ranking.

[0046] Secondly, as supplementary evidence of the inventive step of the claims of this invention, it is also reflected in the following important aspects:

[0047] By achieving dynamic and precise prioritization of maintenance tasks, the efficiency of equipment maintenance resource allocation can be improved by over 30%, the average equipment repair time shortened by 25%, and the continuous combat capability of combat units significantly enhanced. In the commercial application field, this technology can be widely adapted to equipment maintenance management in defense, aerospace, and high-end manufacturing industries, with estimated annual economic benefits reaching tens of millions of yuan and broad market prospects.

[0048] This paper proposes for the first time a maintenance priority evaluation framework that integrates hesitant fuzzy decision-making and dynamic weight adjustment, breaking through the technical bottleneck of traditional static evaluation methods.

[0049] This invention successfully solves long-standing technical challenges in maintenance decision-making, such as quantifying decision-makers' hesitation, real-time response to dynamic tasks, and multi-attribute correlation modeling, achieving a leap from static single evaluation to dynamic comprehensive decision-making.

[0050] By constructing an attribute correlation matrix and a dynamic weight optimization model, we have confirmed the crucial role of considering attribute correlation and scenario adaptability in improving evaluation accuracy. Attached Figure Description

[0051] Figure 1 This is a flowchart of the equipment maintenance task priority evaluation method provided in the embodiments of the present invention.

[0052] Figure 2 This is a structural block diagram of the equipment maintenance task priority evaluation system provided in the embodiments of the present invention. Detailed Implementation

[0053] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0054] Reference Figure 1 As shown, the equipment maintenance task priority evaluation method provided by this invention includes data acquisition steps, scenario discrimination and path selection, hesitant fuzzy evaluation path, and hierarchical analysis-approximation ideal solution evaluation path. The core innovation of this invention lies in constructing a scenario-adaptive dual-path evaluation mechanism: when the task urgency is high, a hesitant fuzzy evaluation path is used to achieve rapid response; when the task constraints are complex, a hierarchical analysis-approximation ideal solution evaluation path is used to achieve refined ranking.

[0055] Step S1: Data acquisition step.

[0056] The data acquisition step is used to obtain the maintenance task dataset and attribute evaluation set, and to determine the evaluation path based on the scenario discrimination parameters.

[0057] Preferably, in one embodiment of the present invention, the maintenance task dataset Include One equipment maintenance task to be evaluated, among which This represents the total number of maintenance tasks, and its value is a positive integer. Each maintenance task For one or more pieces of equipment awaiting repair, the task data includes basic information such as equipment number, fault type, and fault occurrence time. In a preferred embodiment of the present invention, The value ranges from 2 to 100, and more preferably from 5 to 50, in order to balance computational efficiency and evaluation comprehensiveness.

[0058] Attribute Evaluation Set Include There are several evaluation attributes, among which This indicates the number of evaluation attributes, and its value ranges from positive integers. In response to the characteristics of equipment maintenance tasks, this invention constructs an evaluation system comprising four core attributes, covering four dimensions: operational value, time constraints, resource constraints, and spatial constraints. Specifically, the first attribute... Equipment importance, used to reflect the degree to which equipment contributes to combat missions, needs to be quantified by combining the importance of equipment attributes and the importance of the system architecture. Its quantification range is... The larger the value in the range, the more important the equipment, and the higher its priority should be. Second attribute. The repair time for equipment awaiting repair includes two parts: pre-repair preparation time and repair time. Pre-repair preparation time is affected by terrain and environmental factors, while repair time is determined by the characteristics of the malfunction and repair capabilities. The unit is hours (h). A lower value indicates a faster recovery of combat effectiveness, and the higher the priority should be. (Third attribute) Resource demand for equipment under repair is defined as the ratio of resource demand to total resources. It quantifies the intensity of resource consumption during the repair process; a smaller value indicates less resource consumption and a higher priority. (Fourth attribute) The location of the equipment to be repaired is represented by the equivalent distance, which is determined by the product of the difficulty coefficient and the physical distance, and the unit is meters (m). The smaller the value, the shorter the repair maneuver time, and the higher the priority should be.

[0059] Scene discrimination parameters To characterize the urgency of the current task environment, the scenario discrimination parameter in this invention is set as the average remaining task completion time, in hours (h). When the scenario discrimination parameter... Less than the preset scene threshold When the system indicates that it is currently in a wartime dynamic scenario, it automatically selects a hesitant fuzzy evaluation path; when the scenario discrimination parameters... Greater than or equal to the preset scene threshold When this occurs, it indicates that the system is currently in a routine multi-constraint scenario, and the system automatically selects the analytic hierarchy process (AHP) - approximation of the ideal solution evaluation path. Preferably, the scenario threshold... The value ranges from 24h to 72h, with 48h being more preferred. The basis for setting this threshold is that when the average remaining completion time of the task is less than 48h, the decision time is tight, and a hesitant fuzzy evaluation path with a faster response speed is required; when the average remaining completion time of the task is greater than or equal to 48h, the decision time is relatively sufficient, and a more accurate hierarchical analysis-approximation ideal solution evaluation path can be used.

[0060] In a preferred embodiment of the present invention, the data acquisition step further includes a data preprocessing sub-step for performing integrity checks and outlier removal on the raw data. Specifically, when the missing rate of attribute data for a certain maintenance task exceeds a preset missing threshold (preferably 20%), the task data is marked as invalid data and removed; when an attribute value exceeds a preset reasonable range, the three-standard-deviation rule is used for outlier identification and correction.

[0061] Step S2: Hesitation fuzzy modeling steps.

[0062] When the scene discrimination parameter is less than the preset scene threshold, the hesitant fuzzy evaluation path is executed. The hesitant fuzzy modeling step is used to construct a hesitant fuzzy matrix based on the evaluation data of multiple decision-makers, and to calculate the expected value and variance value of each hesitant fuzzy element.

[0063] This invention employs hesitant fuzzy set theory to characterize the fuzzy attitudes of decision-makers. (The following is a partial sentence fragment: "Assuming...") Equipment maintenance tasks and Each evaluation attribute, invitation An evaluation was conducted by several experts with extensive repair experience, among whom... This represents the number of decision-making experts, and its value is a positive integer. Ideally, the number of participants should be 3 to 5. Hesitant fuzzy matrix. for The first order matrix, where the second order matrix is ​​the first order matrix. Line number Column elements Let hesitant fuzzy elements represent the task. Satisfying attributes The degree of hesitation and ambiguity. Multiple values ​​can be taken, represented as ,in For the first Each decision-maker on the task Satisfying attributes The evaluation score ranges from 100 to 100. Interval.

