改进BP网络及其在传感器非线性校正中的应用

AN IMPROVED BP NEURAL NETWORK AND ITS APPLICATION IN NONLINEAR CORRECTION OF SENSOR

针对传统BP网络收敛速度慢、容易陷入局部最小点等问题,采用附加动量因子和自适应学习速率进行了改进,并将其用于对传感器的非线性误差进行补偿。用MATLAB语言编制相应的训练程序,仿真结果表明,在相同的条件下,改进后算法节省了大量的训练时间,同时提高了数据拟合的精度。

An improved BP algorithm with additional momentum and self-adaptive learning rate was presented, which can overcome the disadvantages of traditional back propagation （BP） artificial neural network （ANN）, such as slow convergence and easily falling into local minimum. It has been used in the compensation of nonlinear errors of the sensor. The Matlab language was used to programming corresponding training programs. The simulating result shown that, under the same condition, the improved BP algorithm saved a lot of training time and enhanced the precision of data fitness at the same time.