The invention discloses an intelligent optimization method for vane profiles of fixed guide vanes of a water turbine. The method comprises the steps of building a parametric geometric model of the fixed guide vanes according to vane profile parameters of the fixed guide vanes of the original water turbine; building a three-dimensional CFD flow field simulation model of the fixed guide vanes on thebasis of the parametric geometric model of the fixed guide vanes, and then performing CFD simulation calculation on the fixed guide vanes of different vane profiles, thereby obtaining the influence of the vane profile parameters on flow field characteristics; establishing a training sample set by utilizing the influence of the vane profile parameters on the flow field characteristics, and fittinga neural network by adopting a method of an extreme learning machine, thereby obtaining corresponding relationships between the vane profile parameters and performance indexes; and according to the corresponding relationships between the vane profile parameters and the performance indexes, establishing an objective function, and optimizing the vane profile parameters in the neural network in combination with a genetic algorithm by utilizing the objective function, thereby obtaining optimal vane profile parameters. According to the intelligent optimization method, the performance of the fixedguide vanes of the water turbine can be effectively improved according to actual engineering requirements, the design time is shortened, and the design cost is reduced.