[0064] Hesitant fuzzy matrix The expression is:

[0065] ,

[0066] in: For hesitant fuzzy matrices; For the position located in the matrix Line number The hesitant and ambiguous elements of the column; The number of repair tasks, with a value that is a positive integer; The value is a positive integer to evaluate the number of attributes.

[0067] For each hesitant fuzzy element Calculate its expected value With variance value The expected value is used to characterize the central tendency of the evaluation results from multiple decision-makers, while the variance is used to characterize the dispersion of the evaluation results from multiple decision-makers. The formula for calculating the expected value is:

[0068] ,

[0069] in: For hesitant fuzzy elements The expected value, with a range of values ​​being ; The number of decision-making experts, whose value is a positive integer and ; For the first Each decision-maker on the task Satisfying attributes The evaluation score ranges from 100 to 100. ; The decision-maker's serial number, with a value range of [value range missing]. .

[0070] The formula for calculating the variance is:

[0071] ,

[0072] in: For hesitant fuzzy elements The variance value, with a range of values ​​of . ; For the hesitant fuzzy element The expected value is given; the meanings of the other symbols are the same as described above. The smaller the variance value, the more consistent the evaluation opinions of multiple decision-makers; the larger the variance value, the greater the disagreement among decision-makers.

[0073] In a preferred embodiment of the present invention, when the variance value of a certain hesitant fuzzy element exceeds a preset variance threshold (preferably 0.15), it indicates that there is a large disagreement in the expert evaluation of the attribute. At this time, the Delphi method can be used to reconcile opinions, or the number of experts can be increased to improve the reliability of the evaluation.

[0074] In this invention, the joint modeling of the expected and variance values ​​of hesitant fuzzy elements is a key technical means to accurately characterize decision fuzziness. Compared with traditional methods that only consider the evaluation mean, this invention also introduces variance information, which can more comprehensively reflect the consistency of expert evaluations and provide a richer information basis for subsequent weight optimization.

[0075] Step S3: Dynamic weight calculation step.

[0076] The dynamic weight calculation step is used to construct the correlation matrix and construct the weight optimization objective function based on the hesitant fuzzy element features, attribute dispersion, and attribute correlation. The attribute weight vector is obtained by solving the Lagrange multiplier method.

[0077] First, construct the correlation matrix. Relationship matrix for A symmetric matrix of order n, where the nth order n is the nth order n. Line number Column elements Represents attributes With attributes The degree of correlation between them, with a range of values. .when hour, ;when hour, The correlation matrix is ​​evaluated by domain experts based on the interrelationships between attributes. Preferably, the elements of the correlation matrix are selected according to the following principle: if there is a strong correlation between two attributes, then... The range of values ​​is If there is a moderate correlation between two attributes, then The range of values ​​is If a weak association exists between two attributes, then The range of values ​​is .

[0078] In a preferred embodiment of the present invention, the typical values ​​of the correlation matrix for the four core attributes of equipment maintenance tasks are as follows: the correlation between equipment importance and repair time is 0.30, indicating a moderately weak correlation between the two; the correlation between equipment importance and resource demand is 0.25, indicating a weak correlation between the two; the correlation between equipment importance and location is 0.40, indicating a moderate correlation between the two; the correlation between repair time and resource demand is 0.50, indicating a moderate correlation between the two; the correlation between repair time and location is 0.60, indicating a strong correlation between the two; and the correlation between resource demand and location is 0.35, indicating a moderately weak correlation between the two.

[0079] Based on hesitant fuzzy meta-features, attribute dispersion, and attribute correlation, a weight optimization objective function is constructed. The core idea of ​​weight optimization is: attributes with larger expected values ​​should receive higher weights; attributes with smaller variance should receive higher weights; attributes with greater dispersion should receive higher weights; and attributes with lower correlation to other attributes should receive higher weights. The weight optimization objective function is:

[0080] ,

[0081] ,

[0082] in: Optimize the objective function for the weights; This is the attribute weight vector; For the first The weights of each attribute, with values ​​ranging from 1 to 2. ; For the balance coefficient, satisfying Preferably ; For the hesitant fuzzy element The expected value; For the hesitant fuzzy element The variance value; For the task With the task In attributes The Euclidean distance value on; For attributes With attributes The degree of correlation between them; The number of maintenance tasks; To evaluate the number of attributes.

[0083] Task With the task In attributes Euclidean distance value on The calculation formula is:

[0084] ,

[0085] in: For the task With the task In attributes The Euclidean distance on the surface is a value that takes the range of 1 / 2. ; For the first Each decision-maker on the task Satisfying attributes Evaluation score; For the first Each decision-maker on the task Satisfying attributes Evaluation score; The number of decision-making experts.

[0086] To solve the above constrained optimization problem, a Lagrange auxiliary function is constructed:

[0087] ,

[0088] in: It is a Lagrange auxiliary function; It is a Lagrange multiplier.

[0089] right and Calculate the partial derivatives separately and set them equal to zero, then solve the simultaneous equations to obtain the initial weights. Next, normalize the initial weights to ensure the sum of the weights is 1, finally obtaining the attribute weight vector. .

[0090] In this invention, the attribute-relatedness-driven dynamic weight optimization model is the core technology for achieving real-time weight adjustment. When the task set changes, the dispersion of each attribute changes accordingly, leading to a redistribution of optimal weights. This dynamic adjustment mechanism enables this invention to effectively respond to real-time changes in tasks and improve the adaptability of evaluation results.

[0091] Step S4: Comprehensive evaluation and ranking steps.

[0092] The comprehensive evaluation and ranking step is used to normalize the attribute values ​​to obtain a normalized attribute matrix, calculate the comprehensive evaluation score based on the normalized attribute matrix and the attribute weight vector, and generate a priority ranking sequence by arranging the comprehensive evaluation scores in descending order.

[0093] First, the attribute values ​​are normalized to eliminate dimensional differences. This invention employs different normalization formulas depending on the attribute type: for benefit-type attributes (higher values, higher priority), a forward normalization formula is used; for cost-type attributes (lower values, higher priority), a reverse normalization formula is used.

[0094] The normalization formula for benefit-type attributes is:

[0095] ,

[0096] in: For the task In attributes The normalized attribute value on the above has a range of values. ; For the task In attributes The original attribute value on; For attributes The maximum value of the original value.

[0097] The normalization formula for cost-type attributes is:

[0098] ,

[0099] in: For the task In attributes The normalized attribute value on the above has a range of values. ; For the task In attributes The original attribute value on; For attributes The minimum value of the original value.

