摘要
根据“逆模型”的思想,利用神经元网络良好的逼近能力,提出了基于函数联接型神经网络的传感器动态补偿方法。该方法设计的动态补偿器实现简单,实时性好;不依赖于传感器的模型,鲁棒性强;可以优化补偿器的参数。该方法的补偿效果比零极点配置方法的好,是一种非常有效的新方法。
According to the idea of inverse model,a sensors dynamic copmensating method based on functional link artificial neural network(FLANN) is proposed in this paper by using the excellent approaching ability of artificial neural network.Dynamic compensating device designed by this method is of easy realization and good real time.It is independent for the sensors model,and of strong robust.This method can optimum the parameters of the compensating device.Its compensating results are better than that of the pole zero locating method.It is a effective and new method and is well worth extension and application.
出处
《仪器仪表学报》
EI
CAS
CSCD
北大核心
1999年第5期541-544,共4页
Chinese Journal of Scientific Instrument
基金
国家自然科学基金
八六三计划 5 12主题
机械工业部高校跨世纪优秀人才专项基金
关键词
传感器
神经网络
动态补偿
机器人
腕力传感器
Sensor Functional link artificial neural network Dynamic compensation
