An intelligent scheduling system and method for repairing war injuries of an airplane

Abstract

The application provides an airplane battle damage repair intelligent scheduling system and method, relates to the technical field of aircraft maintenance, and comprises a damage assessment module, a maintenance resource management module, a team scheduling decision module, a knowledge sharing module and a data analysis and feedback module. The damage assessment module quantitatively evaluates the battle damage to obtain damage severity; the maintenance resource management module evaluates a repair team to obtain resource availability data; the team scheduling decision module makes scheduling decisions based on the dynamic weights of multidimensional decision factors; the knowledge sharing module records verified repair experience; and the data analysis and feedback module optimizes the weights and transfers parameters. The application solves the problem that the decision weights of the traditional scheduling system are fixed and cannot adapt to complex battlefield environments by establishing a dynamic weight self-adaptive adjustment mechanism, enables the system to learn from each repair experience and continuously optimize the decision process, and realizes a fundamental change from static rules to intelligent self-adaptation.

Description

Technical Field

[0001] This invention relates to the field of aircraft maintenance technology, specifically to an intelligent scheduling system and method for aircraft combat damage repair. Background Technology

[0002] Traditional aircraft combat damage repair scheduling systems primarily rely on static scheduling rules and expert experience for decision-making. These systems typically employ pre-defined priority rules, determining the repair sequence based on single factors such as damage severity, repair time, or resource requirements. However, this static scheduling model has several limitations. First, the fixed decision weights cannot be dynamically adjusted based on actual combat damage and repair effectiveness, making it difficult to adapt to complex and ever-changing battlefield environments. Second, severe information isolation exists between different specialized repair teams, lacking effective experience-sharing mechanisms. This prevents valuable repair experience from being transferred between teams, resulting in resource waste and inefficiency. Third, resource allocation is inefficient, often leading to excessively long waiting times for repair tasks and the idle or over-utilized use of critical resources. Finally, the system lacks adaptive learning capabilities, failing to summarize patterns from past repair experiences and optimize subsequent decision-making processes.

[0003] In existing technologies, some studies have attempted to improve repair efficiency by optimizing scheduling algorithms. For example, scheduling schemes based on the critical path method can identify critical maintenance tasks that affect the overall repair time, but they still use a fixed weight allocation mechanism. Other studies have proposed scheduling strategies based on multi-objective optimization, considering multiple factors such as time, cost, and resources. However, these methods are essentially still within the scope of static optimization and lack the ability to dynamically adjust based on feedback from repair results.

[0004] In terms of team collaboration, traditional systems typically treat repair teams from different specialties as independent execution units. Each team operates only within its own area of ​​expertise, lacking information exchange and experience sharing. This model leads to severe information silos, preventing other teams from learning from the valuable repair experience accumulated by one team, greatly limiting the improvement of overall repair capabilities. Especially when facing complex damage or novel faults, teams often need to repeatedly explore solutions, resulting in a huge waste of time and resources.

[0005] Furthermore, existing scheduling systems have shortcomings in handling resource constraints. In real-world battlefield environments, maintenance resources are often limited and dynamically changing, including professional personnel, maintenance tools, and spare parts inventory. Traditional systems struggle to perceive changes in resource status in real time and cannot dynamically adjust repair strategies based on resource availability, frequently resulting in inappropriate resource allocation.

[0006] More importantly, traditional scheduling systems lack mechanisms to learn and improve from repair practices. After each repair task is completed, the system cannot automatically analyze the successes and failures of the repair process, nor can it identify which decision-making factors had a positive or negative impact on the repair effect. Therefore, it cannot achieve continuous optimization of system performance. This lack of self-evolutionary capability keeps the system at a relatively low efficiency level.

[0007] Chinese patent document CN116205500B discloses a design and evaluation method for aircraft battle damage repair schemes, and discloses a scheduling decision-making technology scheme based on critical path algorithm and multiple priority strategies. It achieves the technical effect of generating multiple reasonable repair scheduling schemes and improving the objectivity of repair schemes. However, it still has problems such as fixed decision weights that cannot be dynamically adjusted, lack of knowledge sharing mechanism among repair teams, and inability of the system to adaptively learn and optimize based on repair results.

[0008] Chinese patent document CN116029700A discloses a method for determining damage assessment and maintenance information of aircraft systems based on missions. It discloses a technical solution that establishes a mission database, performs damage mode and impact analysis, and constructs a wartime availability criterion library. This method has the technical effect of quickly determining the war damage maintenance list under a specific mission and shortening the damage assessment time. However, it still has problems such as the inability to dynamically optimize based on a static criterion library, the lack of inter-team collaboration and knowledge transfer mechanisms, and the inability to achieve intelligent scheduling of repair resources.

[0009] With the rapid development of artificial intelligence technology, biomimetic algorithms have shown great potential in the optimization of complex systems. Bee swarm intelligence, as an important biomimetic algorithm, simulates the collective behavior of a bee colony, enabling information sharing and collaborative decision-making among individuals, and performs exceptionally well in solving complex optimization problems. However, there is currently no research applying the principles of bee swarm intelligence to aircraft combat damage repair scheduling, and the level of intelligence in this field remains relatively low. Summary of the Invention

[0010] The purpose of this invention is to provide an intelligent scheduling system and method for aircraft combat damage repair that can dynamically adjust decision weights, promote knowledge sharing among teams, achieve optimal resource allocation, and possess adaptive learning capabilities.

