论文部分内容阅读
本文提出了一种基于人工神经网络的铂电阻传感器非线性估计方法．该方法用二次幂级数多项式拟合温度传感器的非线性模型，多项式的系数可由神经网络学习算法得到．当条件发生变化时，只要给出几组测量数据对，通过该方法可自动重新训练网络，获得新的多项式系数，实现传感器的非线性估计．
This paper presents a non-linear estimation method of platinum resistance sensor based on artificial neural network. In this method, a non-linear model of the temperature sensor is fitted by a power-of-two polynomial. The coefficients of the polynomial can be obtained by a neural network learning algorithm. When the conditions change, as long as several pairs of measurement data are given, the network can be re-trained automatically and a new polynomial coefficient can be obtained to realize the nonlinear estimation of the sensor.