Passive Electromagnetic Field Positioning Method Based on BP Neural Network in Underwater 3-D Space

Chaoyi Wang,Yidong Xu,Junwei Qi,Wenjing Shang,Mingxin Liu,Wenjian Chen

DOI: https://doi.org/10.1007/978-3-030-94551-0_24

2022-01-01

Abstract:This paper studies the positioning method of combining the passive electric field positioning and passive magnetic field positioning under three-dimensional (3-D) water. This technology can be applied to underwater submarine positioning, underwater leakage power supply positioning, underwater rescue and other occasions. We collect the samples data by electromagnetic sensors array, and recover the location of the targets. After the data preprocessing process includes data normalization, de-redundancy process, and generalization process, we use Back Propagation (BP) neural networks to build electric field and magnetic field distribution models of electric dipole source. Finally, we enter the test data to obtain the target position in the well-trained positioning model. We take the Euclidean distance between the ideal position and the model output target position as an absolute error. The results show that this method can effectively improve the accuracy of underwater target positioning and anti-interference ability of the training model, and the nonlinear function model trained by the neural network can be applied to the complex and changeable underwater environment.