Fuzzy Multi-objective Expected Value Optimization Models for Locating an Automotive Service Enterprise
Qiang Li,Hongfei Jia
DOI: https://doi.org/10.14257/ijunesst.2016.9.5.07
2016-01-01
Abstract:Automotive service enterprise location is an interesting and important issue in the logistic field. In practice, some factors of its facility location allocation (FLA) problem, i.e., customer demands, allocations, even locations of customers and facilities, are usually changing, and thus FLA problem features with uncertainty. To account for this uncertainty, some researchers have addressed the fuzzy time and cost issues for locating an automotive service enterprise. However, a decision-maker hopes to minimize the transportation time of customers meanwhile minimizing their transportation cost when locating a facility. Also, they prefer to arrive at the destination within the specific time and cost. To handle this issue via a more practical manner, by taking the vehicle inspection station as a typical automotive service enterprise example, this work presents a fuzzy multi-objective expected value optimization approach to address it. Moreover, some region constraints can greatly influence FLA and travel velocity is also an uncertain variable due to the influence of some unpredictable factors in the location process. To do so, this work builds two practical fuzzy multi-objectives programming models of its location with regional constraints, fuzzy inspection demand, and varying velocity. A hybrid algorithm integrating fuzzy simulation, neural networks (NN), and Genetic Algorithms (GA), namely a random weight based multi-objective NN-GA, is proposed to solve the proposed models. A numerical example is given to illustrate the proposed models and the effectiveness of the proposed algorithm.