Improved Genetic Algorithm To Solve The Vehicle Routing Problem

In the logistics process, the most important aspect is distribution, and distribution costs hasbecome a large portion of the cost of the entire logistics,so how reasonable and effectivearrangements for the distribution to become the primary task to reduce logistics costs. Thedistribution is the core vehicle transport, it can be said to solve the Vehicle Routing Problemif we solved the distribution problem.With the popularity of electronic commerce, it makes the distribution is moredifficult.And growing vehicle transportation arrangements also become the primary problem,due to vehicle transport problem is very difficult,and has enormous economic value, it hasbeen attracting more and more competing research scholars and enterprises.This paper first analyzes both domestic and foreign of VRP With reality, established aresearch model of this article-to take delivery vehicle routing with time windows(VRP/SDP/TW). By the model gives a mathematical formula to clarify the constraints of theproblem: reasonable arrangements for the path to meet the capacity of the vehicle andcustomer service time under the conditions of transport, in order to meet customer demand forservices and make the goal of minimization of the transport path.Next, we introduce and analyze the genetic algorithm in detail, illustrates the advantagesand disadvantages of genetic algorithm, genetic algorithm is easy to precocious and slowoperation shortcomings, propose three strategies for improvement:1、the use of adaptiveparameter settings;2、Particle swarm optimization and genetic algorithm fusion;3、the geneticparallelization. The first two strategies to solve the premature convergence problem, a thirdstrategy to solve the problem of slow computation speed.Finally, the proposed method is applied to take delivery vehicle routing problem with timewindows (VRP/SDP/TW) to go in, and gives a detailed operating procedures, and use JAVAlanguage implement the algorithms. The comparative analysis of two experiments with thebasic genetic algorithm and the improved ant colony algorithm, the effectiveness of thestrategy proposed in this paper: can effectively avoid premature enhance the global searchability, but also at the same time computing speed have a larger increase.Overall，the use of PSOPGA provided a higher value for logistic enterprises，which woulddecrease the cost，improve the level of logistics management,response customers more quickly and increase the competitiveness of the enterprise．...