摘要
传感器或仪表的非线性特性是影响其测量精度的重要指标之一。文章采用支持向量机(SVM)的方法进行智能差压变送器非线性校正,选取RBF核函数,由于它数值限制条件少、模型相对简单,因此利用它可以隐式地将训练数据映射到高维空间,实现非线性变换中的线性分类,得到用于函数拟合模型的SVM。实验数据仿真表明,算法取得了较好的效果,非线性度达到±0.03%F.S,为差压变送器的非线性校正提供了一种新的方法。
The nonlinear character is the main index which affects measuring accuracy.The research of correct nonlinear errors making use of support vector machine for intelligent differential pressure transmitter is introduced in this paper.RBF is chosen as kernel function,which has fewer numerical constraint conditions and comparatively simple model.Thus,we can use the function to map the training data into the high dimension space,realize the linear classification of the nonlinear transform,gets the SVM for function modeling.In the end,through simulating the experimental data of transmitter,the numerical results show that the nonlinearity of ±0.03%,and thus a realistic new technique is proved for the non-linear compensating of differential pressure transmitter.
出处
《仪表技术》
2011年第6期1-3,6,共4页
Instrumentation Technology
关键词
差压变送器
非线性校正
支持向量机
differential pressure transmitter
non-linearity correction
support vector machine（SVM）
