基于径向基函数神经网络的压电式六维力传感器解耦算法

摘要

针对四点支撑结构的压电式六维力传感器线性度差,维间耦合严重的问题,提出了基于径向基函数(RBF)神经网络的解耦算法.分析了耦合产生的主要原因,建立了RBF神经网络模型.通过对六维力传感器进行标定实验获取解耦所需的实验数据,并对实验数据进行处理. 然后采用RBF神经网络优化传感器输出系统的多维非线性解耦算法,解耦出传感器的输入输出映射关系,得到解耦后的传感器输出数据.对传感器解耦后的数据分析表明:采用RBF神经网络的解耦算法得到的最大Ⅰ类误差和Ⅱ类误差分别为1.29%、1.56%.结果显示:采用RBF神经网络的解耦算法,能够更加有效地减小传感器的Ⅰ类误差和Ⅱ类误差,满足了传感器两类误差指标均低于2%的要求.该算法有效地提高了传感器的测量精度,基本解决了传感器解耦困难的难题.%For problems of poor linearity and too many inter-dimensional coupling errors of a four-point supporting piezoelectric six-dimensional force sensor, the decoupling algorithms based on Redial Basis Function (RBF) neural network were proposed.Main factors to produce coupling errors were analyzed and the RBF neural network was established.The six-dimensional force sensor was calibrated experimentally to obtain experimental data for decoupling, and the data were processed by the nonlinear decoupling algorithm based on RBF neural network.Then the mapping relation between input and output was acquired by decoupling and the decoupled data from the sensor was obtained.These data were analyzed, and the result shows that the biggest classⅠerror and classⅡerror by the proposed nonlinear decoupling algorithm based on RBF neural network are 1.29% and 1.56% respectively.The experimental analysis shows that it will effectively reduce the classⅠerrors and the classⅡerrors through nonlinear decoupling algorithm based on RBF neural network, and meets the requirements that the two kinds of error indicators of the sensor should be less than 2%.The proposed algorithm improves the measuring accuracy of sensors and overcomes the difficulty on decoupling.