基于BP神经网络的固定翼航空电磁线圈姿态校正

固定翼航空电磁系统在测量过程中受风向、飞机飞行颠簸等影响，线圈姿态会发生变化，使接收到的电磁响应产生偏差。本文基于法拉第电磁感应定律，利用线圈坐标系和系统坐标系之间的转换关系，推导了线圈姿态变化时的电磁响应计算公式，得出电磁响应由线圈姿态变化过程中切割地磁场所产生的动态响应以及线圈处于倾斜状态下所感生的静态响应组成。并研究了基于BP神经网络的线圈姿态校正算法，首先利用地磁场和线圈姿态角度计算去除动态响应，然后构建BP神经网络模型，用该网络得出静态响应系数，实现静态响应校正。仿真结果表明，经过校正后的电磁响应数据质量得到很大改善。

Attitude Correction of Fixed-Wing Airborne Electromagnetic Coil Based on BP Neural Network

During the survey of fixed-wing airborne electromagnetic system, the transmitting and receiving coils are rotated by wind direction, flight bumpy etc., which makes the change of coil attitude,and it,in turn,causes error to receiving electromagnetic responses. Based on Faraday's law of electromagnetic induction, by using the transformation relationship between the coil coordinate system and the system coordinate system, we deduced the formula for calculating electromagnetic response,and concluded that electromagnetic response is composed of two parts:The dynamic response generated by cutting geomagnetic field in the movement process of coil and the static response induced in the sloping coil. We further studied the coil attitude correction algorithm based on BP neural network through the following steps:Firstly, the dynamic response was removed through calculation by using geomagnetic field and coil attitude angle,then BP neural network model was constructed, and the static response was corrected by using of the static response coefficient obtained in the network. The simulation results show that the quality of the corrected electromagnetic response data is greatly improved.