Research On The CLRIP Optimization Of B2C E-commerce Self Distribution System Under Fuzzy Random Environment

E-commerce has an important position in the national economy and social development, and also has become a new driving force for the economic development. The rapid advance of E-commerce has imposed great pressure on the logistics and distribution. Especially in recent years, as B2 C E-commerce develops swiftly, increasing competition among B2 C enterprises have put forward a higher demand for the distribution efficiency and reaction speed, and also make the demand of distribution system optimization more urgent. To carry out distribution system optimization scientifically and reasonably not only enables B2 C enterprises to reduce operation costs, improve the overall benefit and better serve their customers, but also helps build and enhance the core competitiveness of enterprises, improve their logistics and distribution capabilities and reduce the pressure on social logistics. Therefore, it has important theoretical and practical significance for further study on the optimization of B2 C self distribution system.Three levels of decisions should be made to optimize self distribution system of B2 C enterprises: strategic, tactical and operational level. The strategic layer makes decision on the location of distribution centers and other facilities, the tactical layer on the inventory control and the operational level layer on Vehicle Route Problem(VRP). Because the decisions for solving these three levels are interrelated to one another, Combined Location Routing and Inventory Problem(CLRIP) should be fully considered. In previous literature, most of the different levels of decisionmaking are considered separately. There are few studies on the combination of multilevel decision factors and even less research on the integrated optimization of CLRIP. Meanwhile, the current documents discussed CLRIP on the assumption that system variables are random or fuzzy, but in the practice of enterprise distribution, system variables are often in the coexistence of random and fuzzy. Therefore, the current study did not consider the double uncertainty problem with fuzzy random variables. According to these problems, this dissertation has a deeper insight to offer a new perspective and approach of study.Firstly, combining the characteristics of B2 C distribution system, this dissertation analyzes the operation processes for B2 C self distribution system, and summarizes the operation process of the general B2 C self distribution system. In order to better reflect the practical situation of the system, this dissertation also analyzes the fuzzy random environment of this system, summarizes the crisp methods of fuzzy random variables, and further studies the decision factors of the system, such as the location-allocation, inventory control, vehicle routing and service time, etc. Then, a deep study is made on the relationship between the essential factors, arguing that they are not independent, but interacting. Therefore, when the optimization of B2 C self distribution system is concerned decisions should be made from the perspective of a multi-level integration. On this basis, this paper constructs the overall framework of study on the CLRIP model of B2 C self distribution system under fuzzy random environment.Aiming at the fuzzy randomness of B2 C enterprises customer demand, under the policy of continuous and periodic review inventory, the CLRIP model with fuzzy random demand is built in this paper and the method of fuzzy possibilistic mean value is used to transform into the crisp equivalent mode. Then, a two-stage heuristic algorithm based on Tabu Search(TS) is designed. The first stage of this algorithm is to construct initial solution stage, and the second stage is to continuously improve the solution quality in location-allocation stage and inventory-routing stage. This algorithm can effectively reduce the distribution cost of B2 C enterprises.Combined with the characteristics of high requirements and personalized service time of B2 C Ecommerce customers, this paper introduces fuzzy random time windows, proposes the multi-objective CLRIP model in the policy of continuous and periodic review inventory respectively. The target of the multi-objective model is to minimize the total cost of system and maximize the average customer’s satisfaction of service time windows. Then the multi-objective problem is converted into a single objective problem by using constraint method. And the fuzzy random time windows are converted into the determined equivalent form through the method of fuzzy random expected value. Furthermore, this dissertation develops the improved genetic algorithm, adopts adaptive selection mechanism, the crossover operator based on chromosome gene location of spouse and the swap mutation operation in this algorithm. Although considering the constraint of customer’s service time windows enables system total cost to increase, it also can improve the efficiency of distribution and the customer satisfaction effectively.In view of the high return rate of B2 C E-commerce, this paper constructs CLRIP models with fuzzy random demand, fuzzy random time windows and the two fuzzy random stations occurring at the same time separately when considering the return. Then the fuzzy random variables are transformed into the crisp equivalent form by using the method of fuzzy possibilistic mean value, fuzzy random expected value and fuzzy random simulation. A two-stage heuristic algorithm based on Tabu Search(TS) and Simulated Annealing(SA) is developed to solve the CLRIP model. Even though the total cost is higher than that without considering the return, return service can increase customer satisfaction and bring more potential customers for B2 C enterprises.Employing this CLRIP model with fuzzy random demand when considering the return and the two-stage heuristic algorithm based on TS, a demonstrated study is conducted on optimization for B2 C self distribution system of green and organic vegetable base in the Great Northern Wilderness Hongqi farm. Based on the analysis on the present situation of the distribution system, the CLRIP model with fuzzy random demand of B2 C self distribution system is optimized. The results show that the model and algorithm in this paper have proved to be reliable and effective. It also provides scientific evidence for distribution system of B2 C E-commerce and related enterprises.