Correction Model of Pressure Sensor Based on Support Vector Machine

Abstract

The temperature and voltage fluctuation characteristics of pressure sensor was analyzed and found that the sensor output is nonlinear and easy to be affected by temperature and voltage fluctuation over a wide measuring range, a correction model of pressure sensor based on Support Vector Machine was presented. The approximate ability of the SVM to any nonlinear function was utilized to drill the correction model. so as to enable it to be setup at different temperatures and voltage fluctuation, thus allowing the sensor output can be in a nonlinear mapping relation to the voltage values the sensor actually sensed. The experimental results showed that the max comes down from 22.2% for 0.64%; the model can not only eliminate the influence of temperature fluctuation and voltage fluctuation but obtain the expected linear output from the output terminal of correction model.