Research on the Movement Control of Non-Traditional Six Propeller ROV Based on BP-PID Algorithm

In order to solve the problem that traditional PID control is difficult to adapt to the highly nonlinear and strongly coupled characteristics of ROV, this research, focusing on the motion control mechanism of ROV, constructs a non-traditional six-propeller-driven ROV physical platform and designs a PID control algorithm based on BP neural network. Through dynamic modeling and physical tests, the control effects of BP-PID and traditional PID are compared, and the improvement of system response characteristics and performance is evaluated. The experimental results indicate that the BP-PID control algorithm has significant advantages in enhancing the stability and response speed of the ROV system. After optimization, Pitch stabilizes after approximately 40 iterations, with the steady-state error controlled at about 4° and the maximum deviation also at about 4°; Roll stabilizes after 55 iterations, with the steady-state error at about 2° and the maximum deviation at 17°; Yaw stabilizes after 50 iterations, with the steady-state error at about 10° and the maximum deviation at 13°. These data suggest that the optimized system shows a shorter adjustment time before reaching the steady state and a significantly reduced steady-state error. In terms of dynamic performance, the overshoot of the system optimized by BP-PID is significantly reduced, and the recovery time is shortened, demonstrating superior anti-interference ability and robustness.