The Nonlinear Correction Model Research based on Wavelet Support Vector Regression Machine Sensor 

Ning TIAN, Yongsheng WAN 

School of Mechanical and Electrical Engineering, Southwest Petroleum University, Chengdu, CHINA 

Abstract: In view of the sensor due to factors such as aging, drift, temperature change caused by the sensor input and output of nonlinear problems, the support vector machine (SVM) based on wavelet kernel data correction, the nonlinear data correction method and constructed based on wavelet support vector machine (SVM) model for the nonlinear data correction. Through the Matlab simulation results show that the calibration results based on the wavelet support vector machine (SVM) is better than RBF and polynomial kernel correction, provides a very good method for sensor nonlinear correction, which has extensive promotion of significance. 

Keywords: Sensors; Wavelet kernel; Support vector regression machine; Nonlinear correction