[0100] In the four-attribute evaluation system of this invention, equipment importance As a benefit-oriented attribute, repair time Resource demand ,Location All are cost-related attributes.

[0101] After normalization, the normalized attribute matrix is ​​obtained. Based on the normalized attribute matrix and attribute weight vector, a weighted summation method is used to calculate the comprehensive evaluation score:

[0102] ,

[0103] in: For the task The overall evaluation score, with a value range of 100%. ; For the task In attributes Normalized attribute values; For attributes The optimal weight; To evaluate the number of attributes.

[0104] Based on overall evaluation score Sort in descending order to obtain the priority ranking sequence. The higher the overall evaluation score, the higher the priority of the maintenance task, and it should be executed first.

[0105] Preferably, the present invention also supports dynamic scheduling. When a new task is added or an existing task is completed, the system automatically recalculates the comprehensive evaluation score of each task and updates the priority ranking sequence. Because the weight optimization model can respond to changes in the task set, dynamic scheduling will not cause inconsistencies in the ranking.

[0106] Step S5: Hierarchical Analysis Weight Calculation Steps

[0107] When the scene discrimination parameter is greater than or equal to the preset scene threshold, the analytic hierarchy process (AHP) – approximation of the ideal solution evaluation path is executed. The AHP weight calculation step is used to construct the hierarchical structure model and judgment matrix, calculate the weights of the criterion layer and the index layer, and verify the effectiveness of the judgment matrix through the consistency ratio value.

[0108] First, a hierarchical structure model is constructed. This invention decomposes the problem of prioritizing equipment maintenance and support tasks into a three-level progressive structure: the target layer is the priority of equipment maintenance and support tasks (A); the criterion layer includes four criteria: task urgency (B1), equipment importance (B2), maintenance resource requirements (B3), and fault severity (B4); the indicator layer includes eight secondary indicators: remaining completion time (C11), task urgency level (C12), equipment strategic value (C21), fault impact range (C22), maintenance man-hours (C31), spare parts preparation time (C32), fault escalation risk (C41), and economic loss estimation (C42).

[0109] Construct the judgment matrix of the criterion layer to the target layer ,in The number of criteria layer indicators. Elements of the judgment matrix. Representation Criteria Relative to criteria The importance of [the metric] is quantified using a 1-9 scale. Specifically, when [the metric is determined by the number of metric points]... When, it indicates the criterion with guidelines Equally important; when When, it indicates the criterion Ratio Criterion Slightly important; when When, it indicates the criterion Ratio Criterion Clearly important; when When, it indicates the criterion Ratio Criterion Strongly important; when When, it indicates the criterion Ratio Criterion Extremely important; 2, 4, 6, and 8 are the intermediate values ​​of the above adjacent scales. The judgment matrix must satisfy reciprocity, i.e. And the diagonal element is 1, that is .

[0110] The weights of the criterion layer are calculated using the summation method. First, calculate the sum of the elements in each row of the judgment matrix. Then, the sum of each row is divided by the total sum of the judgment matrix to obtain the normalized criterion layer weights:

[0111] ,

[0112] in: As a standard The weight, with a value range of . And satisfy ; To determine the matrix of the first The sum of the elements in the row; The number of indicators at the criteria layer.

[0113] To eliminate logical contradictions in expert subjective judgments, a consistency check is required. First, the largest eigenvalue of the judgment matrix is ​​calculated. :

[0114] ,

[0115] in: To determine the largest eigenvalue of a matrix, the range of values ​​is: ; For the judgment matrix; The criterion layer weight vector; For matrix with vector The product of the first Each element.

[0116] Then calculate the consistency index. :

[0117] ,

[0118] in: This is a consistency indicator, with a value range of [value range missing]. ; The largest eigenvalue of the judgment matrix; The number of indicators at the criteria layer. The smaller the value, the fewer logical contradictions there are in the subjective judgment.

[0119] Finally, calculate the consistency ratio. :

[0120] ,

[0121] in: This represents the consistency ratio. The consistency index is mentioned above; As a random consistency indicator, it needs to be obtained by looking up a table based on the number of indicators. hour ,when hour ,when hour ,when hour ,when hour ,when hour .

[0122] When the consistency ratio value When the consistency of the judgment matrix is ​​qualified, the weight calculation result is valid; when When this happens, the judgment matrix needs to be readjusted until the consistency requirement is met.

[0123] The same method is used to construct the judgment matrix of the indicator layer to the criterion layer and calculate the weights of the indicator layer. Finally, the total weight of each secondary indicator is obtained by multiplying the weights of the criterion layer and the weights of the indicator layer.

[0124] Step S6: Sort steps to approximate the ideal solution.

[0125] The approximation of ideal solution sorting step is used to standardize attribute data using a nonlinear value function to obtain a standardized decision matrix, determine the positive and negative ideal solution vectors, calculate the weighted Euclidean distance from each task to the positive and negative ideal solution vectors, calculate the proximity value based on the weighted Euclidean distance value, and generate a priority sorting sequence by arranging the proximity values ​​in descending order.

[0126] First, the attribute data is standardized using the Logistic nonlinear value function. Compared to the traditional linear normalization method, the Logistic function can better handle uneven distribution of attribute values, making the standardization results more reasonable.

[0127] For positive indicators (higher values ​​have higher priority), the Logistic standardization formula is:

[0128] ,

[0129] in: For the task In indicators The standardized attribute values ​​on the surface have a range of values. ; For the task In indicators The original attribute value on; As an indicator The median threshold is usually taken as the median of the indicator value or the median of the industry standard. As an indicator The dispersion parameter is usually taken as 1 / 3 of the half-width of the index value range.

[0130] For negative indicators (the smaller the value, the higher the priority), the Logistic standardization formula is:

[0131] ,

[0132] The meanings of the symbols are the same as those of the positive indicators.

[0133] In the 8-index evaluation system of this invention, the remaining completion time C11, maintenance man-hours C31, and spare parts preparation time C32 are negative indicators, while the mission urgency level C12, equipment strategic value C21, fault impact range C22, fault deterioration risk C41, and economic loss estimate C42 are positive indicators.

[0134] After standardization, the standardized decision matrix is ​​obtained. ,in This refers to the number of secondary indicators.

[0135] Determine the positive ideal solution vector With negative ideal solution vector The positive ideal solution vector represents a virtual task where all metrics reach their theoretical optimal values, while the negative ideal solution vector represents a virtual task where all metrics reach their theoretical worst values. For positive metrics, the positive ideal solution takes the maximum value of all standardized task values, and the negative ideal solution takes the minimum value; for negative metrics, the positive ideal solution takes the minimum value of all standardized task values, and the negative ideal solution takes the maximum value.

