基于支持向量回归的非线性多功能传感器信号重构

在多功能传感器信号重构中,通常采用经验风险最小化准则实现函数回归,在小样本情况下,该方法易导致泛化性差和过拟合问题.本文利用支持向量回归方法实现非线性多功能传感器信号重构,支持向量机是基于结构风险最小化准则的新型机器学习方法,可有效抑制过拟合问题并改善泛化性能.仿真结果表明经该算法重构后的信号重构误差率在0.4%以下,重构效果较好,验证了该算法的有效性.

Nonlinear Multifunctional Sensor Signal Reconstruction based on Support Vector Regression

The ordinary empirical risk minimization method is often used to estimate the regression function in multifunctional sensor signal reconstruction. If the size of sample data is small, this method will lead to the problem of overfitting and poor generalization capability. This paper applied support vector regression (SVR) method to nonlinear multifunctional sensor signal reconstruction. Support vector machine (SVM) is a novel machine learning method based on structural risk minimization, and it can restrain overfitting and improve generalization capability. The emulation result shows that the ratio of signal reconstruction error is less than 0.4% and verify the feasibility of this algorithm.