In this paper, the hybrid genetic-alopex algorithm (GAA) is proposed to solve the problem of synthesis of waste interception and allocation networks (WINs). Algorithm of pattern extraction (alopex) adopts 'noises' generated randomly to cast off local optima and enhances the capacity of hillclimbing. Genetic algorithms (GA) is good at keeping evolving of total population and parallel computing. The combination of GA and alopex enhances the algorithm in that it converges at the global optimum easily and accelerates the convergence to some extent. In addition, several advanced genetic operators are adopted to protect diversity of the population. Results of testing functions and a case study show that GAA is stable and can converge to global optimum well for the problem of high dimensions.