Research And Application Of Vehicle Routing Problems Based On Ant Colony Algorithm

In the logistics cost, the transportation cost occupied more than half of the total cost, sotransportation cost savings have a lot of space.Especially in recent years, with thedevelopment of the economy, customers looking forward to higher service level. Theywant companies to provide with high-quality services, also to reduce the cost oftransporting goods. The meaning of the vehicle routing problem research, not only hastheoretical significance, but also has certain practical significance.The VRP is the problem of a non-deterministic polynomial algorithm (nondeterministicpolynomial, NP). Therefore, it is necessary to check on a large multi-dimensional,non-contiguous space in order to find the optimal solution of the VRP. In general,NP-hard problems have a non-linear variables and constraints, and with the furtherincrease of the variables and constraints, it will make rapid rise of the difficulty ofproblem solving. Therefore, it has yet to find the most effective solution. The currentsolution algorithms can be divided into exact algorithms and heuristic algorithms. Thisarticle mainly uses the ant colony algorithm to solve the VRP problem with timewindows. Ant colony algorithm uses the principle of positive feedback, which canaccelerate the process of looking for the optimal solution; because of the nature of antcolony algorithm is a parallel algorithm, individuals can the transfer and exchange ofinformation.This paper firstly introduces the research background of the vehicle schedulingproblem, significance, propose and current development status, we can know someknowledge about the vehicle scheduling issues. The second chapter briefly described theprinciples of logistics problems, the characteristics of ant colony algorithm, as well asgeneral mathematical model of ant colony algorithm. The third chapter analyzes basic,limited VRP problems, and the pros and cons of the general solution of VRP.In the fourth chapter of the, according to the characteristics of VRP with time windows, this paperanalyzes the general path of the ant colony algorithm selection mechanism, thepheromone update mechanism and coordination mechanisms, also introduces Improvedregional search method to solve VRP problem with time windows. After that, this papermakes design in places such as pheromone initialization, solution build, andregionalization search. At last, paper used the Benchmark Problems in this design andimplementation of the VRP with time windows for testing. Finally, the ant colonyalgorithm and its path optimization problem in the study are summarized and pointed outthe direction in the next step study.