摘要
盲源分离是在对源信号和传输通道几乎没有可利用信息的情况下,仅从观测到的混合信号中提取或恢复出源信号的信号处理方法。采用改进型算法作为核心学习算法计算分离矩阵W,即基于负熵的快速独立成分分析算法;对该种算法进行编程及仿真,并将现场采集的数据利用盲源分离算法进行实验室分离实验,最后将该分离算法植入到故障诊断系统中,投入现场使用,验证其可行性。
This study uses the modified algorithms as a core learning algorithm to calculate the separation matrix W,and analysis algorithm based on negative entropy fast independent component.The programming and Simulation of the algorithms,and sends the collected data by using blind source separation algorithm for laboratory experiment,finally the separation algorithm into the fault diagnosis system,put into use,verify their quality.
出处
《工业控制计算机》
2017年第6期100-102,共3页
Industrial Control Computer
