基于小波网络的非线性传感特性的线性化

针对最小二乘回归、几何逼近、分段线性化和神经网络等方法的不足 ,提出了实现非线性传感特性线性化的小波网络方法 ,给出了具体的算法和计算公式以及自适应确定小波基数目的方法 ,该方法可大大减小非线性误差 .应用实验 ,验证了该方法的有效性

Linearization of sensor''''s nonlinearity based on wavelet networks

A method on how to realize the linearization of the sensor's nonlinearity based on wavelet is put forward in this paper for the shortcomings existing in the conventional methods such as least square regression,geometric approximation, segmental linearization and neural network etc. The algorithm, the formulae, and the self-adaptive method on deciding the numbers of wavelet radices are also presented in this paper. The nonlinear errors can be greatly reduced by employing this method. Experimental results show that this method is quite effective.