| Routing optimization is one of the most necessary and important part of fourth-party logistics study. Excellent and efficient distribution solution plays a very important role in improving the ability to integrate supply chain, the logistics qulity and efficiency, the service level for customers, reducing the distribution cost, resource waste, and satisfying the various demand of material products for the whole society.The distribution costs and delivery time are the key factors in4PL routing optimization problem, and the timeliness of distribution services is highly desired, therefore,4PL providers need to consider the effect on distribution solution.4PL routing optimization problem with four different distribution requirements are studied, including hard time windows requirements, soft time windows requirements, tardiness risk requirements, and eariliness/tardiness risk requirements, based on the background of4PL and the present researches on4PL.This dissertation reviews the background of4PL, the importance of routing optimization in4PL, and the present researches on4PL, firstly. And then, an overview of a new metaheuristic algorithm, the Harmony Search (HS) algorithm, is proposed. The detail description of the original of HS, the principle of HS, the flowchart and procedure of HS, the extensions of HS, the present researches on HS are proposed. The applications of the improved HS on combinational optimization and continuous optimization are also introduced. Finally, brief overviews of two routing optimiazation algorithms are also given, which provides an idea for sloving routing optimization problem in4PL.In order to solve the routing optimization problem in4PL with hard time windows, a mathematical model and algorithm are proposed. Considering the practical situation that the distribution task can not be completed on some fixed time point, the hard time windows are studied. Fistly, a mathematical model is proposed based on the path definition, taking the distribution costs as objective and the hard time windows as constraint. Repair strategy based on harmony search and deletion algorithm embedded harmony search are proposed. The results of these two algotihms are compared with enumeration algorithm and show that the proposed algorithms are effective. Secondly, a0-1integer programming model is set up, a commercial software package, ILOG, is adopted to solve the model. The experiment results verify the effectiveness of the model and algorithms. Finally, the different kinds of models and corresponding methods are compared with each other.The model and algorithm of the routing optimization in4PL with soft time windows are proposed. The distribution task can be completed in advance or delay with some punishment in practical situation, which called soft time widonws. Therefore, the routing optimization problem in4PL with soft time windows is studied on as follws:the mathematical model minimizing distribution costs and penalty cost of violating the soft time windows is proposed, repair strategy based on harmony search and deletion algorithm embedded harmony search are introduced, the results are compared with enumeration algorithm, and the effectiveness of the model and algorithms is verified. Then, a0-1integer programming model minimizing distribution costs and penalty cost is build, and the commercial software package, ILOQ is used to solve this problem. Results show that0-1integer programming model and the method are effective. The two kinds of models and the corresponding methods are compared and analyzed.For solving the routing optimization problem in4PL with delay risk, a mathematical model and the algorithm are proposed. The omnipresent uncertainty of delivery times on network links could result from weather conditions, traffic jams or accidents, temporary construction projects. For this reason, it is neccessary to consider the risk that the distribution task can not be completed in time. Value-at-risk is introduced to measure the delay risk, a mathematical model minimizing the tardiness risk is proposed, a deletion algorithm embedded harmony search is designed. To test the effectiveness of the model and the algorithm, sevral different scales of cases are solved. The experiment results show that the model and the algorithm are effective.As to the practical situation that the distiribution task can be completed in advance or delay with a penalty cost, the routing optimization problem in4PL with consideration of earliness/tardiness risk is studied. A mathematical model minimizing the distribution costs is set up and taking the earliness/tardiness risk as constraint, a deletion algorithm embedded harmony search is proposed. The effectiveness of model and algorithm is verified through sovling the different scales of cases.Finally a summary conclusion of this dissertation is drawn and some suggestions are given to future research. |
PDF Full Text Request