自适应变异混沌BA多中心智能故障识别策略  

Multiple Center Intelligent Fault Recognition Strategy Based on Adaptive Mutation Chaos BA

作　　者：徐岩[1] XU Yan(Branch of Jilin Normal University,Jilin Siping 136000,China)

机构地区：[1]吉林师范大学分院,吉林四平136000

出　　处：《机械设计与制造》2024年第12期33-39,共7页Machinery Design & Manufacture

基　　金：全国教育科学“十三五”教育部规划课题(QGCZ1819)。

摘　　要：这里用一种基于自适应变异混沌蝙蝠算法优化的多中心优先级识别方法通过多个中心识别具有非线性分离的数据集解决了单一中心策略无法识别非线性分离数据的问题。首先,将优先级引入到多目标优化中,包括识别精度,优化中心的数量和距离关系。根据各种数据的特征,调整优先级,以确保自适应优化中心的数量并保持原始精度,并且通过自适应变异混沌蝙蝠算法进行优化。仿真结果表明,所提出的策略对数据的不同分布特征更广泛的识别能力,对于非线性分离数据具备良好的识别性能。We propose an adaptive mutation algorithm based on chaos to solve the problem of multiple center identification.Firstly,priority is introduced into multiple objective optimization,including recognition accuracy,number of optimization centers and distance relationship.According to the characteristics of various data,the priority is adjusted to ensure the number of adaptive optimization centers and maintain the original accuracy,and the adaptive mutation chaotic bat algorithm is used for optimization.The simulation results show that the proposed strategy has more extensive recognition ability for different distribution characteristics of data,and has good recognition performance for nonlinear separation data.

关 键 词：自适应 优先级 智能识别 多中心 蝙蝠算法 

分 类 号：TH16[机械工程—机械制造及自动化] TP301.6[自动化与计算机技术—计算机系统结构]