Balance parameters calculation method of underwater glider based on BP neural network

Underwater glider is a long-range autonomous underwater vehicle for oceanographic research that propels them through the ocean by changing their buoyancy. Balance parameters usually obtained through experiments. These kind of inaccurate parameters have a great influence on gliders gliding. In this article, a balance parameters calculation method of underwater glider based on BP neural network is presented. This study aims to find one of the most suitable balance parameters to improve the control effect of vehicles. Three steps are put forward to implement this approach. First, do at least two groups of pitch adjusting tests in the water and record the related data at the same time. Second, extract data and data preprocessing. Finally, use the BP Neural Network model to calculate balance parameters. The results of Qian-dao Lake test show that these parameters are appropriate and this calculation method is feasible. A further sea trial continues.