Summary: | 碩士 === 東海大學 === 工業工程學系 === 88 === As substantial portion of the cost of logistic is the cost of distribution, many researchers have been devoting themselves to the development of algorithms for the vehicle routing problem. Also, as the customers nowadays have higher demand on punctual delivery, and can not be satisfied if the deliver can only be made on the scheduled “day” but not the scheduled “hour”. Therefore, the traditional algorithms for the VRP can no longer meet the demand of the customers, and the adoption of new method such as time window become inevitable.
Because a variety of algorithms introduced by many researchers in the past have been tested and examined by many different ways such as random selection or actual case study, there is no general consensus on the strength and weakness on certain particular algorithm, nor is any consensus on which algorithm can be test used in which particular circumstance. Therefore, the analysis of this study will base on some of the best introductory algorithms found in the past research documentation. Also, in order to be objective, the logistic standard which has been more commonly adopted in the industry today has been used as a basis for comparison.
The following are the three results and conclusions of this study: First, it is found that the TABU method has a very good performance in the aspects of total length, total scheduled time, and the number of routing, therefore, it warrants a further study, Second, the common characteristics of many problems such as the location of the customers, the width of the time window and the establishment of the end goal are determinative factors in choosing the algorithms; Third, there is no single algorithms can suit every problem, and certain algorithm better suits certain problem and the algorithm has outstanding performance normally is the one that has already integrated the better algorithm .
|