摘要
根据当前结冰传感器非线性校正存在的问题,提出了利用BP网络建立传感器逆模型的校正方法。文中采用功能强大的MATLAB工具软件,对神经网络进行训练,获得权值、阈值。实际应用结果表明,该方法简单、实用,大大方便了产品性能一致性不高的结冰传感器在测控系统中的应用。
This paper presents a compensation method using the BP network to establish a reversable model for an ice-detectors so as to solve the problems existing in present non-linearity compensation.We use the powerful MATLAB to train the neural network in order to get the weights and thresholds.It has shown in practical applications that this method is not only correct and practical,but also very convenient for the ice-detector with low consistence to be applied to measurement and control systems.
出处
《工业仪表与自动化装置》
2005年第2期49-50,17,共3页
Industrial Instrumentation & Automation
关键词
BP网络
非线性校正
结冰传感器
BP network
non-linearity compensation
ice-detector
