基于BP神经网络处理测斜实验数据的方法

Processing method of the experiment data of a dip meter through BP neural network.

介绍了利用人工神经网络进行测斜仪传感器误差校正的原理,提出了基于BP神经网络的测斜仪传感器(加速度传感器,磁通门传感器)安装误差校正方法,并与其它优化算法校正误差进行了比较,最后给出了一个仿真实例。实验结果表明采用BP神经网络可以提高网络收敛速度,大大减小传感器线性误差。

This paper introduces the principle of correcting linear errors of the sensors of dip meters by neural network. The method of back propagation neural network(BPNN)is given to correct linear errors of the sensors of dip meter. The experimental results show that the network learning speed can be increased and the nonlinear errors of the sensors can be reduced by using BPNN.