基于BP神经网络的超声波测距非线性误差校正  

Nonlinearerrors correction of ultrasonicwave measure distance system based on neuralnetwork

作　　者：刘建生[1] 杨丽荣[2] 程铁栋[1] 

机构地区：[1]江西理工大学理学院 [2]江西理工大学机电学院

出　　处：《微计算机信息》2007年第01S期187-188,20,共3页Control & Automation

摘　　要：本文结合笔者所设计的超声波测距仪,介绍了用神经网络校正超声波传感器的非线性误差的原理与方法,并提出了基于BP神经网络的超声波测距非线性误差校正的模型、算法及其硬件实现。通过理论分析和硬件实验,显示出BP神经网络对超声波传感器的温度补偿和非线性校正的效果良好,充分表明了应用神经网络在提高超声波测距精度方面是一种行之有效的方法。The paper introduce the principle and the methods for correcting the nonlinear errors of the ultrasonic wave sensor with a neural network, combining the designed system of ultrasonic wave measuring distance.Nonlinear errors correction of the system based on BP neural network, which includes its model, algorithm and realized techniques on hardware are presented. The outcome of theory analysis and experiment results shows that this method is very effectual.

关 键 词：超声波测距 BP神经网络 非线性误差 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]