Based On Ant Colony Algorithm Dynamic Vehicle Scheduling Problem Research

Vehicle scheduling problem involved in the industry is very extensive, high-paced life with real-time requirements of the transport industry more and more intense. How to be resolved on the basis of the static vehicle scheduling dynamic vehicle scheduling problem and make the optimal scheduling results become the focus and difficulty of the work of many researchers. Vehicle scheduling problem is NP-hard problem the algorithm exact algorithms and heuristic algorithms can be divided into two categories, the exact algorithm due to the introduction of a strict mathematical methods, can not avoid the exponential explosion can only effectively solve small-scale vehicle scheduling problem. The vast majority of researchers is focusing on the structure high-quality heuristic algorithm.In this paper, a variety of heuristic algorithms compare Choose improve the ant colony algorithm to solve transit the arrival of new orders this dynamic vehicle scheduling problem.1, A simple description of the vehicle scheduling problem, and then study the dynamic vehicle scheduling problem, compare the difference between static vehicle scheduling and dynamic vehicle scheduling, analysis of changes in the processing of information in the dynamic vehicle scheduling problem. Finally, the algorithm for vehicle scheduling problem, a precise algorithm, heuristic algorithms and sub-heuristic algorithm.2, various algorithms to compare the final choice ants swarm algorithm, a more in-depth research through a series of simulation experiments, the reasonable selection of the parameters of the ant colony algorithm, the optimal combination of algorithm parameters. Compared with previously fully use their experience and tentative to select the parameters, greatly improving efficiency. Simultaneously in three areas to improve the basic ant colony algorithm.(1) volatile factor from constant to variable functions;(2) the introduction of the incentive mechanism;(3) Max Min Ant System; The corresponding mathematical model. Improved ant colony algorithm with the basic ant colony algorithm, genetic algorithm to analyze an example compared to verify the effectiveness of the improved ant colony algorithm.3, simulation example:the use of matlab simulation software for the traveling salesman problem the static vehicle scheduling problem with time windows, the ideal road conditions and instances of realistic traffic simulation and compared by simulation comparison with the basic ant colony algorithm, genetic algorithm derived to improvement colony algorithm to get better optimization results,proved that the effectiveness of the improved ant colony algorithm.4.Having chosen a class of dynamic vehicle scheduling occur more frequently in real life-the arrival of new orders in transit. New orders processing and improved ant colony algorithm to accept or reject, accepted new orders on the basis of the delivery on the way to complete the task of the new orders.In this papers, the basic ant colony algorithm and the improved ant colony algorithm makes a class of dynamic vehicle scheduling problem is a good solution. To solve the purpose of vehicle scheduling problem has practical significance.