[0011] To achieve the above objectives, the present invention provides the following technical solution: an intelligent scheduling system for aircraft combat damage repair, comprising a damage assessment module, a maintenance resource management module, a team scheduling decision module, a knowledge sharing module, and a data analysis and feedback module;

[0012] The damage assessment module is used to receive aircraft battle damage information and quantitatively assess the damage location, type and severity to obtain the damage severity.

[0013] The maintenance resource management module assesses available repair teams, parts inventory, and equipment tools to obtain resource availability data, which includes skill matching, wait time, and parts availability.

[0014] The team scheduling decision module obtains scheduling information from the damage assessment module, maintenance resource management module, and data analysis and feedback module. It makes scheduling decisions based on the weights of multi-dimensional decision factors, transmits repair experience to the knowledge sharing module, and transmits repair result data to the data analysis and management module. The weights are dynamically adjusted according to the aircraft repair result feedback. The multi-dimensional decision factors include skill matching degree, waiting time, parts availability, and damage severity.

[0015] The knowledge sharing module receives repair experience and transmission parameters, records and verifies the received repair experience using decentralized distributed ledger technology, transmits the knowledge to relevant teams, and transmits the knowledge transmission effect data to the data analysis and feedback module.

[0016] The data analysis and feedback module collects repair result data and knowledge transfer effect data, optimizes decision weights and transfer parameters based on the obtained data, and feeds back to the corresponding modules.

[0017] Furthermore, the specific steps of the team scheduling decision module in making scheduling decisions based on the weights of multi-dimensional decision factors include:

[0018] S1. Obtain quantitative data on the location, type, and severity of aircraft damage from the damage assessment module, analyze the battle damage information, and obtain the damage severity.

[0019] S2. Obtain information from the maintenance resource management module, check the skill level of available repair teams, current workload, parts inventory status and equipment and tool availability, and obtain resource availability data;

[0020] S3. Based on damage severity and resource availability data, calculate multi-dimensional weighted decision values ​​using the weights of multi-dimensional decision factors;

[0021] S4. Generate the optimal scheduling scheme based on the calculated multi-dimensional weight decision values, and select the team with the highest scheduling decision value and that meets the resource constraints as the execution team.

[0022] Furthermore, the specific steps for the team scheduling decision module to dynamically adjust the weights include:

[0023] S1. The team scheduling decision module initializes the weight values ​​of each decision factor;

[0024] S2. Based on the completion status of each repair task, provide repair result data to the data analysis and feedback module;

[0025] S3. The data analysis and feedback module evaluates the repair results and adjusts the weights. If the repair is successful and efficient, the weight of the corresponding success factor is increased; if the repair is delayed or inefficient, the weight of the corresponding influencing factor is decreased.

[0026] Furthermore: the calculation formula used by the team scheduling decision module is as follows:

[0027] D = W1×S1 + W2×S2 + W3×S3 + W4×S4

[0028] Where D is the scheduling decision value, W1 to W4 are the corresponding decision factors, S1 to S4 are the standardized scores of the corresponding decision factors, and the value of S ranges from 0 to 1.

[0029] Furthermore: the knowledge-sharing module records and verifies the repair experience of each team, and the knowledge sharing specifically includes the following steps:

[0030] S1. Establish professional field scores and non-professional field competence scores for each repair team;

[0031] S2. When a team successfully repairs battle damage, it receives repair experience from the team scheduling decision module, records and verifies the received repair experience.

[0032] S3. Transfer the repair experience to the relevant teams according to the preset knowledge transfer path;

[0033] S4. Update the non-professional area capability score of the receiving team based on the repair experience received;

[0034] S5. Based on the frequency and effectiveness of knowledge transfer between teams, dynamically optimize the knowledge transfer path by combining transfer parameters.

[0035] Furthermore: the data analysis and feedback module includes a data acquisition layer, a data processing layer, an algorithm optimization layer, and a closed-loop feedback application layer connected in sequence;

[0036] The data acquisition layer is used to collect data, including repair results and knowledge transfer effect data.

[0037] The data processing layer standardizes the raw data using a sliding time window strategy;

[0038] The algorithm optimization layer includes an adaptive weight optimization algorithm and a knowledge sharing parameter optimization algorithm. The adaptive weight optimization algorithm updates the weights based on the repair results, and the knowledge sharing parameter optimization algorithm updates the transmission parameters based on the knowledge transmission effect.

[0039] The closed-loop feedback application layer optimizes the entire system through three time scales: short-cycle feedback, medium-cycle feedback, and long-cycle feedback.

[0040] Furthermore: the data processing layer standardizes the original data using a sliding time window strategy. The formula used for standardization is:

[0041] β=(R×T 理想 ) / T 实际

[0042] Where β is the repair efficiency index, R is the repair quality score, and its value ranges from 0 to 1, T 理想 For ideal repair time, T 实际 This refers to the actual repair time.

