Abstract: With a view to the difficulty of the traditional hydrodynamic parameters experiments and theoretical calculation methods for autonomous underwater vehicle (AUV) and the drawback of optimum results tending to local optimization of the ever identification methods, a genetic algorithm based hydrodynamic parameters identification method of AUV is presented. This method has better robustness and can find the globally optimal point without initial values of parameters. It can provide a dynamic model for motion control, condition prediction and control system exploitation of AUV. Aimed at an open-frame AUV prototype named by CRanger-01, the genetic algorithm is used to identify the horizontal plane hydrodynamic parameters based on the analysis of the dynamic modeling. Then nineteen horizontal plane hydrodynamic parameters of CRanger-01 are identified, compared with the actual values, the identification errors are in the allowable range. By the simulation of the dynamic model, it demonstrates that the method can identify the hydrodynamic parameters of AUV effectively and can provide a reference for engineering practice.

Key words: Autonomous underwater vehicle, Genetic algorithm, Hydrodynamic parameters, Parameter identification