Genetic Algorithm Based On Mechanism Of Positive Feedback Solves The Expert Assignment Problem

With the development of science fund system, staking fund sum and applying projects become more and more, and the peer review becomes more important in project management. The expert assignment that is chief work of peer review affects the quality and the income of projects staked directly. Recently, some researchers begin using optimization technique to solve expert assignment problem. However, because genetic algorithm loses a lot of excellent information without using feedback pheromone in the evolution, there are a large redundant repeats of searching at the last stage of evolution, and the efficiency to obtain précis results is reduced. All in all, the optimized result of genetic algorithm solving expert assignment is not satisfying.Firstly, in order to improve the disadvantage that genetic algorithm doesn't use feedback pheromone of system, we present genetic algorithm based on mechanism of positive feedback by putting mechanism of positive feedback of ant algorithm into genetic algorithm. The algorithm may search along optimization path by using the feedback pheromone of excellent individuals in former populations to instruct the evolution. The method improves the efficiency and the quality of searching of genetic algorithm. The TSP simulation results show the feasibility and the effectiveness of the above idea.Secondly, we put the mechanism of positive feedback into the strategy of genetic algorithm solving expert assignment. The problem that genetic algorithm has a large redundant repeats at the last stage of evolution is solved, the ability of breaking away from the local optimum is strengthened, and the precision of searching is improved by using the feedback pheromone of system instructing the mutation,Finally, this paper puts the method that genetic algorithm based on mechanism of positive feedback solves the expert assignment into practice, processes use case and class design according to the software development, and implements the system under Delphi 6.0 integrated development environment.