[0043] Furthermore: the specific method by which the adaptive weight optimization algorithm updates the weights based on the repair results is as follows:

[0044] When β>1, W i新 =W i旧 ×(1+0.15×(β-1)×C i );

[0045] When β < 1, W i新 =W i旧 ×(1-0.2×(1-β)×C i );

[0046] Where C i This is the contribution coefficient of this factor to the current repair task, with a value ranging from 0 to 1; the updated weight value W i新 Through formula W i归一化 =W i新 / ∑W j新 Normalization is performed, W i旧 The weight values ​​are those before the update.

[0047] Furthermore: when the knowledge-sharing module transfers knowledge to relevant teams, it calculates the knowledge increment, and the formula used for calculating the knowledge increment is:

[0048] Z = K × Y × F

[0049] Where Z represents the knowledge increment, K represents the transmission parameter, Y represents the professional level of the providing team, and F represents the severity of the damage;

[0050] A time decay function is introduced in the knowledge transfer process. and efficiency improvement upper limit function ,

[0051] Where K0 is the initial repair experience transfer parameter, λ is the decay rate, n is the number of times the team receives repair experience, and t is the time interval since the repair experience was generated.

[0052] A scheduling method for an intelligent scheduling system for aircraft combat damage repair includes the following steps:

[0053] S1: The damage assessment module receives aircraft battle damage data and assesses the aircraft's battle damage status to obtain the severity of the damage.

[0054] S2: The maintenance resource management module checks available repair resources and obtains resource availability data;

[0055] S3: The team scheduling decision module calculates the scheduling decision value for each team based on damage severity and resource availability data, selects the team with the highest scheduling decision value to execute the repair task, and transmits the repair results to the data analysis module and the repair experience to the knowledge sharing module after the task is completed.

[0056] S5: The knowledge sharing module receives repair experience, transmits the repair experience to relevant teams based on the received repair experience, updates the capability score of the receiving teams, and transmits the knowledge transmission effect data to the data analysis and feedback module.

[0057] S6: The data analysis and feedback module receives the repair results and dynamically adjusts the decision weights, receives the knowledge transfer effect data and optimizes the transfer parameters, and sends the transfer parameters to the knowledge sharing module. The knowledge sharing module updates the repair experience transfer path according to the received transfer parameters.

[0058] Compared with the prior art, the present invention has the following advantages:

[0059] I. This invention solves the problem that traditional scheduling systems with fixed decision weights cannot adapt to complex battlefield environments by establishing a dynamic weight adaptive adjustment mechanism. This invention adopts a multi-module architecture combined with iterative algorithm updates, dynamically adjusting the weights of decision factors such as skill matching degree, waiting time, component availability, and damage severity based on feedback from repair results. This allows the system to learn from each repair experience and continuously optimize the decision-making process, achieving a fundamental shift from static rules to intelligent adaptation.

[0060] Second, this invention solves the problems of information isolation and lack of experience transfer between different specialized repair teams through a knowledge-sharing network. This invention employs a knowledge-sharing module based on decentralized distributed ledger technology to enable rapid transfer of repair experience among multiple specialized teams, such as those working with hydraulic systems, engines, electronic systems, avionics systems, and fuel systems. This improves the efficiency of each team in areas outside their expertise, effectively breaking down information barriers in traditional maintenance systems and realizing a shift from individual specialization to collective intelligence.

[0061] Third, this invention solves the problem of the system's lack of self-learning and continuous improvement capabilities through a multi-timescale closed-loop feedback optimization mechanism. This invention establishes a three-layer feedback system comprising short-cycle, medium-cycle, and long-cycle phases. Through the collaborative work of the data acquisition layer, data processing layer, algorithm optimization layer, and closed-loop feedback application layer, it achieves continuous optimization and self-evolution of system performance, ensuring the continuous improvement of scheduling decision-making capabilities.

[0062] Fourth, this invention significantly improves the overall efficiency of aircraft combat damage repair. It increases the utilization efficiency of repair resources, improves the efficiency of repair knowledge transfer, and makes scheduling and feedback more timely and effective. It effectively shortens the repair cycle of combat-damaged aircraft, enhances the continuous combat capability of the troops, and provides strong technical support for the rapid repair needs in modern warfare. Attached Figure Description

[0063] Figure 1 A schematic diagram of the architecture of an intelligent scheduling system for aircraft combat damage repair provided by the present invention;

[0064] Figure 2 A schematic diagram of the workflow of the team scheduling decision module of an intelligent scheduling system for aircraft combat damage repair provided by the present invention.

[0065] Figure 3 This invention provides a schematic diagram of the dynamic evolution mechanism of decision weights in an intelligent scheduling system for aircraft combat damage repair.

[0066] Figure 4 A schematic diagram of the data analysis and feedback module architecture in an intelligent scheduling system for aircraft combat damage repair provided by this invention;

[0067] Figure 5 The flowchart of the scheduling method of an intelligent scheduling system for aircraft combat damage repair provided by the present invention is shown.

[0068] In the picture:

[0069] 11. Damage Assessment Module; 12. Maintenance Resource Management Module; 13. Team Scheduling Decision Module; 14. Knowledge Sharing Module; 15. Data Analysis and Feedback Module. Detailed Implementation

[0070] The technical solution of the present invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0071] Example 1

[0072] like Figure 1As shown, this embodiment provides an intelligent scheduling system for aircraft combat damage repair, including a damage assessment module 11, a maintenance resource management module 12, a team scheduling decision module 13, a knowledge sharing module 14, and a data analysis and feedback module 15. These modules are interconnected through data interfaces to form a complete intelligent scheduling system.

