FastICA遗传神经网络算法

针对反向传播（BP）算法和基于负熵固定点迭代快速独立分量分析（FastICA）方法各自的优缺点，提出了FastICA遗传神经网络算法，对滚动轴承进行故障识别.首先对信号进行FastICA分离，得到振动信号故障信息的独立分量，每个独立分量对应着相应的能量，将各个独立分量的能量构成特征向量；其次利用遗传算法对BP神经网络的初始权值和阈值进行优化，得到遗传神经网络；最后将特征向量作为遗传神经网络的输入样本进行故障识别.利用该方法对滚动轴承多类故障信号进行识别，提高了故障识别能力.

FastICA Genetic Neural Networks Method

Depending on the intrinsic weakness and advantages of back propagation(BP) neural network and Fast Independent Component Analysis(FastICA), a Fast Independent Component Analysis(FastICA) Genetic Neural Networks Method was proposed for fault characteristic signal recognition. The FastICA is used to decompose signals to obtain the independent components successively, each of Independent components corresponding to an energy band, and feature vector of each energy band is used as input sample to optimize neural network. Secondly, the genetic algorithm is used to optimize the weights and thresholds of BP neural network to obtain the Genetic Neural Network. Thirdly,the feature vector is used as input sample of the genetic neural network to identify the fault. Using this method can analysis and identify many kinds of rolling bearings fault signal, and through this method the ability of fault identification isimproved.