改进PSO优化SVM的滚动轴承故障诊断

提出一种基于改进粒子群算法和支持向量机的滚动轴承故障诊断方法.首先分析基本粒子群算法的不足及其关键参数,提出多方面改进的粒子群算法,利用10种基准测试函数对比多种粒子群算法,证明该改进算法的优势.然后结合支持向量机,建立滚动轴承故障诊断模型,并提取滚动轴承振动信号的时域、频域、小波包节点能量和CEEMDAN分量排列熵四种特征,构成单一特征和组合特征作为诊断模型的输入特征向量.最后利用凯斯西储大学滚动轴承数据进行验证,并与网格算法、遗传算法和多种不同粒子群算法进行对比.试验证明,本改进粒子群算法优化支持向量机模型在滚动轴承故障诊断中更具优势.

Fault diagnosis of rolling bearing based on improved PSO optimized SVM

Rolling bearing is one of the most common and easily damaged parts in mechanical equipment. Therefore, a fault diagnosis method based on Improved Particle Swarm Optimization and support vector machine is proposed. Firstly, the shortcomings and key parameters of the basic particle swarm optimization (PSO) algorithm are analyzed, and an improved PSO algorithm is proposed. Ten benchmark functions are used to compare the advantages of the improved PSO algorithm. Then, a fault diagnosis model of rolling bearings is established based on support vector machine. Four features of rolling bearing vibration signal are extracted, including time domain, frequency domain, wavelet packet node energy and CEEMDAN component permutation entropy. A single feature and a combination feature are constituted as input eigenvectors of the diagnosis model. Finally, the rolling bearing data of Case Western Reserve University are used to validate the grid algorithm, genetic algorithm and a variety of different particle swarm optimization algorithms. Experiments show that the improved particle swarm optimization support vector machine model has more advantages in rolling bearing fault diagnosis.