[0136] ,

[0137] ,

[0138] in: The first positive ideal solution vector One component; The first of the negative ideal solution vectors One component; For the task In indicators Standardized attribute values.

[0139] Calculate the weighted Euclidean distance from each task to the positive ideal solution vector. The weighted Euclidean distance to the negative ideal solution vector :

[0140] ,

[0141] ,

[0142] in: For the task The weighted Euclidean distance to the positive ideal solution vector, with values ​​ranging from 1 to 2. ; For the task The weighted Euclidean distance to the negative ideal solution vector, with values ​​ranging from 1 to 2. ; As an indicator The total weight; For the task In indicators Standardized attribute values; The first positive ideal solution vector is the... One component; The first of the negative ideal solution vectors One component; This refers to the number of secondary indicators.

[0143] Closeness values ​​are calculated based on weighted Euclidean distance. :

[0144] ,

[0145] in: For the task The proximity value, with a range of values ​​being: ; For the task The weighted Euclidean distance to the positive ideal solution vector; For the task The weighted Euclidean distance to the negative ideal solution vector. The closer to 1, the better the task. The closer a solution is to a positive ideal solution and the further it is from a negative ideal solution, the higher its priority. The closer to 0, the better the task. The closer a solution is to the negative ideal solution and the further it is from the positive ideal solution, the lower its priority.

[0146] According to proximity value Arrange in descending order to obtain the priority sorting sequence. The hierarchical analysis-approximation ideal solution evaluation path of this invention also supports dynamic scheduling: when a new task is added, it is only necessary to standardize the new task, calculate its distance and proximity to the positive and negative ideal solutions, and quickly determine its position in the existing sorting sequence without reconstructing the judgment matrix and calculating the weights.

[0147] To verify the effectiveness of the method of the present invention, an equipment maintenance task during a live-fire exercise of a combined arms brigade in a high-altitude region is used as an example. In this scenario, a total of 8 main combat equipment experienced sudden malfunctions, resulting in 8 maintenance tasks to be handled. to The average remaining task completion time is 36 hours, which is less than the scenario threshold of 48 hours. Therefore, the system automatically selects the hesitant fuzzy evaluation path. This scenario has typical wartime characteristics: time is tight, information is incomplete, and decision-makers are under great pressure, making it very suitable for verifying the practical effect of the hesitant fuzzy evaluation path of this invention.

[0148] First, data collection is performed to obtain the original attribute values ​​for each maintenance task. Task The equipment importance is 0.56, repair time is 0.7 hours, resource requirement is 0.08, and location is 800 meters. (Mission details omitted) The equipment importance is 0.37, the repair time is 1.5 hours, the resource requirement is 0.15, and the location is 900 meters. (Mission details omitted) The equipment importance is 0.35, repair time is 0.8 hours, resource requirement is 0.24, and location is 900 meters. (Mission details omitted) The equipment importance is 0.52, repair time is 0.5 hours, resource requirement is 0.53, and location is 1200 meters. (Task) The equipment importance is 0.36, the repair time is 2.0 hours, the resource requirement is 0.18, and the location is 1400 meters. (Mission details omitted) The equipment importance is 0.32, the repair time is 1.2 hours, the resource requirement is 0.39, and the location is 1400 meters. (Mission details omitted) The equipment importance is 0.50, repair time is 0.4 hours, resource requirement is 0.05, and location is 300 meters. (Task) The equipment importance is 0.38, the repair time is 1.2 hours, the resource requirement is 0.36, and the location is 200 meters.

[0149] Three senior experts with over 10 years of experience in armored equipment maintenance were invited to conduct hesitant and fuzzy assessments of each task. The expert assessments used a 1-0 range rating method, with each expert independently assigning a score. (Based on the task...) In terms of equipment importance Taking the above assessment as an example, the three experts gave evaluation scores of 0.80, 0.85, and 0.95 respectively, forming a hesitant fuzzy element. This hesitant ambiguity reflects the expert's understanding of the task. The comprehensive assessment of equipment importance reveals three numerical values, indicating some disagreement among experts, but a general consensus. These values ​​are calculated using the expected value formula. Calculated according to the variance formula The relatively small variance of 0.003 indicates that the experts' opinions are largely consistent, and the assessment result is highly reliable.

[0150] The expected value and variance of all hesitant fuzzy elements were calculated using the same method. (Task...) For example, its importance in equipment The hesitant fuzzy element is The expected value is 0.783, and the variance is 0.0039; during the repair time... The hesitant fuzzy element is The expected value is 0.867, and the variance is 0.0039; in terms of resource demand... The hesitant fuzzy element is The expected value is 0.900, and the variance is 0.0017; at position The hesitant fuzzy element is The expected value is 0.717, and the variance is 0.0039. These values ​​indicate that the task... It performs exceptionally well in terms of repair time and resource requirements, which is consistent with its original attribute values ​​of only 0.4 hours for repair and only 0.05 for resource requirements.

[0151] Next, we construct the attribute association matrix. Based on the experience of experts in the field of equipment maintenance, the importance of equipment is... With repair time A correlation coefficient of 0.30 indicates a moderately weak correlation between the two, meaning that more important equipment does not necessarily require a longer repair time; equipment importance With resource demand A correlation score of 0.25 indicates a weak correlation between the two; equipment importance With position A correlation coefficient of 0.40 indicates a moderate correlation between the two, suggesting that important equipment is typically deployed in critical locations; repair time With resource demand A correlation score of 0.50 indicates a moderate correlation between the two tasks; tasks with longer repair times often consume more resources. With position A correlation coefficient of 0.60 indicates a strong correlation between the two; equipment located further away has a longer maneuver time, indirectly extending the total repair time; resource demand... With position The correlation coefficient is 0.35, indicating that there is a moderate to weak correlation between the two.

[0152] Set balance coefficient The optimal attribute weight vector is obtained by solving the correlation matrix and weight optimization model. The weighting results show that: location attribute has the highest weight (0.271), reflecting the key impact of mobility on maintenance response in high-altitude environments. The complex terrain and poor road conditions in high-altitude areas significantly increase the difficulty of mobility; repair time and resource requirement have similar weights (0.247 and 0.245), reflecting a balanced consideration of time and resource constraints; equipment importance has a slightly lower weight (0.237), which is consistent with the actual need in wartime scenarios to prioritize restoring combat effectiveness rather than emphasizing the value of individual equipment. When multiple pieces of equipment fail simultaneously, priority should be given to repairing those that can be quickly restored and consume fewer resources.

