基于ANFIS的温度传感器非线性校正方法

Nonlinear Calibration for Temperature Sensors Based on Adaptive Neural-fuzzy Inference System (ANFIS)

介绍了用神经网络进行传感器非线性误差校正的原理与方法,分析了自适应神经模糊推理系统(ANFIS)的基本原理。通过模糊聚类和混合学习算法,ANFIS可以逼近高阶输入输出非线性系统,将该算法用于两个典型非线性系统建模,均能获得满意结果。之后,将ANFIS算法用于温度传感器非线性校正中,试验结果表明该方法与基于CMAC网络和BP网络的校正方法相比,校正的精度高于以上两种校正方法。

The principle and the method for correcting the nonlinear errors of the sensor system with neural networks are introduced. The basic principle of adaptive neural-fuzzy inference system (ANFIS) is presented. By using fuzzy clustering and hybrid learning procedures, the ANFIS can construct the highly nonlinear input-output mapping. Then, the ANFIS is used to model two nonlinear systems, both yield remarkable results. Finally, the ANFIS is used in temperature sensor nonlinear calibration system. Experimental results show that the nonlinear calibration system based on ANFIS has higher precision than the nonlinear calibration methods based on CMAC or BP.