摘要
针对定性概念与定量表示之间不确定性转化的模糊性和随机性问题,并为避免小波网络陷入局部极小、过拟合现象,将云模型、遗传算法和小波神经网络相结合,不但解决了定性定量之间的完好转换,而且通过遗传算法的全局优化搜索得到了网络的最优参数。仿真实验验证了这种新方法对于空气增压机故障诊断的有效性。
Aiming at the fuzziness and stochastic problems of the uncertainty transformation between the qualitative concept and the quantitative representation, to avoid the wavelet network falling into local minimum or over fitting, combine the cloud model ,genetic algorithm with wavelet neural network not only realize the transformation between the qualitative and quantitative problems perfectly, the global optimization search of genetic algorithm also get ＂the optimal network parameter. The simulation experiment shows that the new method is effective on the fault diagnosis of air compressor.
出处
《电子技术应用》
北大核心
2012年第3期139-141,148,共4页
Application of Electronic Technique
基金
国家自然基金项目(61102039)
关键词
小波神经网络
遗传算法
云模型
故障诊断
wavelet neural network
genetic algorithm
cloud model
fault diagnosis