[0153] Normalize the original attribute values. (Based on the task) For example, its equipment importance normalized value is (because The equipment importance score of 0.56 is the highest among all missions; the normalized repair time value is... (because The minimum repair time is 0.4 hours; the normalized resource requirement value is... (because The resource demand degree is 0.05 (minimum value); the location normalization value is... (because The minimum value is 200m from the location.

[0154] Calculate the comprehensive evaluation score using the weighted summation method. (Task) Overall evaluation score The overall evaluation score for all tasks was calculated using the same method: , , , , , , , The final priority-sorted sequence is obtained as follows: .

[0155] The sorting results show that the task It ranked first with the highest overall score of 0.885. This is because the mission has significant advantages such as short repair time (0.4h), low resource requirement (0.05), and close location (300m). Although the equipment importance is not the highest (0.50), its overall performance is the best. It ranks second, thanks to its highest equipment importance (0.56). Mission It ranks last, mainly because of its remote location (1400m) and high resource demand (0.39).

[0156] To verify the dynamic adjustment capability, the task set was reduced to the first 6 tasks. to And recalculate the weights. Due to the task and When an attribute is removed, the dispersion of its values ​​changes, and the input parameters of the weight optimization model change accordingly. A new weight vector is then obtained by resolving the problem. Comparing the two sets of weights, we can see that the equipment importance weight increased from 0.237 to 0.258, an increase of 8.9%; while the location weight decreased from 0.271 to 0.249, a decrease of 8.1%. This change is reasonable: the optimal location was removed. (300m) and After 200m, the differences in the location attributes of the remaining tasks decrease, and their distinguishability decreases, so their weights are adjusted downwards accordingly. At the same time, the differences in equipment importance among the remaining tasks become more obvious, and their distinguishing effect is enhanced, so their weights are adjusted upwards accordingly. This result fully verifies the effectiveness of the dynamic weight adjustment mechanism of this invention, which can automatically redistribute weights according to changes in the task set, making the evaluation results more adaptable to the actual situation.

[0157] The comprehensive evaluation scores and priority rankings of the six tasks were recalculated based on the new weight vectors. The results show that, under the new weight configuration, the tasks... Its overall score rose to 0.928, ranking first; Task Score 0.694, ranked second; Task With a score of 0.620, it ranks third. The new ranking result is different from the original ranking (excluding...). , The differences between the two results (the results of which are not explicitly stated in the original text) reflect the substantial impact of weight changes on the final ranking. This further demonstrates the necessity and effectiveness of the present invention in considering dynamic changes in tasks to adjust weights.

[0158] Reference Figure 2 As shown, the present invention also provides an equipment maintenance task priority evaluation system, including a data acquisition module 1, a scenario discrimination module 2, a hesitant fuzzy modeling module 3, a dynamic weight calculation module 4, a comprehensive evaluation ranking module 5, a hierarchical analysis weight module 6, and an approximation ideal solution ranking module 7.

[0159] Data acquisition module 1 is used to acquire maintenance task datasets and attribute evaluation sets, and transmit the data to scenario discrimination module 2. Specifically, data acquisition module 1 includes a task information acquisition unit, an attribute data acquisition unit, and an expert evaluation acquisition unit. The task information acquisition unit is used to acquire basic information such as equipment number, fault type, and fault occurrence time for each maintenance task; the attribute data acquisition unit is used to acquire the raw values ​​of each maintenance task on each evaluation attribute; and the expert evaluation acquisition unit is used to acquire hesitant fuzzy evaluation data from multiple decision-making experts. Data acquisition module 1 also includes a data preprocessing unit for performing integrity checks and outlier removal on the raw data.

[0160] Scene discrimination module 2 compares scene discrimination parameters with preset scene thresholds. When the scene discrimination parameters are less than the scene threshold, a hesitant fuzzy evaluation path is triggered, and the data is transmitted to the hesitant fuzzy modeling module 3. When the scene discrimination parameters are greater than or equal to the scene threshold, a hierarchical analysis-approximation ideal solution evaluation path is triggered, and the data is transmitted to the hierarchical analysis weight module 6. Scene discrimination module 2 implements adaptive selection of evaluation paths, and can automatically switch evaluation strategies according to the urgency of the task environment.

[0161] The hesitant fuzzy modeling module 3 is used to construct a hesitant fuzzy matrix based on the evaluation data of multiple decision-makers, calculate the expected value and variance value of each hesitant fuzzy element, and transmit the calculation results to the dynamic weight calculation module 4. The specific implementation of the hesitant fuzzy modeling module 3 corresponds to the content of step S2 in the aforementioned method embodiment.

[0162] The dynamic weight calculation module 4 is used to construct the correlation matrix and, based on the hesitant fuzzy element features, attribute dispersion, and attribute correlation, construct the weight optimization objective function. The attribute weight vector is obtained by solving the Lagrange multiplier method and then transmitted to the comprehensive evaluation and ranking module 5. The specific implementation of the dynamic weight calculation module 4 corresponds to step S3 in the aforementioned method embodiment.

[0163] The comprehensive evaluation ranking module 5 is used to normalize the attribute values ​​to obtain a normalized attribute matrix, calculate the comprehensive evaluation score based on the normalized attribute matrix and the attribute weight vector, and generate a priority ranking sequence by arranging the comprehensive evaluation scores in descending order. The specific implementation of the comprehensive evaluation ranking module 5 corresponds to the content of step S4 in the aforementioned method embodiment.

[0164] The hierarchical analysis weight module 6 is used to construct the hierarchical structure model and judgment matrix, calculate the weights of the criterion layer and the index layer, verify the validity of the judgment matrix through the consistency ratio value, and transmit the weight calculation results to the approximation ideal solution sorting module 7. The specific implementation of the hierarchical analysis weight module 6 corresponds to the content of step S5 in the aforementioned method embodiment.

[0165] The approximation-ideal-solution sorting module 7 is used to standardize the attribute data using a nonlinear value function to obtain a standardized decision matrix, determine the positive and negative ideal solution vectors, calculate the weighted Euclidean distance and proximity values, and generate a priority sorting sequence by arranging the proximity values ​​in descending order. The specific implementation of the approximation-ideal-solution sorting module 7 corresponds to the content of step S6 in the aforementioned method embodiment.

[0166] Preferably, the equipment maintenance task priority evaluation system of the present invention further includes a result output module 8 and a dynamic update module 9. The result output module 8 is used to present the priority ranking sequence to the decision-maker in a visual manner, including various display formats such as a ranking list, a score bar chart, and a task attribute radar chart. The dynamic update module 9 is used to monitor changes in the task set, and automatically triggers a re-evaluation process and updates the priority ranking sequence when a new task is added or an existing task is completed.

