压力传感器的支持向量机非线性回归建模

摘要

The output characteristics of pressure sensor affected by environmental factors,especially temperature changes. In order to solve this problem,a nonlinear regression model is presented based on support vector machine (SVM). The approximate ability of the SVM to nonlinear function is utilized to build the nonlinear regression correction model. A nonlinear mapping relation between sensor output and the actual voltage values is established to achieve the correction of pressure sensor. The experimental results show that the model can decrease the correction of the temperature changes effectively, while both the measurement precision and temperature stability are improved.%压力传感器的输出特性易受到环境因素,尤其是温度变化的影响.针对该问题,提出了利用支持向量机( SVM)对压力传感器输出特性进行非线性补偿的校正模型.校正模型利用SVM的回归算法来逼近非线性函数的特点,通过建立压力传感器输出特性与其实际电压值之间非线性映射关系的校正模型来实现压力传感器的校正.实例表明:该方法能有效地减少温度变化对传感器输出的影响,且校正后的压力传感器具有更高的测量精度和温度稳定性.