H∞-ESO Based Robust Path Tracking with Multi-Model Fusion for Underwater Robots in Nonlinear Disturbance Environments

Abstract

As an important equipment for Marine resource development and environmental monitoring, the path tracking accuracy of underwater robots is directly related to the reliability and stability of task execution. However, in the dynamic Marine environment, factors such as water flow disturbance, hydrodynamic nonlinearity, and uncertainty of system parameters can cause significant path deviations. Traditional control methods such as PID or sliding mode control have deficiencies in terms of robustness and real-time performance. To this end, a composite control strategy integrating H∞ robust control, extended state observer (ESO) and multi-model dynamic compensation mechanism is proposed. This strategy takes H∞ control as the core to enhance the lower bound of the system's stability against the most adverse disturbances. The unmeasured states and synthetic disturbances are estimated in real time through ESO to enhance the system's perception ability of complex disturbances. Combined with the multi-model fusion mechanism, the control model is adaptively switched for different interference modes, effectively enhancing the adaptability and flexibility of the control strategy. A six-degree-of-freedom simulation model was constructed in typical path and complex disturbance scenarios. Experiments were carried out by setting multiple performance indicators such as root mean square error (RMSE), maximum error, steady-state error, disturbance recovery time and performance retention rate. The results show that the RMSE of the H∞-ESO fusion controller is controlled at 0.103 meters in the undisturbed scenario, with a maximum error of 0.26 meters. In a strongly disturbed environment, the RMSE was 0.136 meters, the error recovery time was only 3.2 seconds, and the performance retention rate reached 87.5%, all of which were significantly better than the traditional H∞, ASMC and PID methods. This strategy performs outstandingly in improving control accuracy, enhancing system robustness and ensuring real-time response. It has good engineering deployability and promotion prospects, providing technical support for high-precision path tracking of underwater robots in unstructured environments.