LPG Transportation Research On Vehicle Routing Problem

In recent years, due to the shortage of oil resources and people’s awareness Of environmental protection, LPG, with its clean and green, energy-efficient, cost and economic characteristics, as high-quality energy is gradually in various fields for people of all ages. Thus, with the increasing amount of LPG use, closely related to the LPG transport demand is also increasing. The LPG for exclusive use shipping is limited due to the impact of the shipping market, each Refining and Chemical Company, through flexible road tankers distribute to the demand areas. As for LPG companies obtaining higher profits in transit is a fundamental goal of the pursuit of business, but LPG as dangerous goods, to ensure their safety during transport, to reduce the risks associated with transport as much as possible, in this context, timely and efficient transport to customer demand is also essential. In order to better serve its customers, make full use of the resources of the enterprise, to create a better economic, social environment, choosing an optimal transport path is scientific and rational.Currently, Transportation and distribution of LPG Refining and Chemical Company, is using more traditional methods based on the driver’s experience of artificial selection of routes. When faced with an emergency situation, it often caused by LPG not arrived in time, greatly reducing the quality of service. Path selection optimization problem is a combinatorial optimization problem, and the most effective is to use heuristic search algorithm to solve this type of combinatorial optimization problems. In this paper, the actual situation of LPG vehicle transport, in the analysis of existing research results based on the characteristics of the transport of LPG, and establish practical problems with a more appropriate vehicle routing problem with time windows multi-objective mathematical model. For particularity, we constraint the model to improve the model targeted and strengthen the guiding role of the actual problem. At last, we select the risk, time and cost representative goal and establish a multi-objective optimization problem, Using this instead of the traditional multi-objective optimization method to solve the model, and Improvement the genetic algorithm to make the results more reasonable.