[0167] The equipment maintenance task priority evaluation system of this invention can be deployed on various computing platforms such as cloud servers, local workstations, or embedded devices. In wartime scenarios, local deployment is preferred to ensure communication security and response speed; in routine training scenarios, cloud deployment can be used to facilitate data sharing and remote collaboration.

[0168] Taking the armored equipment maintenance and support scenario of a military industrial enterprise as an example, this paper verifies the application effect of the system of the present invention under the analytic hierarchy process (AHP)-approximation ideal solution evaluation path. The enterprise is responsible for the maintenance of 8 core armored equipment (including main battle tanks, armored personnel carriers, etc.). There are currently 10 tasks A1 to A10 to be executed, with an average remaining completion time of 72 hours, which is greater than or equal to the scenario threshold of 48 hours. Therefore, the system automatically selects the AHP-approximation ideal solution evaluation path. This scenario is a typical daily multi-constraint scenario with diverse task types, including emergency fault repair (A1, A4, A8), early warning repair (A2, A3, A6, A9), preventive repair (A5, A7), and technical upgrade (A10), requiring full consideration of resources, time, strategy, and other factors.

[0169] Data acquisition module 1 acquires the raw data for each task. Using the eight secondary indicators at the indicator layer as dimensions, the data for task A1 is as follows: remaining completion time 20 hours, task urgency level 4 (emergency), equipment strategic value level 5 (critical), failure impact range level 4 (affecting 6-7 units), maintenance man-hours 6 hours, spare parts preparation time 3 hours, failure deterioration risk level 5 (extremely high), and estimated economic loss level 5 (greater than 500,000 yuan). The data for task A8 is as follows: remaining completion time 18 hours, task urgency level 4, equipment strategic value level 5, failure impact range level 3 (affecting 4-5 units), maintenance man-hours 5 hours, spare parts preparation time 1 hour, failure deterioration risk level 5, and estimated economic loss level 5. The data for task A10 is as follows: remaining completion time 144 hours, task urgency level 1 (normal), equipment strategic value level 1 (auxiliary), failure impact range level 1 (single unit), maintenance man-hours 11 hours, spare parts preparation time 10 hours, failure deterioration risk level 1 (extremely low), and estimated economic loss level 1 (less than 50,000 yuan).

[0170] The weighting module 6 of the hierarchical analysis first constructs a hierarchical structure model. The target layer is the priority of equipment maintenance and support tasks (A); the criterion layer includes four criteria: task urgency (B1), equipment importance (B2), maintenance resource requirements (B3), and fault severity (B4); the indicator layer includes eight secondary indicators (C11 to C42). Subsequently, the judgment matrix of the criterion layer on the target layer is constructed and the weights are calculated.

[0171] Three military maintenance experts were invited to compare the importance of criteria layers pairwise using a 1-9 scale. The experts unanimously agreed that equipment importance B2 was the most critical, as the recovery of core equipment's combat effectiveness directly impacts mission execution; mission urgency B1 was second, as time constraints directly affect maintenance scheduling; fault severity B4 ranked third, as severe faults may worsen if not addressed promptly; and maintenance resource requirements B3 had the lowest weight, as resource constraints are not a primary concern in routine scenarios with relatively abundant resources. The average of the three experts' judgments was used to obtain the following judgment matrix: the first row contains elements 1, 1 / 3, 4, 2; the second row contains elements 3, 1, 7, 4; the third row contains elements 1 / 4, 1 / 7, 1, 1 / 3; and the fourth row contains elements 1 / 2, 1 / 4, 3, 1.

[0172] The summation method is used to calculate the weights of the criterion layer. First, the sum of the elements in each row is calculated: , , , The sum is 28.809. After normalization, the criterion layer weights are obtained: , , , The weight distribution shows that equipment importance has the highest weight (0.521), which is dominant; mission urgency has the second highest weight (0.254); failure severity has a medium weight (0.165); and maintenance resource demand has the lowest weight (0.060).

[0173] Perform a consistency check. Calculate the product of the judgment matrix and the weight vector: Calculate the largest eigenvalue: Calculate the consistency index: From the table, we can find... hour Calculate the consistency ratio: The consistency of the matrix is ​​deemed satisfactory, and the weight calculation results are valid.

[0174] The same method is used to construct the indicator-level judgment matrix and calculate the weights. Taking the indicators under criterion B2 (equipment importance) as an example, the judgment matrices for C21 (equipment strategic value) and C22 (failure impact scope) are as follows: Calculations yielded , By combining the weights of the criterion layer and the indicator layer, the total weight of each secondary indicator is obtained: Remaining Completion Time. Mission urgency level Strategic value of equipment Scope of Fault Impact Repair time Spare parts preparation time Risk of malfunction worsening Economic loss estimate The weight distribution shows that the strategic value of equipment has the highest weight (0.347), reflecting the importance attached to the protection of core equipment in daily scenarios; the impact range of the failure has the second highest weight (0.173), reflecting the serious consequences of systemic failures.

[0175] The sorting module 7, which approximates the ideal solution, first performs Logistic standardization. It sets intermediate thresholds for each indicator. and dispersion parameter Remaining completion time , Mission urgency level , Strategic value of equipment , ; Scope of the fault , Repair man-hours , Spare parts preparation time , Risk of worsening fault , Economic loss estimate , .

[0176] Calculate the standardized value using task A8 as an example. Remaining completion time is a negative indicator. The mission urgency level is a positive indicator. The strategic value of equipment is a positive indicator. The scope of the fault's impact is a positive indicator. Maintenance man-hours are a negative indicator. Spare parts preparation time is a negative indicator. The risk of worsening failure is a positive indicator. The economic loss forecast is a positive indicator. .

[0177] Determine the positive ideal solution vector and negative ideal solution vector The ideal solution vector represents the optimal value for all indices. The negative ideal solution vector represents the worst value of all indices: .

[0178] Calculate the weighted Euclidean distance and proximity to the positive and negative ideal solutions for each task. Distance from task A8 to the positive ideal solution. Distance to the negative ideal solution Proximity The similarity score for all tasks was calculated using the same method: , , , , , , , , , .

[0179] The final priority sorted sequence is: A8 (0.815). A1 (0.782) A4 (0.698) A6 (0.632) A2 (0.603) A3 (0.553) A7 (0.333) A9 (0.290) A5 (0.268) A10 (0.132). This ranking result conforms to the business logic of prioritizing emergency tasks over early warning tasks, early warning tasks over preventive tasks, and preventive tasks over technical upgrade tasks: the top three, A8, A1, and A4, are all emergency fault repair tasks; A6, A2, and A3 are early warning repair tasks, ranked 4th to 6th; A7 and A5 are preventive repair tasks, ranked 7th to 9th; and A10 is a technical upgrade task, ranked last. This result is highly consistent with the maintenance strategy of military enterprises to protect core components and reduce losses, verifying the effectiveness of the system of this invention.

