旋转设备声学故障特征提取与优化方法

Acoustical Fault Feature Extraction and Optimization of Rotating Machinery

针对旋转设备原始故障特征空间中存在的冗余特征问题,提出一种基于支持向量数据描述(SVDD)和遗传算法的故障特征优化方法.通过理论和实验分析构造了相对完备的设备声学故障特征空间;依据特征可分离性评价准则和SVDD识别率从原始故障样本数据集中提取出先验知识,指导种群的初始化;以类内-类间距离判据和故障分类器的识别率评价种群中个体的适应度,在此基础上建立改进的遗传算法搜索最优故障特征子集.基于转子振动台所模拟的不平衡故障实验样本数据集,验证了该方法的有效性.

To eliminate redundant features in original fault feature space, a novel feature selection algorithm based on support vector data description （SVDD） and modified genetic algorithm is proposed. Firstly, it constructs a complete acoustical fault feature space through theoretic analysis and experiments. According to established criterion of feature separation and SVDD classifier prediction accuracy, prior knowledge is extracted from training data set and used as initialization to improve the efficiency of genetic algorithm. Inner and intra-class distance criterion and classifier prediction accuracy arc introduced to establish fitness function and evaluate the degree of importance of every gene, thus the optimized feature subset is obtained. Experiments with unbalance-fault data set simulated on rotor vibration test-bed show that the proposed algorithm can improve the diagnosis accuracy.