基于神经网络的巨磁阻位移测量系统研究  

Research on GMR Measurement System Based on Neural Network

作　　者：刘日 董全林[1] 朱天宇 ZHAO Dong-qing;CHENG Hong-tao;CHU Cheng-qun;YUAN Xiao-kang(Science and Technology on Electronic Test and Measurement Laboratory,North University of China;)

机构地区：[1]北京航空航天大学仪器科学与光电工程学院微纳测控与低维物理教育部重点实验室,北京100191

出　　处：《仪表技术与传感器》2018年第6期53-57,共5页Instrument Technique and Sensor

基　　金：国家科技支撑计划资助项目(2006BAK03A24)

摘　　要：针对工业过程中直线行程的高精度测量问题,提出了基于神经网络函数逼近原理,以巨磁阻(GMR)传感器为核心的非接触式测量系统。运用等效电流法对提供磁场的圆柱形永磁体进行建模并仿真,得到永磁体磁场在空间中的分布规律,从而确定磁体的位置及形状参数。通过Matlab建立BP神经网络,用遗传算法进行优化,最终获取了位移与电压的函数关系,研究结果表明:该测量系统可以精确地测量直线位移。A non-contact measurement system based on the principle of neural network function approximation and GMR sensor is proposed for the linear travel measurement in industrial process.The regular current method was used to model and simulate the cylindrical permanent magnet that provided the magnetic field,obtaining the distribution law of the magnetic field in space,determining the position and shape parameters of the magnet.The BP neural network was established by Matlab and optimized by genetic algorithm. Last,the relationship between displacement and voltage was obtained. The result shows that the system can measure the linear displacement accurately.

关 键 词：神经网络 函数逼近 非接触式 巨磁阻 磁场 直线位移 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程] TP212[自动化与计算机技术—控制科学与工程]