[0073] The damage assessment module 11, serving as the system's information input, is responsible for receiving aircraft combat damage information, including damage location, damage type (e.g., structural damage, system failure), and a preliminary description of the damage severity. The damage assessment module 11 quantifies the location, type, and severity of damage by analyzing battlefield images, sensor data, and preliminary assessment reports from maintenance personnel. Specifically, the system extracts damage features using image recognition algorithms and combines this with the preliminary assessment from maintenance experts to generate a damage severity rating between 0 and 1, where 0 represents no damage and 1 represents severe damage requiring immediate attention. This provides a quantitative reference for subsequent dispatch decisions.

[0074] The Maintenance Resource Management Module 12 manages available repair teams, parts inventory, and equipment tools, ensuring the system has a comprehensive understanding of the current availability of repair resources. Specifically, Module 12 employs a graph database-based resource monitoring system to track the location, status, and availability of various resources in real time. This module assesses available repair teams, parts inventory, and equipment tools, generating resource availability data. This data includes three main indicators: skill matching, waiting time, and parts availability. Skill matching is determined by analyzing the degree to which the repair team's professional skills match the current damage type, with a value ranging from 0 to 1. Waiting time reflects the current workload of the repair team; the system calculates the available time based on the team's current task schedule. Parts availability assesses the inventory and supply cycle of the parts needed to complete the repair. By comprehensively analyzing this resource information, the module provides accurate resource constraints for scheduling decisions.

[0075] The team scheduling decision module 13 is responsible for obtaining scheduling information from the damage assessment module 11 and the maintenance resource management module 12, while also receiving optimization weight parameters from the data analysis and feedback module 15. It makes scheduling decisions based on the dynamic weights of multi-dimensional decision factors. These multi-dimensional decision factors include four key elements: skill matching, waiting time, component availability, and damage severity. The core responsibility of this module is to execute decisions based on parameters. Specifically, it receives optimization weights from the data analysis module, performs the decision calculation D=W1×S1+W2×S2+W3×S3+W4×S4, and outputs the optimal scheduling scheme, but does not participate in the weight calculation and optimization process. After completing the repair task, this module transfers the repair experience to the knowledge sharing module 14 and simultaneously transmits the repair result data to the data analysis and feedback module, forming a closed-loop optimization mechanism.

[0076] The knowledge-sharing module 14 enables the transfer of experience and skills among teams. Its core responsibility is to execute knowledge transfer based on parameters, receiving optimized transfer parameters from the data analysis module, executing the knowledge transfer, and managing the knowledge transfer network and decentralized ledger, but it does not participate in the optimization calculation of the transfer parameters. To ensure the credibility and integrity of knowledge transfer, this module uses decentralized distributed ledger technology to record and verify each team's repair experience. After verification, the module transfers knowledge to relevant teams according to a preset knowledge transfer path, enabling professional teams to break through traditional isolation models and achieve experience sharing and collective intelligence. The effectiveness of knowledge transfer is quantitatively evaluated using efficiency improvement indicators, and the relevant data is transmitted to the data analysis and feedback module 15 for subsequent parameter optimization.

[0077] The Data Analysis and Feedback Module 15 is responsible for parameter optimization and system learning. This module is responsible for weight optimization, calculating and updating the weight values ​​of W1-W4, and optimizing transmission parameters, dynamically optimizing key parameters such as the transmission coefficient and attenuation rate λ. This module does not directly perform decision calculations or knowledge transfer, but focuses on optimizing the parameters of these processes. The Data Analysis and Feedback Module 15 collects repair result data from the Team Scheduling Decision Module 13 and knowledge transfer effect data from the Knowledge Sharing Module 14. Based on the collected data, it uses algorithms to optimize parameters and transmits the optimization results to the corresponding modules through a closed-loop feedback mechanism. Specifically, this module pushes new weight parameters to the decision module through short-cycle feedback (4 hours), pushes new transmission parameters to the knowledge sharing module 14 through medium-cycle feedback (24 hours), and performs global system parameter tuning through long-cycle feedback (1 week). This closed-loop feedback mechanism ensures that the system can continuously learn and evolve in practical applications, gradually improving overall performance.

[0078] In one specific implementation of this embodiment, such as Figure 2As shown, the team scheduling decision module 13's scheduling decision process based on the weights of multi-dimensional decision factors includes four specific steps. First, the module obtains quantitative data on the location, type, and severity of aircraft damage from the damage assessment module 11. By comprehensively analyzing this damage information, a standardized damage severity index is calculated. Second, the module obtains current resource status information from the maintenance resource management module 12, comprehensively checking the skill level, current workload, parts inventory status, and equipment availability of available repair teams, generating complete resource availability data. Third, based on the obtained damage severity and resource availability data, the system calculates a multi-dimensional weighted decision value using the current weight values ​​of the multi-dimensional decision factors. This calculation process comprehensively considers four key factors: skill matching, waiting time, parts availability, and damage severity. Finally, based on the calculated multi-dimensional weighted decision value, an optimal scheduling plan is generated. The system selects the team with the highest scheduling decision value that meets the resource constraints as the execution team, ensuring that the repair task can be completed efficiently.

