论文部分内容阅读
依据压力传感器样本,提出了一种采用最小二乘支持向量机(LS-SVM)辨识传感器逆模特征的校正压力传感器非线性误差的方法,该方法将实测数据由径向基函数把非线性逼近问题转化为线性逼近问题,不需逆模型函数形式的先验知识,能够保证得到的极值解就是局最优解,具有较好的泛化能力。实验结果表明,采用该方法校正后的传感器的检测精度可达到1%,效果令人满意。
According to the pressure sensor sample, a method of correcting the nonlinearity error of the pressure sensor using LS-SVM to identify the inverse model of the sensor is proposed. The measured data is obtained by radial basis function and nonlinear approximation The problem is transformed into the problem of linear approximation without the prior knowledge in the form of an inverse model function, which guarantees that the extreme solution obtained is the optimal solution of the bureau and has better generalization ability. The experimental results show that the detection accuracy of the sensor calibrated with this method can reach 1%, and the result is satisfactory.