航空发动机在使用寿命周期内会不断磨损最终出现故障,通过对发动机油液监测铁谱分析数据的挖掘可实现磨损故障的诊断。本文研究免疫算法优化的支持向量机(SVM)在航空发动机磨损故障诊断中的运用。首先,总结了支持向量机和免疫算法的运行流程和关键算法。然后,用改进的免疫算法优化支持向量机惩罚因子、松弛变量及核函数参数。某型航空发动机的油液铁谱分析数据和加入噪声数据验证结果表明,该方法可有效实现航空发动机磨损故障诊断且具有较好的鲁棒性。最后,研究了核函数、多分类决策方法、初始种群大小、亲和力计算公式、支持向量机优化方法和归一化方法对磨损故障诊断准确率的影响,得到了最佳诊断方法。 Aeroengine will eventually wear and tear during the life cycle of the final failure, through the engine oil analysis of iron spectrum analysis data mining fault diagnosis can be realized. This paper studies the application of immune algorithm-optimized support vector machine (SVM) in fault diagnosis of aeroengine wear. First, the operating flow and key algorithms of support vector machine and immune algorithm are summarized. Then, the improved immune algorithm is used to optimize SVM penalty factor, slack variable and kernel function parameters. Aeroengine oil ferrographic analysis data and added noise data validation results show that this method can effectively achieve the aero-engine wear fault diagnosis and has good robustness. Finally, the effects of kernel function, multi-classification decision-making method, initial population size, affinity calculation formula, support vector machine optimization method and normalization method on the accuracy of wear fault diagnosis are studied, and the best diagnosis method is obtained.