[0079] In one specific implementation of this embodiment, such as Figure 2 As shown, the team scheduling decision module 13 and the data analysis and feedback module 15 assume different responsibilities during the dynamic weight adjustment process, reflecting the system's collaborative working mechanism. The weight adjustment responsibility of the data analysis and feedback module 15 is to perform the specific calculations for weight adjustment. This module first initializes the weight values ​​of each decision factor to equal values, typically set to 0.25. During system operation, upon receiving repair result data, the module uses a non-linear function to update the weights: when the repair efficiency index β is greater than 1, the formula W is used... i新 =W i旧 ×(1+0.15×(β-1)×C i Increase the weight of relevant success factors; when β is less than 1, use formula W. i新 =W i旧 ×(1-0.2×(1-β)×C i Reduce the weights of relevant influencing factors. Then perform weight normalization processing. i归一化 =W i新 / ∑W j新 This ensures that the total weights are equal to 1. The weight adjustment responsibility of the team scheduling decision module 13 is to receive and apply the weight parameters from the data analysis and feedback module 15, and to use the dynamic weight execution formula D=W1×S1+W2×S2+W3×S3+W4×S4 in the decision calculation. It participates in the dynamic evolution process of weights but does not perform weight calculation optimization.

[0080] In one specific implementation of this embodiment, the team scheduling decision module 13 calculates the scheduling decision value using a formula. The formula for calculating the scheduling decision value is D = W1 × S1 + W2 × S2 + W3 × S3 + W4 × S4, where D represents the final scheduling decision value, used to quantitatively evaluate the overall suitability of each repair team to perform the current task. W1 to W4 represent the weight coefficients corresponding to the four decision factors: skill matching, waiting time, component availability, and damage severity, respectively. These weight values ​​change dynamically during the system learning process. S1 to S4 are the standardized scores for the corresponding decision factors, with all S values ​​ranging from 0 to 1, where 0 indicates the worst condition for that factor and 1 indicates the best condition for that factor. After calculating the scheduling decision value for each team, the system selects the team with the highest D value to perform the repair task.

[0081] In one specific implementation of this embodiment, the knowledge-sharing module 14 achieves effective transfer of repair experience among teams through five steps. First, the system establishes a detailed capability assessment file for each repair team, including professional domain scores and non-professional domain capability scores. The professional domain score reflects the team's technical level in its area of ​​expertise, while the non-professional domain capability score reflects the team's auxiliary capabilities in other professional fields. When a team successfully completes a battle damage repair task, the knowledge-sharing module 14 receives the team's repair experience from the team scheduling decision module 13, including detailed information such as the technical solutions used, problems encountered and their solutions, and resource usage. Subsequently, the module uses distributed ledger technology to record and verify the received repair experience, ensuring the authenticity and completeness of the experience data. Fourth, the system transfers the verified repair experience to relevant teams according to a preset knowledge transfer path. The determination of the transfer path considers multiple factors such as the professional relevance between teams, geographical location, and current work status. Finally, the receiving team updates its non-professional domain capability score based on the acquired repair experience. Simultaneously, the system dynamically optimizes the knowledge transfer path based on the frequency and actual effect of knowledge transfer between teams, combined with transfer parameters, to ensure the continuous improvement of the knowledge-sharing network.

[0082] In one specific implementation of this embodiment, such as Figure 4As shown, the data analysis and feedback module 15 adopts a four-layer architecture. The data acquisition layer is responsible for collecting four types of key data: repair time series, team status information, repair success rate, and component consumption data. The data processing layer standardizes the collected raw data using a sliding time window strategy and calculates the repair efficiency index β. The algorithm optimization layer contains two key algorithms: an adaptive weight optimization algorithm that updates weight parameters, and a knowledge-sharing parameter optimization algorithm that updates parameters such as the transfer coefficient and decay rate. The closed-loop feedback application layer is responsible for transmitting the optimization results to the corresponding modules according to different time scales: short-cycle feedback pushes new weights to the decision-making module every 4 hours, medium-cycle feedback pushes new transfer parameters to the knowledge-sharing module 14 every 24 hours, and long-cycle feedback performs global system parameter tuning weekly. This layered architecture ensures the integrity of the parameter optimization processing system and enables timely feedback.

[0083] In one specific implementation of this embodiment, the data processing layer uses the repair efficiency index as the core evaluation criterion and employs a sliding time window strategy to scientifically standardize the original data. The formula for calculating the repair efficiency index is β=(R×T) 理想 ) / T 实际 This formula comprehensively considers two key dimensions: repair quality and time efficiency. β is the repair efficiency index, a comprehensive indicator measuring the completion of the repair task; R is the repair quality score, strictly limited to a range of 0 to 1, and is comprehensively evaluated by professional technicians based on factors such as the degree of functional recovery of the aircraft after repair, safety indicators, and impact on service life; T... 理想 This value, representing the standard time required to complete the same repair task under ideal conditions, is determined based on historical maintenance data and theoretical analysis; T 实际 This represents the actual time taken for this repair task. When the β value is greater than 1, it indicates that the repair effect exceeded expectations, and the system will increase the weight of relevant success factors; when the β value is less than 1, it indicates that the repair effect was poor, and the system will decrease the weight of relevant influencing factors.

