| Summary: | 碩士 === 國立勤益科技大學 === 工業工程與管理系 === 107 === Logistics distribution have become the key conditions for the interconnection of various industries. In the rapid development of Internet, technology and transportation, how to improve the global competitive advantage of their enterprises, provide customers with instant logistics system, ordering to receiving in one day, so how can decision makers quickly choose the logistics distribution route to save transportation time, vehicle cost and reduce vehicle fuel consumption, so as to avoid unnecessary waste, it is already considered as a decision maker, so this study explored the topic of transportation problems.
In order to better conform to the actual case, this study will consider the vehicle routing problem with soft time windows (VRPSTW), let each customer's business hours be variable, and then restrict the consistency of each vehicle type, the same vehicle loading limitation, the same vehicle distance limitation, the same total vehicle service time, avoid problems such as employees overtime and excess workloads, make the problems closer to the current industry through the above restrictions, and the goal of minimizing the total cost is calculated. The total cost is divided into four parts: distance cost, vehicle fixed cost, vehicle early arrival cost, and vehicle late cost. The four costs are firstly formulated mathematical mixed integer programming model, and using LINGO 10.0 to solve the problem to get the best answer solution, and then using the genetic algorithm to solve the problem with MATLAB to find the approximate optimal solution, and save time and improve efficiency.
This study assumes three cases to verify the proposed MIP model and enhanced genetic algorithm. After comparison, it is found that the results in Case 1 and Case 2 are the same, while the results in Case 3 are inconsistent, but the error rate is only 8.68%. This method can effectively solve the solution and get the approximate optimal solution to replace the MIP model.
|
|---|