[0180] To verify the dynamic scheduling function, a new task A11 was added, and its priority was quickly calculated. The original data for A11 was: remaining time 15 hours, emergency level 4 (emergency), strategic value level 5 (critical), impact range level 4 (affecting 6-7 units), maintenance man-hours 5 hours, spare parts time 2 hours, deterioration risk level 5 (extremely high), and economic loss level 5 (greater than 500,000 RMB). This task represents a newly discovered emergency fault, and its position in the existing task queue needs to be quickly determined.

[0181] After receiving new task data, the dynamic update module 9 triggers the approximation of ideal solution sorting module 7 for rapid calculation. First, Logistic standardization is performed to obtain the standardized vector. Since the positive and negative ideal solutions were determined in the initial calculation, there is no need to recalculate. Directly calculate the distance from A11 to the positive and negative ideal solutions: , Calculate the closeness: Comparing A11's proximity score of 0.784 with the existing task ranking, this value falls between A8 (0.815) and A1 (0.782). Therefore, A11's priority is determined to be second, ranking after A8 and before A1. The entire calculation process takes no more than 2 seconds, meeting the requirements for real-time response.

[0182] The updated priority sort sequence is: A8 (0.815) A11 (0.784) A1 (0.782) A4 (0.698) A6 (0.632) A2 (0.603) A3 (0.553) A7 (0.333) A9 (0.290) A5 (0.268) A10 (0.132). The results output module 8 presents the updated sorted sequence to the decision-maker in a visual manner, including a sorted list, a closeness bar chart, and a distance comparison chart with the ideal solution, so that the decision-maker can intuitively understand the priority status of each task and the reasons for it.

[0183] This invention relates to specific application areas or related products. The following are application examples of the technical solution of this invention in specific products and technical fields, intended to verify its inventiveness and technical value:

[0184] Application Example 1: Wartime Equipment Maintenance Decision Support System

[0185] This invention is applied to the battlefield maintenance and support system of a certain type of armored force. The data acquisition module acquires real-time tank damage data (such as armor damage level, fire control system status, and power system operating conditions) and combat mission parameters (combat priority, remaining mission time). The scenario discrimination module triggers a hesitant fuzzy evaluation path based on a threshold of "combat urgency > 0.8." Three maintenance experts input evaluation data (equipment importance, repair time, resource requirement) via portable terminals. The system automatically constructs a hesitant fuzzy matrix and calculates the expected value and variance. The dynamic weight calculation module analyzes the strong correlation between "repair time and resource requirement" (correlation coefficient 0.78) using a correlation matrix and obtains the dynamic weight vector using the Lagrange multiplier method. The comprehensive evaluation and ranking module completes the priority ranking of 20 damaged pieces of equipment within 3 minutes, improving the critical equipment repair response speed by 40% compared to traditional static ranking methods, and increasing the continuous combat capability of the armored assault group by 25% in actual combat exercises.

[0186] Application Example 2: Aircraft Engine Maintenance Priority Scheduling Platform

[0187] The system of this invention is deployed in the engine maintenance center of a civil airline, integrating real-time monitoring data from the Aircraft Health Management System (AHMS) (such as vibration spectrum, temperature field distribution, and fuel consumption deviation) and a maintenance resource database (spare parts inventory, technician qualifications, and tool status). When the scenario discrimination parameter (daily maintenance workload < 15 engines) triggers the hierarchical analysis-approximation ideal solution evaluation path, the hierarchical analysis weight module constructs a three-layer evaluation model of "safety-economy-task urgency," and calculates the weights of indicators such as "turbine blade damage degree" (weight 0.32), "maintenance cost" (weight 0.25), and "flight delay loss" (weight 0.43) through the judgment matrix. The approximation ideal solution sorting module standardizes the 12 engines to be repaired, calculates the proximity value through weighted Euclidean distance, and generates a priority sequence that improves maintenance resource utilization by 35%, shortens the average engine maintenance cycle by 22%, and reduces flight delay losses by approximately 12 million yuan annually.

[0188] Application Example 3: Intelligent Factory Equipment Maintenance Management System

[0189] The system automatically switches evaluation paths based on scenario parameters of "equipment failure risk level ≥ 3": a hesitant fuzzy evaluation path is used for high-risk equipment (such as welding robots), incorporating the fuzzy assessments of 5 equipment engineers; a hierarchical analysis-approximation ideal solution path is used for conventional equipment. The dynamic weight calculation module, through attribute correlation analysis, discovered a significant negative correlation (correlation coefficient -0.69) between "robotic arm positioning error" and "welding yield," and optimized weight allocation accordingly. After applying this system, production line equipment downtime was reduced by 38%, maintenance costs by 28%, and product defect rate by 15%, validating the technical value of this invention in the field of intelligent industrial manufacturing.

[0190] It should be noted that embodiments of the present invention can be implemented in hardware, software, or a combination of both. The hardware portion can be implemented using dedicated logic; the software portion can be stored in memory and executed by a suitable instruction execution system, such as a microprocessor or dedicated-design hardware. Those skilled in the art will understand that the above-described devices and methods can be implemented using computer-executable instructions and / or included in processor control code, for example, such code provided on a carrier medium such as a disk, CD, or DVD-ROM, a programmable memory such as read-only memory (firmware), or a data carrier such as an optical or electronic signal carrier. The devices and modules of the present invention can be implemented by hardware circuitry such as very large-scale integrated circuits or gate arrays, semiconductors such as logic chips, transistors, or programmable hardware devices such as field-programmable gate arrays, programmable logic devices, etc., or by software executed by various types of processors, or by a combination of the above-described hardware circuitry and software, such as firmware.

