文章描述了一種用反函數(shù)校正傳感器非線性誤差的方法,闡述了校正原理,提出了利用
BP 神經(jīng)網(wǎng)絡(luò)和遺傳算法相結(jié)合,擬合傳感器傳輸特性反函數(shù)的算法,該算法可將傳感器傳輸特性的非線性模型,改造成為與實際物理過程相一致的不失真的線性模型,給出了一個應(yīng)用實例,其結(jié)果表明,可使傳感器的非線性誤差有較大的減少。
關(guān)鍵詞:傳感器;非線性誤差;反函數(shù);神經(jīng)網(wǎng)絡(luò);遺傳算法
Abstract : This paper describe a method that the nonlinear error of the sensor is corrected using a inverse function. The correction principle is expounded. A neural network of using genetic algorithms is showed , the network can close with sensor input and output essence , and the nonlinear model sensor can be retrofitted into a non2distortion linear model that is
consistent with the actual physical process. Finally , a applied example is introduced , the experimental results show that sensor nonlinear errors are reduce more than tenfold.
Key words : sensor ;linear error ;inverse function ;neural network ;genetic algorithms