基于改进的KELM轴承故障诊断算法  

Bearing Fault Diagnosis Algorithm Based on Improved KELM

作　　者：张成[1] 李朝阳 

机构地区：[1]沈阳化工大学理学院,辽宁沈阳110142

出　　处：《工业控制计算机》2024年第9期59-61,共3页Industrial Control Computer

基　　金：辽宁省教育厅基本科研项目(LJKMZ20220792)。

摘　　要：提出了一种使用核主成分分析(KPCA)、改进的麻雀搜索算法(OCSSA)和核极限学习机(KELM)相结合的算法来解决工厂化工过程中数据非高斯性和数据耦合性强的故障检测问题。首先使用KPCA算法降维,然后使用OCSSA算法寻找KELM中核参数γ和正则化系数C的最优取值。其中对SSA算法的改进如下:通过引入混沌映射技术增加SSA算法的种群丰富性,采用鱼鹰优化算法在第一阶段的全局勘探策略来替换SSA算法的发现者位置更新公式,利用柯西变异策略来替换SSA算法的跟随者位置更新公式。最终建立了一种KPCA、OCSSA和KELM三种算法相结合的轴承故障诊断分类算法。实验结果显示,经过OCSSA优化后,该算法在解决轴承故障时表现出了较高的准确性和有效性。In this paper,an algorithm combining kernel principal component analysis(KPCA),osprey cauchy sparrow search algorithm(OCSSA)and kernel extreme learning machine(KELM)is proposed to solve the problem of non-Gaussian data and strong data coupling in the process of chemical industry.Firstly,the KPCA algorithm is employed for dimensionality reduction,followed by the utilization of the OCSSA algorithm to determine the optimal values for kernel parameterγand regularization coefficient C in KELM.Enhancements made to the sparrow algorithm include:augmenting population diversity through incorporating chaotic mapping technology,substituting the original sparrow algorithm's finder position update formula with a global exploration strategy derived from Osprey optimization,and replacing the follower position update formula of the original sparrow algorithm with Cauchy variation strategy.Ultimately,a bearing fault diagnosis and classification algorithm that integrates KPCA,OCSSA,and KELM algorithms is established.Experimental results demonstrate that after undergoing OCSSA optimization,this approach exhibits remarkable accuracy and effectiveness in addressing bearing faults.

关 键 词：故障诊断 核主成分分析 麻雀算法 核极限学习机 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程] TQ050.7[自动化与计算机技术—控制科学与工程]