Research On Vehicle Scheduling Problem Based On Cloud Model

Logistics is seen as the last border of reducing the cost, and it rows after thereducing material consumption and raising productivity. Logistics as the third sourceof profit plays an important role in our country’s economic and social development. Inlogistics distribution activities whether the vehicle scheduling is reasonable hasgreat influence on the distribution speed0 cost and the effects. So how to establish thescientific and scientific and reasonable method to optimize the vehicle schedulingproblem has been a hot issue in academic and business areas.In the actual city’s logistic system or the area’s logistics system, there are alwaysmany distribution centers. Thus the study on vehicle scheduling problem withmulti-distribution centers has its real meaning. The existing research documents usedto solve the vehicle problem with multi-distribution centers basically can be classifiedinto decomposition and global optimization. For decomposition has the defect that itcan not allocate the tasks suitably and the sub-question’s independence will destroythe integrity of the question, so using intelligent optimize algorithm to solve thevehicle scheduling problem from the global optimization aspect become popular.Genetic algorithm and its improving algorithm have been widely used in vehiclescheduling problem, but Genetic Algorithm has its immanent limitation and many ofthe previous algorithms just consider the trend of evolutionary process and ignore therandomness of evolution in natural environment. So this paper makes use of theuniversal character of the normal cloud mode and the properties of randomness andstable trend of a cloud model(randomness could keep the individual’s diversity inorder not to rolling into the shortcomings such as slow searching, easy to localoptimization solutions. And the stable trend can protect good individuals so that tolocate the global optimization adaptively), and the X-conditional cloud generatoralgorithm of the cloud model is used to generate the adaptive cross probability andmutation probability in the process of evolution, so we developed CGA to solve thevehicle scheduling problem.In this thesis it has firstly given a brief summary on Vehicle Scheduling Problem.By simplifying the Problem it gives the corresponding mathematical model and theintroduction of regular way of solving the problem and the application of GeneticAlgorithm in Vehicle Scheduling Problem. Secondly, it introduces the theory of cloudmodel and describes the nature and characters of the cloud model, and expatiate the working principle of the cloud generator. It talks about the start point of how toconstruct CGA and the application of CGA in vehicle scheduling problem. Finally,validate the practicability and effectiveness of using CGA to solve vehicle schedulingproblem with multi-distribution centers with specific examples.