Research On Intelligent Evaluation Of Equipment Capability Based On Key Elements Sensitivity Of Index System

Equipment capability evaluation plays a vital role in the development and construction of equipment.The combat aircraft is an important part of the weapon equipment system.By evaluating the capabilities of combat aircraft,it is of practical significance to seek the relationship between key elements of index and equipment capabilities,and study the intelligent evaluation method of equipment capabilities based on the equipment index system in operation.Based on the construction of the combat capability index system of combat aircraft,this paper proposes an integrated learning algorithm for strategy optimization,which is used to evaluate the combat capability of combat aircraft,and solves the problem of neural network over-fitting and global optimization.Finally,we use combat aircraft air,ground,and total combat capability data to train the network,and evaluate the accuracy,stability,and rapidity of the algorithm through combat capability estimates.At the same time,through the index sensitivity calculation,the sensitivity analysis of the impact of key indicators of equipment capability on equipment performance can be achieved.The main work of this paper is as follows:1.Introduce the basic theories and concepts in equipment capability evaluation,summarize the research status of capability evaluation technology,and expound the contents and basic steps of capability evaluation.According to the research project,the index system of capability evaluation of typical equipment is established by analyzing the operational use of typical equipment system and the principle of index selection.In order to provide data for the evaluation of combat aircraft capability,build a global index system,establish a database based on equipment indicators and combat capabilities,and provide data for combat aircraft capability evaluation.It is of great scientific significance and application value to study the general method of equipment evaluation,and to provide support for equipment demonstration and equipment capability evaluation during the whole process from operation to combat.2.Aiming at the problem of small sample equipment capability evaluation,according to the theory of Grey system index correlation analysis,the index correlation degree is calculated and the equipment capability is evaluated to seek the relationship between the key elements of the index and the equipment capability.According to the space characteristics of the combat aircraft index system,quantify the index system space index,construct the equipment capability index matrix,standardize the index matrix,select the optimal mode vector,calculate the index correlation coefficient and information entropy,and calculate the combat capability index weight of the combat aircraft objectively.At the same time,through the mapping relationship between the optimal mode correlation degree of combat aircraft and the database of air combat capability,the correlation degree-combat capability curve is fitted to estimate the combat aircraft capability.The effectiveness and feasibility of the method are verified by the capability evaluation of typical equipment of combat aircraft.3.According to the characteristics of the relationship between the equipment index system and equipment capability,the search strategy of BP neural network is optimized on the basis of neural network theory,the gradient strategy is used to solve the problem of parameter optimization and convergence acceleration in the local optimum and global minimum problem of neural network algorithm,and the anti-overfitting method of neural network is found by the strategy of early stopping and regularization,solving the problem of singular point optimization in estimation.Finally,based on the research of intelligent algorithm,with the support of equipment index and combat capability library,the combat aircraft capability evaluation model is established.The effectiveness and feasibility of the method are verified by the capability evaluation of typical equipment of combat aircraft.4.According to the characteristics of the relationship between the equipment index system and equipment capability,based on the BP neural network model,the methods of proportional selection operator,single-point crossover operator and single-point mutation operator are used to compress and reduce the dimension in the index space of the index system,the key indexes are extracted,and the evaluation model of the reduced dimension network is constructed.The method is proved to be fast by the capability evaluation of typical equipment and the record of the evaluation time.5.Aiming at the requirement of accuracy and stability for equipment capability evaluation,the paper starts with searching strategy and gradient strategy based on neural network,an intelligent evaluation model of equipment capability based on neural network strategy optimization ensemble learning(ELPONN)is established to solve the problems of poor generalization performance and local minima.The accuracy and stability of the operational capability evaluation of the model are solved by iteratively verifying the model through the operational aircraft capability estimation and data validation.At the same time,the sensitivity of the indexes in the integrated learning network is calculated,the sensitivity of the key elements in the equipment capability is analyzed,the analytical solution of the indicators is explored,and the absolute sensitivity and relative sensitivity of the indexes are calculated and analyzed,then the correlation model between sensitivity of equipment index and equipment capability is established.The experimental results show that the average relative error of ELPONN method is 4.10%,which is superior to the improved neural network method and has better accuracy and stability.In summary,this paper analyzes the problems faced in the field of intelligent evaluation of equipment capability and the correlation analysis between the sensitivity of key elements of indicators and equipment capability,and also studies the related algorithm theory.Based on neural network and integrated network model,an equipment capability evaluation model based on neural network strategy optimization integrated learning is constructed.At the same time,the sensitivity of key elements is analyzed and some research results are obtained.The related achievements in this paper provide the theoretical basis and practical method for the intelligent evaluation and the sensitivity of key index to equipment capability.