双轴分排涡扇发动机气路故障诊断测量参数选择方法

针对地面台架试车气路故障诊断测量参数的优化选择提出了一种四步优化方法.该方法包括测量参数敏感性分析、部件性能参数相关性分析、影响系数矩阵条件数分析和遗传算法检验.第1步选择出了总空气流量;第2步选择出了风扇出口总温和压气机出口总温;第3步选择出了12种对故障诊断有利的测量参数组合;第4步选择出了最有利于故障隔离的测量参数组合.利用遗传算法进行单一故障辨识的结果表明:所选的12种测量参数组合的适应度都大于0.85,接近于最优的适应度1.对于最有利的测量参数组合,利用遗传算法的故障辨识验证结果表明,其适应度均大于0.9,验证了四步优化方法的有效性.

Measurement parameters selection method for gas path fault diagnosis of two-shaft split flow turbofan engine

Aiming at measurement parameters optimal selection for gas path fault diagnosis in ground test bed, a four-step optimal method was presented. The four-step optimal method includes measurement parameters sensitivity analysis, component performance parameters correlation analysis, influence coefficient matrix condition number analysis and genetic algorithm validation. According to the first step, total air mass flow was picked out. In the second step, total temperature at fan exit and total temperature at compressor exit were picked out. In the third step, twelve measurement parameters combinations beneficial to fault diagnosis were picked out. In the final step, the best measurement parameters combination was obtained. Based on the genetic algorithm, the simulated diagnostic results show that, with these twelve measurement parameters combinations, all the fitness values for each single fault diagnosis are more than 0.85, which approximate the optimal fitness value 1. As to the most promising measurement parameter combination, all the fitness values are greater than 0.9 by genetic algorithm validation, which demonstrate the validity of this four-step optimal method.