On the basis of neural network theory,this paper investigated the hydrodynamic coefficients of the linear type,weakly nonlinear type and strongly nonlinear type.To determine the coefficients of vessel movement,all neural network identifiers were optimized through the fuzzy immune genetic algorithm.Al so,simulations on wheel movements for the 10°/10° Z shape,30° circumgyratetion and 35° circumgyrate tion were performed.The results show that the corresponding curve for each wheel movement agrees well with the input parameters,which validates the effectiveness of the proposed approach.