摘要
针对传统的PSO优化BP网络的局限性,提出了一种混沌PSO-DV算法和BP神经网络的混合算法.该算法具有混沌算法的局部搜索遍历性,DE算法的种群多样性及BP神经网络的快速搜索能力等优势.仿真结果表明,混沌PSO-DV优化的BP神经网络应用于汽车发动机故障诊断,使得故障诊断的效率和准确率得到了很大的提高.
Aimed at the limitation of BP neural network optimazed by traditional PSO, a novel chaos PSO-DV algorithm was proposed. This algorithm has the advantages of the local search er- godicity of chaos algorithm, the population diversity of the DE algorithm and fast search ability of BP neural network. The simulation results showed that when BP neural network which optimized by chaos PSO-DV was applied on the automotive engine fault diagnosis, the efficiency and accura-cy of fault diagnosis was higher improved.
出处
《山东理工大学学报（自然科学版）》
CAS
2013年第1期67-70,共4页
Journal of Shandong University of Technology:Natural Science Edition