[0084] In one specific embodiment of this example, the adaptive weight optimization algorithm and the nonlinear function in the team scheduling decision module 13 play different roles. The goal of the nonlinear function in the data analysis and feedback module 15 is to control the magnitude and speed of weight updates. This is achieved by using coefficients 0.15 and 0.2 to control the adjustment intensity. These coefficients are verified optimal values ​​determined through a comprehensive method, and the contribution coefficient C is also considered. i Adjusting the precision of the weighting provides stability assurance to prevent large single weight adjustments and ensures system convergence.

[0085] The goal of the nonlinear function in the team scheduling decision module 13 is to smoothly handle weight changes during the decision calculation process. Its mechanism involves receiving weights that have already undergone nonlinear adjustment, maintaining computational stability in multi-dimensional decision-making, and providing stability guarantees to ensure the continuity and reliability of the decision output. The specific weight update method is divided into two cases: when the repair efficiency index β is greater than 1, it indicates that the repair effect is better than expected, and the system uses formula W... i新 =W i旧 ×(1+0.15×(β-1)×C i Increase the weight of relevant factors; when β is less than 1, it indicates that the repair effect is not ideal, and the system uses formula W. i新 =W i旧 ×(1-0.2×(1-β)×C i Reduce the weight of relevant factors. In these formulas, C i This represents the contribution coefficient of the decision factor to the current repair task, ranging from 0 to 1. This coefficient is determined by analyzing the actual role of each factor in the repair process. To ensure that the sum of all weights remains 1, the system uses the normalization formula W after the update is completed. i归一化 =W i新 / ∑W j新 All weights are standardized, where W i旧 The weight values ​​are those before the update.

[0086] In one specific embodiment of this example, the knowledge sharing module 14 uses a precise mathematical model to calculate the knowledge increment when transferring knowledge between teams, wherein the transfer coefficient is dynamically optimized by the data analysis and feedback module 15. The knowledge increment is calculated using the formula Z=K×Y×F, where Z represents the knowledge increment that the receiving team can obtain, K is the current transfer parameter, which is dynamically optimized by the data analysis and feedback module 15 based on historical transfer effects and pushed to the knowledge sharing module 14; Y is the professional level score of the team providing repair experience, reflecting its technical capabilities in the relevant field; and F is the severity of damage in the current repair task, the more severe the damage, the higher the value of the resulting repair experience.

[0087] To more realistically simulate the natural laws of knowledge transfer, the data analysis and feedback module 15 provides the knowledge sharing module 14 with optimized parameters such as the decay rate λ, based on which the knowledge sharing module 14 executes the time decay function K(t) = K0 × e -λt and the upper limit function of efficiency improvement Y max =0.4×(1-e -0.5nIn the time decay function, K0 is the initial repair experience transfer parameter, and t is the time interval since the repair experience was generated. This function reflects the time-sensitive nature of knowledge. In the efficiency improvement cap function, n is the number of times the team receives repair experience. This function ensures that the team's ability improvement follows a diminishing returns principle, avoiding unlimited growth.

[0088] Example 2

[0089] like Figure 5 As shown, this embodiment provides a scheduling method applied to the above-mentioned intelligent scheduling system for aircraft combat damage repair. The method includes the following steps:

[0090] Step S1: Damage Assessment Module 11 receives and assesses the aircraft's combat damage. Damage Assessment Module 11 receives aircraft combat damage data from the battlefield, including damage images, sensor detection data, preliminary assessment reports from on-site personnel, and other information. Damage Assessment Module 11 uses image recognition algorithms and an expert knowledge system to comprehensively analyze the aircraft's damage, determining the specific location, type, and extent of the damage, and ultimately generating a standardized damage severity index. This not only considers the physical characteristics of the damage but also the degree of impact of the damage on the overall aircraft performance, providing a scientific basis for subsequent dispatch decisions.

[0091] Step S2: The maintenance resource management module 12 conducts a comprehensive check and evaluation of currently available repair resources. This module monitors the working status, skill level, and geographical location of each repair team in real time, while tracking the quantity, specifications, and supply cycle of spare parts inventory, as well as the availability and distribution of maintenance equipment and tools. By comprehensively analyzing this resource information, the module generates detailed resource availability data, including key indicators such as the skill matching degree of each team with the current damage type, estimated waiting time, and the availability of required spare parts.

[0092] Step S3: The team scheduling decision module 13 receives damage severity data from the damage assessment module 11 and resource availability data from the maintenance resource management module 12. Simultaneously, it receives optimized weight parameters pushed by the data analysis and feedback module 15, and uses a dynamic weighting mechanism to calculate the scheduling decision value for each repair team. During the calculation, the system, based on the currently received weight configuration, comprehensively considers four key factors: skill matching, waiting time, component availability, and damage severity, generating a comprehensive score for each potential repair team. Subsequently, the system selects the team with the highest scheduling decision value that meets the current resource constraints as the final execution team. After the repair task is completed, the work results and experience of the execution team will be transmitted to the data analysis module and the knowledge sharing module 14 respectively, forming a complete information transmission loop.

