基于支持向量机的电子鼻参数补偿的研究

摘要

简要介绍采用HART协议的电子鼻系统和其电子元件,利用连续相位的FSK调制和解调技术使数据传输受到最小扰动. 着重说明支持向量机算法适合在小样本数据情况,其核函数可以向高维空间非线性映射的优点.对系统前端传感器构建其非线性回归预测的数学模型, 实现在软件上对参数补偿, 提高了电子鼻系统的精度, 其拟合度R=99.777 7%,方差MSE=9.354 13e-05.%First of all. This paper briefly introduces electric sensor system use HART protocol and other electric hardware parts. On For the purpose of minimal interference data transmission on Electric Sensor System use continuous phase FSK modulation and demodulation techniques,.Secondly, the use of the advantages of small data and kernel function to the higher space nonlinear map on support vector machine regression algorithm , To predict a mathematical nonlinear model to fit the front sensor of the Electric Sensor System, implement the compensation of parameters in software part. By combining software and hardware methods to improve the accuracy of the Electronic Sensor System.Its fit R=99.777 7%,MSE=9.354 13e-05.