Research on Optimization of the Customer-Enterprise Interacting Mechanism based on the GA-WNN Method

To raise the effective and efficiency of the quality-improvement, the Customer-Enterprise interacting mechanism for the Quality-Improvement oriented to the customer loyalty should be built. As a complex nonlinear system composed by suppliers, manufacturers and customers, this mechanism´s state is very uncertain & unpredictable. It makes that the optimization of the Mechanism has significance. To resolve the problem of optimizing this system, the GA-WNN optimization method that combining genetic algorithm (GA) with wavelet neural network (WNN) has been proposed in this paper. In the GA-WNN method, wavelet neural network has been applied to identify and model on the system according the high approximation ability of WNN for identifying the complex nonlinear system, then genetic algorithm as a kind of overall search optimization methods has been used to the system optimization of this system. To accelerate the convergence speed and learning speed of GA-WNN method, the parameters of the wavelet networks are trained by using hybrid algorithm, and the mixed GA modified by selected strategy and combined local search algorithms is adopted. It has been proved by simulation results and practices in enterprises that the GA-WNN method can optimize the customer-enterprise interacting Mechanism effectively and make this system tend to the optimum operation point so as to raise the effective and efficiency of the quality-improvement