基于MEA-BP的NTC热敏电阻温度传感器非线性补偿研究

Research on Nonlinear Compensation of NTC Thermistor Temperature Sensor Based on MEA-BP

热敏电阻作为温度传感器其阻温特性表现为非线性,在工程实际温度测量中存在一定的非线性误差,精度较低。论文使用NTC热敏电阻温度传感器作为研究对象,针对其存在的非线性问题,通过MEA(思维进化算法)优化BP神经网络模型从而实现对NTC热敏电阻温度传感器的非线性补偿,论文简要阐述了有关温度传感器补偿的相关方法与研究成果,分析了热敏电阻的阻温特性、工作原理,介绍了MEA-BP的模型构建,补偿原理与方法,利用5种评估标准对比传统RBF与BP神经网络模型。结果表明该补偿模型在各个评价指标上均优于传统RBF与BP神经网络,具有补偿精度更好、稳定性更强等优势,在测控、军工、航空等众多领域有一定实用价值。

As a commonly used temperature sensor,thermistor's resistance-temperature characteristics are nonlinear.Howev⁃er,its resistance to temperature is nonlinear,and there is a certain nonlinear error in engineering actual temperature measurement,and its accuracy is low.In this paper,NTC thermistor temperature sensor is used as the research object.In view of its nonlinear prob⁃lems,the nonlinear compensation of NTC thermistor temperature sensor is realized by optimizing BP neural network model through MEA(Mind Evolutionary Algorithm).The related methods and research results of temperature sensor compensation are briefly intro⁃duced,the resistance temperature characteristics and working principle of the thermistor are analyzed,the MEA-BP model con⁃struction,compensation principle and method are introduced,five evaluation criteria are used to compare the traditional RBF and BP neural network models.The results show that the compensation model is better than traditional RBF and BP neural network in each evaluation index,and has the advantages of better compensation accuracy and stronger stability,and has certain practical val⁃ue in many fields such as observe and control,military project and aviation.