Research on UUV Recovery Active Disturbance Rejection Control Based on LMNN Compensation

As a kind of intelligent marine equipment, its successful recovery is the basis Unmanned Underwater Vehicle (UUV)’s normal working and completing mission successfully. However, mathematical model UUV severely coupled highly nonlinear, may be subject to complex interference in process. Besides, could not determined completely. Active disturbance rejection control (ADRC) method does need beforehand information unknown also can ensure stability. But conventional ADRC with fixed parameters adjust complicated motion control. This paper introduces adaptive wavelet neural network optimized by Levenberg-Marquardt algorithm, puts forward novel improved algorithm for partial system identification uncertain compensation. Moreover, pitching moment change considered process recovery, two dimensionless hydrodynamic coefficients are defined based on near-wall effect. The simulation experiments have been tested verify effectiveness proposed results indicate that variable environment more stable reduce computational burden.