The preparation of reinforced metal matrix composites used by the method of in-situ and laser cladding has become one of the interesting topics in the metal matrix composite materials. This method is adopted here to prepare the reinforced iron matrix exterior composites, and the critical task is to obtain the optimized technique parameters. The optimal process of manufacturing composite is traditionally defined via experimental methods of orthogonal arrays design, which is costly and can’t get the exact parameters. Hence, we firstly use the artificial neural network to fit the relationship between the performances of the composite and the parameters so as to construct the model of the process. And the parameters are optimized by genetic algorithm. The results further verified that the method in this paper is effective.