Yaw angle control of a boxfish-like robot based on cascade PID control algorithm

Because of the complexity of water environment, it is difficult to precisely model the fish robot system. Therefore, model-based attitude control algorithm is difficult to apply to the robotic fish. Cascade PID control uses the measurements of controlled variable and its derivative to constitute a double feedback loop. It can adjust the interference that influence the intermediate variable in advance and improve the dynamic quality and working frequency of the whole system. It is better than the traditional PID controller in anti-interference performance, rapidity, adaptation and stability. This article adopts the cascade PID algorithm to control the yaw angle of a robotic fish for the first time. This article innovatively uses a Kalman filter to eliminate the system angle error caused by periodic oscillation of robotic fish. A yaw angle control framework was proposed for biomimetic robotic fish. Based on the proposed framework, systematically yaw angle control experiments with the robotic fish were carried out in situations where the reference inputs were step signal, square wave and sine wave. Experiments showed that the robotic fish can accurately track the reference of the yaw angle in real time, verifying the effectiveness of the proposed algorithm.