| The transportation network is the operation basis of post enterprises. In logisticsindustry, the market competition is becoming fiercer and fiercer. It is one of theimportant problems for post enterprise to make full use of its resource advantages,especially in the location of postal center, vehicle scheduling and in-time service. Thetransportation network of postal universal service (TNPUS) has great advantages inscale, fund and customer and plays a key role in China postal. Thus, the research on theoptimization problems in TNPUS, such as center location, regional planning, vehiclescheduling and route planning is very important to the elaborate management Chinapost and the results have good opportunity to be practiced.From the view of optimization, the thesis analyzes the conditions of the Chinapostal transportation network, such as vehicle scheduling, center location and postaltransportation scheduling. Then, the major research direction of this thesis is found. Themain content of this thesis is shown as follows:(1) Firstly, a single-center postal transportation optimization model is proposed.Some important factors, such as multi-type vehicles, pick-up and delivery of packages,are also considered in this model. Then a hybrid algorithm of combing taboo searchwith genetic algorithm is designed to solve this model. The simulation analysis isperformed with the production data of Zunyi postal.(2) Based on the first model, a hybrid postal transportation model with centerlocation and vehicle routing problem is presented. The solution of the model is dividedinto two stages. In the first stage, the third-level centers and their regions are determined.Moreover, the parcel deliver tasks between the second-level and the third-level centersare distributed. In the second stage, the vehicle scheduling scheme in each center isobtained and the result is returned to the first stage. The center distribution method andtaboo-genetic hybrid algorithm are employed in the two stages, respectively. Finally,the simulation analysis is performed with the production data of Guizhou postal.(3) The time constraint of service is added into the second model. The improvedmodel is much closer to the actual conditions of postal enterprise. The verifiedexperiment for the improved model is performed with the same data.The results show that these models and algorithms are valid and practical, and thedistribution solutions can greatly reduce operation cost of post enterprise. It not only brings a new idea of optimizing the postal transportation network but also gives someexperiences in logistics optimization. |
