Research on Wave Compensation System for a Double-layer Stewart Platform Based on RBF Neural Network Active Disturbance Rejection Control

Aiming at the problem of insufficient control accuracy of the marine wave compensation system in a multi-disturbance environment, the core difficulty lies in the strong coupling characteristics of ship motion and wave disturbance, the control challenges brought by the multi-variable nonlinearity of permanent magnet synchronous motors, and the difficulty for traditional controllers to simultaneously achieve disturbance suppression and dynamic response performance.

Considering the advantages of the active disturbance rejection controller in estimating and compensating for internal and external disturbances without an exact model, as well as the powerful nonlinear approximation and adaptive learning capabilities of radial basis function neural networks, this paper proposes a wave compensation system scheme based on RBF neural network-optimized ADRC. To be more in line with the actual system, a three-loop permanent magnet synchronous motor is used as the actuator to build a complete control system.