[0191] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any modifications, equivalent substitutions, and improvements made by those skilled in the art within the scope of the technology disclosed in the present invention, and within the spirit and principles of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A method for prioritizing equipment maintenance tasks, characterized in that, The method for prioritizing equipment maintenance tasks includes the following steps: Step 1, Data Acquisition Steps; Obtain the maintenance task dataset and attribute evaluation set, and determine the evaluation path based on the scenario discrimination parameters; when the scenario discrimination parameters are less than the preset scenario threshold, execute the hesitant fuzzy evaluation path: Step 2, hesitant fuzzy modeling steps; A hesitant fuzzy matrix is ​​constructed based on the evaluation data of multiple decision-makers, and the expected value and variance value of each hesitant fuzzy element are calculated. Step 3, dynamic weight calculation steps; Construct a correlation matrix, and build a weight optimization objective function based on hesitant fuzzy meta-features, attribute dispersion, and attribute correlation. Solve the attribute weight vector using the Lagrange multiplier method. Step 4: Comprehensive evaluation and ranking steps; The attribute values ​​are normalized to obtain a normalized attribute matrix. A comprehensive evaluation score is calculated based on the normalized attribute matrix and the attribute weight vector. The scores are then sorted in descending order to generate a priority ranking sequence. When the scene discrimination parameter is greater than or equal to a preset scene threshold, the analytic hierarchy process (AHP) – approximation of the ideal solution evaluation path is executed. Step 5, Hierarchical Analysis weight calculation steps; Construct a hierarchical structure model and a judgment matrix, calculate the weights of the criterion layer and the index layer, and verify the effectiveness of the judgment matrix through the consistency ratio value; Step 6, sorting steps to approximate the ideal solution; A standardized decision matrix is ​​obtained by standardizing the attribute data using a nonlinear value function, and positive and negative ideal solution vectors are determined. The weighted Euclidean distance from each task to the positive and negative ideal solution vectors is calculated. Based on the weighted Euclidean distance, the proximity value is calculated, and a priority sorting sequence is generated by arranging the proximity values ​​in descending order.

2. The equipment maintenance task priority evaluation method as described in claim 1, characterized in that, The attribute evaluation set includes equipment importance attribute, equipment repair time attribute, equipment resource demand attribute, and equipment location attribute; the equipment importance attribute is a benefit-type attribute, with a value range of 0 to 1. The repair time attribute, resource demand attribute, and location attribute of the equipment to be repaired are all cost-type attributes. The scenario discrimination parameter is the average remaining time to complete the task, and the scenario threshold ranges from 24h to 72h.

3. The equipment maintenance task priority evaluation method as described in claim 1, characterized in that, The correlation matrix is ​​a symmetric matrix, with its elements ranging from 0 to 1, and the diagonal elements having a value of 1. When there is a strong correlation between two attributes, the corresponding element value ranges from 0.6 to 0.9; when there is a moderate correlation between two attributes, the corresponding element value ranges from 0.3 to 0.6; and when there is a weak correlation between two attributes, the corresponding element value ranges from 0 to 0.

3.

4. The equipment maintenance task priority evaluation method as described in claim 1, characterized in that, The expected value of the hesitant fuzzy element is equal to the arithmetic mean of the evaluation scores of each decision-maker; the variance of the hesitant fuzzy element is equal to the arithmetic mean of the squares of the differences between the evaluation scores of each decision-maker and the expected value. The weight optimization objective function includes three weighted components. The first component is calculated based on the ratio of the expected value to the variance value of the hesitant fuzzy element. The second component is calculated based on the sum of the Euclidean distances of each task on the attribute. The third component is calculated based on the sum of the complements of the correlation between the attribute and other attributes. The sum of the balance coefficients of the three components is 1.

5. The equipment maintenance task priority evaluation method as described in claim 1, characterized in that, The nonlinear value function is the Logistic function; for positive indicators, the standardized value is equal to the Logistic function value; for negative indicators, the standardized value is equal to 1 minus the Logistic function value; the parameters of the Logistic function include the indicator's intermediate threshold and the indicator's dispersion parameter. The hierarchical model includes a target layer, a criterion layer, and an indicator layer. The criterion layer includes the urgency criterion, equipment importance criterion, maintenance resource demand criterion, and fault severity criterion. The indicator layer includes the remaining completion time indicator, the urgency level indicator, the strategic value of the equipment indicator, the scope of the fault impact indicator, the maintenance man-hour indicator, the spare parts preparation time indicator, the fault deterioration risk indicator, and the economic loss estimation indicator.

6. The equipment maintenance task priority evaluation method as described in claim 1, characterized in that, The consistency ratio is equal to the ratio of the consistency index to the random consistency index; the consistency index is equal to the difference between the largest eigenvalue of the judgment matrix and the number of indicators divided by the number of indicators minus 1; when the consistency ratio is less than 0.1, the judgment matrix is ​​deemed to be of good consistency.

7. An equipment maintenance task priority evaluation system implementing the equipment maintenance task priority evaluation method as described in any one of claims 1-6, characterized in that, The equipment maintenance task priority evaluation system includes: The data acquisition module is used to acquire maintenance task datasets and attribute evaluation sets, and determine the evaluation path based on scenario discrimination parameters; The scene discrimination module is used to compare the scene discrimination parameters with the preset scene thresholds. When the scene discrimination parameters are less than the scene thresholds, the hesitant fuzzy evaluation path is triggered. When the scene discrimination parameters are greater than or equal to the scene thresholds, the hierarchical analysis-approximation ideal solution evaluation path is triggered. The hesitant fuzzy modeling module is used to construct a hesitant fuzzy matrix based on evaluation data from multiple decision-makers and to calculate the expected value and variance value of each hesitant fuzzy element. The dynamic weight calculation module is used to construct the correlation matrix and build a weight optimization objective function based on hesitant fuzzy element features, attribute dispersion, and attribute correlation. The attribute weight vector is obtained by solving the Lagrange multiplier method. The comprehensive evaluation and ranking module is used to normalize the attribute values ​​to obtain a normalized attribute matrix, calculate the comprehensive evaluation score based on the normalized attribute matrix and the attribute weight vector, and generate a priority ranking sequence by arranging the comprehensive evaluation scores in descending order. The hierarchical analysis weight module is used to construct a hierarchical structure model and a judgment matrix, calculate the weights of the criterion layer and the index layer, and verify the effectiveness of the judgment matrix through the consistency ratio value. The approximation ideal solution sorting module is used to standardize attribute data using a nonlinear value function to obtain a standardized decision matrix, determine the positive ideal solution vector and the negative ideal solution vector, calculate the weighted Euclidean distance value and the proximity value, and generate a priority sorting sequence by arranging them in descending order according to the proximity value.

8. A computer device, characterized in that, The computer device includes a memory and a processor. The memory stores a computer program, which, when executed by the processor, causes the processor to perform the steps of the equipment maintenance task priority evaluation method as described in any one of claims 1-6.

9. A computer-readable storage medium storing a computer program, which, when executed by a processor, causes the processor to perform the steps of the equipment maintenance task priority evaluation method as described in any one of claims 1-6.

10. An information data processing terminal, characterized in that, The information data processing terminal is used to implement the equipment maintenance task priority evaluation system as described in claim 7.

Images