摘要
以大型直线振动筛侧帮裂纹为研究对象,利用小波分析对振动信号进行降噪和故障特征提取,设计了用于系统故障诊断的BP神经网络,并用遗传算法对网络结构和参数进行优化。通过样本训练、测试和在振动筛侧筛裂纹诊断中的应用,证明了这种小波遗传神经网络具有较高的故障识别能力、分类精度和速度。
Side crack of large-scale linear vibration screen served as the study object, the paper carried out noise reduction of vibration signals and extraction of fault characteristics by using wavelet analysis. It also de- signed BP neural network for system fault diagnosis, and optimized the network construction and parameters by using genetic algorithm. After sample training and test as well as application to diagnosis of side crack of vibra- tion screen, it showed that the wavelet genetic neural network had high fault recognition ability, classification precision and classification speed.
出处
《矿山机械》
北大核心
2012年第8期81-86,共6页
Mining & Processing Equipment
关键词
故障诊断
振动筛
小波分析
神经网络
遗传算法
fault diagnosis
vibration screen
wavelet analysis
neural network
genetic algorithm