[0093] Step S4: The knowledge-sharing module 14 receives repair experience from the team scheduling and decision-making module 13. This experience includes crucial information such as the technical solutions adopted by the execution team during the repair process, the technical difficulties encountered, innovative solutions, and resource optimization strategies. The knowledge-sharing module 14 first uses distributed ledger technology to verify and record this experience, ensuring the authenticity and completeness of the information. Then, based on a pre-established knowledge transfer network, the repair experience is transferred to other repair teams, especially those in similar technical fields or potentially facing similar problems. Receiving teams update their own capability scores based on the acquired knowledge, achieving skill complementarity and experience sharing among teams. Simultaneously, the knowledge-sharing module 14 also transmits data on the effectiveness of knowledge transfer, such as the improvement in the receiving teams' capabilities and the success rate of knowledge application, to the data analysis and feedback module, providing data support for continuous system optimization.

[0094] Step S5: The data analysis and feedback module 15 receives repair result data from the team scheduling decision module 13, including key indicators such as repair time, repair quality, resource consumption, and cost-effectiveness. Through in-depth analysis of this data, it identifies key factors affecting repair efficiency and dynamically adjusts the weights of each decision factor using an adaptive weight optimization algorithm. Simultaneously, the data analysis and feedback module 15 also receives knowledge transfer effect data from the knowledge sharing module 14, analyzes the effectiveness of different knowledge transfer paths, and optimizes transfer parameters and strategies. After optimization, new weight parameters are pushed to the team scheduling decision module 13 via short-cycle feedback (4 hours), and new transfer parameters are sent to the knowledge sharing module 14 via medium-cycle feedback (24 hours), guiding it to update the transfer paths of repair experience and ensuring continuous improvement of the knowledge sharing network. The team scheduling decision module 13 and the knowledge sharing module 14 focus on execution functions, respectively responsible for how to use parameters for decision-making and how to use parameters for transfer, without participating in parameter optimization calculations. This clear division of responsibilities and closed-loop feedback mechanism enables the entire scheduling system to improve after each task execution, gradually enhancing the overall performance and intelligence level of the system.

[0095] By implementing the above scheduling methods, the system can achieve truly intelligent scheduling of aircraft combat damage repair. It not only makes optimal decisions in a single mission, but more importantly, through continuous learning and optimization, it continuously improves the system's scheduling capabilities, ultimately achieving the goal of significantly improving the efficiency of aircraft combat damage repair.

[0096] Example 3

[0097] like Figure 3As shown, the decision weight dynamic evolution mechanism of this invention is a process of continuous adaptive adjustment over time and as the repair task progresses. Initially, all weights are 0.25. As the system learns and optimizes, the skill matching weight gradually rises to a higher level of 0.60, while the component availability weight drops to a lower level of 0.10 due to repeated repair delays. This dynamic adjustment forms an adaptive optimization loop, enabling the system to continuously optimize the decision-making process based on the actual repair results.

[0098] It should be noted that this invention can also be extended to various scenarios such as routine maintenance and repair of civil aircraft, battle damage repair of military vehicles and ships, emergency maintenance of large industrial equipment, emergency infrastructure repair after natural disasters, and predictive maintenance of equipment in intelligent manufacturing environments. By adjusting the corresponding parameter settings and knowledge base content, the system can adapt to the specific needs of different fields, demonstrating good versatility and scalability.

[0099] The above embodiments are only for illustrating the technical concept and features of the present invention, and are intended to enable those skilled in the art to understand the content of the present invention and implement it accordingly. They should not be construed as limiting the scope of protection of the present invention. All equivalent transformations or modifications made in accordance with the spirit and essence of the present invention should be covered within the scope of protection of the present invention.

Claims

1. An intelligent scheduling system for aircraft combat damage repair, characterized in that: It includes a damage assessment module, a maintenance resource management module, a team scheduling decision-making module, a knowledge sharing module, and a data analysis and feedback module; The damage assessment module is used to receive aircraft battle damage information and quantitatively assess the damage location, type and severity to obtain the damage severity. The maintenance resource management module assesses available repair teams, parts inventory, and equipment tools to obtain resource availability data, which includes skill matching, wait time, and parts availability. The team scheduling decision module obtains scheduling information from the damage assessment module, maintenance resource management module, and data analysis and feedback module, makes scheduling decisions based on the weights of multi-dimensional decision factors, transmits repair experience to the knowledge sharing module, and transmits repair result data to the data analysis and management module. The weights are dynamically adjusted according to the aircraft repair result feedback. The multi-dimensional decision-making factors include skill matching, waiting time, parts availability, and damage severity; The knowledge sharing module receives repair experience and transmission parameters, records and verifies the received repair experience using decentralized distributed ledger technology, transmits the knowledge to relevant teams, and transmits the knowledge transmission effect data to the data analysis and feedback module. The data analysis and feedback module collects repair result data and knowledge transfer effect data, optimizes decision weights and transfer parameters based on the obtained data, and feeds back to the corresponding modules. The specific steps of the team scheduling decision module in making scheduling decisions based on the weights of multi-dimensional decision factors include: S1. Obtain quantitative data on the location, type, and severity of aircraft damage from the damage assessment module, analyze the battle damage information, and obtain the damage severity. S2. Obtain information from the maintenance resource management module, check the skill level of available repair teams, current workload, parts inventory status and equipment and tool availability, and obtain resource availability data; S3. Based on damage severity and resource availability data, calculate multi-dimensional weighted decision values ​​using the weights of multi-dimensional decision factors; S4. Generate the optimal scheduling scheme based on the calculated multi-dimensional weight decision values, and select the team with the highest scheduling decision value and that meets the resource constraints as the execution team. The specific steps for the team scheduling decision module to dynamically adjust the weights include: S1. The team scheduling decision module initializes the weight values ​​of each decision factor; S2. Based on the completion status of each repair task, provide repair result data to the data analysis and feedback module; S3. The data analysis and feedback module evaluates the repair results and adjusts the weights. If the repair is successful and efficient, the weight of the corresponding success factor is increased; if the repair is delayed or inefficient, the weight of the corresponding influencing factor is decreased.

