Improved Genetic Algorithm And Its Application On Location Of Logistics Distribution Center

Genetic algorithm is a random searching method which simulates natural selection and evolution. This method has some advantages that other usualmethods don't have because of its two characters------implicit parallelism andglobal searching. But after all, genetic algorithm is a newborn optimizing method and both its theory and its realization need to be improved. Only in this way, can genetic algorithm apply to the practice more effectively and widely.After the characters, development, application and the foundational theory of genetic algorithm being introduced, the simple genetic algorithm is improved on in this thesis aiming at its application limitation. The improving work is as follows.1) A conclusion is made that there is an optimal combination of crossover rate and mutation rate after study on the relationship of that two rates. Thus the blindness of selection of the two rates is reduced in a way.2) For the ability of searching, a formula of adaptive crossover rate decreasing with the increasing of genetic generation.3) Improving the method of penalty function inosculating with simulated annealing.Delivery is very important in logistics system. Logistics distribution center is a connecting link between the preceding and the following. A reasonable location of logistics distribution center is of benefit to the whole logistics system. Once the distribution center is decided, it will run longtime. The center has not only a direct connection with freight but also a big influence upon the work efficiency and the logistics control level.Some models and datum are necessary to location of logistics distribution center when we analyze and design logistics system. There are many solutions to the problems of location, but it is difficult to solve the big size system by these orthodox methods in the practical application.In this thesis, following improving of the simple genetic algorithm, the improved genetic algorithm is used to solve the problem of logistics distributioncenter location, getting the resolution of the location model. And then based onpractical application, a new coding method--hybrid parallel code is presented.The application show that genetic algorithm is simpler and its operation is speedier, especially when the problem is complicated compared with other optimizing methods.