An improved Float-Coded genetic algorithm based on wavelet denoising mutation

Coding is an important and a difficult subject of research on genetic algorithm (GA). Float code (FC) is super to other codes in use. But, noise and its influence on GA performance were ignored by researches in genetic operation. Mutation played an important role of improving GA performance. This paper presents a float-coded genetic algorithm (FCGA) with wavelet denoising mutation. They are called the bound crossover and wavelet denoising mutation. By introducing the proposed genetic operation, both the solution quality and stability are better than GA with conventional genetic operations. Decomposing of FC noise was proved with wavelet in theory. FC denoising mutation was implemented in it. Experiments were made in it. The results of the research and experiments indicate that the theory is credible and the method is feasible in it. FCGA is of active significance to extend application space of GA.