2. The intelligent scheduling system for aircraft combat damage repair according to claim 1, characterized in that: The calculation formula used by the team scheduling decision module is as follows: D = W1×S1 + W2×S2 + W3×S3 + W4×S4 Where D is the scheduling decision value, W1 to W4 are the corresponding decision factors, S1 to S4 are the standardized scores of the corresponding decision factors, and the value of S ranges from 0 to 1.

3. The intelligent scheduling system for aircraft combat damage repair according to claim 1, characterized in that: The knowledge-sharing module records and verifies the repair experience of each team, and the knowledge sharing specifically includes the following steps: S1. Establish professional field scores and non-professional field competence scores for each repair team; S2. When a team successfully repairs battle damage, it receives repair experience from the team scheduling decision module, records and verifies the received repair experience. S3. Transfer the repair experience to the relevant teams according to the preset knowledge transfer path; S4. Update the non-professional area capability score of the receiving team based on the repair experience received; S5. Based on the frequency and effectiveness of knowledge transfer between teams, dynamically optimize the knowledge transfer path by combining transfer parameters.

4. The intelligent scheduling system for aircraft combat damage repair according to claim 1, characterized in that: The data analysis and feedback module includes a data acquisition layer, a data processing layer, an algorithm optimization layer, and a closed-loop feedback application layer connected in sequence. The data acquisition layer is used to collect data, including repair results and knowledge transfer effect data. The data processing layer standardizes the raw data using a sliding time window strategy; The algorithm optimization layer includes an adaptive weight optimization algorithm and a knowledge sharing parameter optimization algorithm. The adaptive weight optimization algorithm updates the weights based on the repair results, and the knowledge sharing parameter optimization algorithm updates the transmission parameters based on the knowledge transmission effect. The closed-loop feedback application layer optimizes the entire system through three time scales: short-cycle feedback, medium-cycle feedback, and long-cycle feedback.

5. The intelligent scheduling system for aircraft combat damage repair according to claim 4, characterized in that: The data processing layer standardizes the raw data using a sliding time window strategy. The formula used for standardization is: β=(R×T 理想 ) / T 实际 Where β is the repair efficiency index, R is the repair quality score, and its value ranges from 0 to 1, T 理想 For ideal repair time, T 实际 This refers to the actual repair time.

6. The intelligent scheduling system for aircraft combat damage repair according to claim 5, characterized in that: The specific method for updating weights based on the repair results in the adaptive weight optimization algorithm is as follows: When β > 1, W i新 = W i旧 × (1 + 0.15 × (β - 1) × C i ); When β < 1, W i新 = W i旧 × (1 - 0.2 × (1 - β) × C i ) Where C i The contribution coefficient of this factor to the current repair task, with a value ranging from 0 to 1; the updated weight value W i新 Through formula W i归一化 =W i新 / ∑W j新 Normalization is performed, W i旧 These are the weight values ​​before the update.

7. The intelligent scheduling system for aircraft combat damage repair according to claim 1, characterized in that: When the knowledge-sharing module transfers knowledge to relevant teams, it calculates the knowledge increment. The formula used for calculating the knowledge increment is as follows: Z = K × Y × F Where Z represents the knowledge increment, K represents the transmission parameter, Y represents the professional level of the providing team, and F represents the severity of the damage; A time decay function is introduced in the knowledge transfer process. and efficiency improvement upper limit function , Where K0 is the initial repair experience transfer parameter, λ is the decay rate, n is the number of times the team receives repair experience, and t is the time interval since the repair experience was generated.

8. A scheduling method applied to an intelligent scheduling system for aircraft combat damage repair according to any one of claims 1-7, characterized in that, Includes the following steps: S1: The damage assessment module receives aircraft battle damage data and assesses the aircraft's battle damage status to obtain the severity of the damage. S2: The maintenance resource management module checks available repair resources and obtains resource availability data; S3: The team scheduling decision module calculates the scheduling decision value for each team based on damage severity and resource availability data, selects the team with the highest scheduling decision value to execute the repair task, and transmits the repair results to the data analysis module and the repair experience to the knowledge sharing module after the task is completed. S4: The knowledge sharing module receives repair experience, transmits the repair experience to relevant teams based on the received repair experience, updates the capability score of the receiving teams, and transmits the knowledge transmission effect data to the data analysis and feedback module. S5: The data analysis and feedback module receives the repair results and dynamically adjusts the decision weights; it also receives knowledge transfer effect data and optimizes the transfer parameters. The transmission parameters are sent to the knowledge sharing module, which then updates and repairs the experience transmission path based on the received transmission parameters.

Images