Fault Diagnosis of Airborne EHA Based on GA-PSO Hybrid Optimization SVM
Aiming at the typical faults of airborne electro-hydrostatic actuator(EHA),the fault principle was analyzed and the simulation model was established in MATLAB/Simulink.To identify fault types efficiently,a fault diagnosis algorithm based on genetic algorithm(GA)and particle swarm optimization(PSO)hybrid optimization support vector machine(SVM)was proposed.GA has good robustness and global search ability,but its convergence speed is slow.PSO is insensitive to sample size and has memory function,but it's easy to fall into local optimization.Thus,the two algorithms were combined to find the optimal parameters of SVM.In addition,the traditional multi-classification methods of SVM'one to many'and'one to one'are easy to be inseparable,to solve the problem,a method with partial binary tree structure was proposed.For solving the collected original data were highly coincident,time domain feature statistics were introduced to improve the classification performance.The experimental results show the proposed GA-PSO has faster optimization speed and better parameters.At the same time,the classification effect of SVM optimized by this algorithm is better than other five commonly used machine learning models,and the accuracy of fault identification is 97.7％.
airborne EHAgenetic algorithmparticle swarm optimizationpartial binary tree structuremulti-class SVM
万方数据
维普
