A Multi-Objective Genetic Approach for Planning Electric Vehicle Charging Station Location Problems with Dynamic Demands

• Main Authors: Chen, Hsin-Kai, 陳信凱
• Other Authors: Chen, Jian-Hung
• Format: Others
• Language: zh-TW
• Published: 2014
• Online Access: http://ndltd.ncl.edu.tw/handle/23330551750412700878

碩士 === 中華大學 === 資訊工程學系碩士班 === 102 === With rapid development of vehicles, demands of fossil fuel dramatic increasing while the total amount of fossil fuel on Earth dramatic decreasing. Moreover, the uses of fossil fuel increase the greenhouse effect and thus aid in heating Earth’s surface and atmosphere. It results from the fact that certain atmospheric gases, such as carbon dioxide, is able to change the energy balance of the planet by absorbing long wave radiation emitted from the Earth’s surface. In order to seek for alternative energy substitutes for fossil fuel, the development of electric vehicle has become an important tendency. In order to promote the usages of electric vehicle, the location selection of electric vehicle charging station and minimizing maintenance costs has become two important issues for the development of electric vehicle. In recent years, researches on electric vehicle charging station always assumed that demands of charging electric vehicle are static. Moreover, the charging of each vehicle is assigned and limited to a pre-determined charging station. However, the demands of charging electric vehicle should be dynamic because the number of drivers may change in different seasons and their location may also vary due to many reasons. Pre-determining charge station is also not reasonable, because most vehicle drivers usually prefer to charge their vehicle in the nearest charge station, instead of charging vehicle at a pre-specific station which could be far from their current location. Despite of assigning charging stations, literature only considered building cost with respect to cost of electric vehicle charging stations. However, with increasing demands and electricity consumption by charging electric vehicles, maintenance of charging stations will definitely increase extra cost. Therefore, solving charging stations location selection problems should be take maintenance cost into considerations. This study will formulate a multi-objective optimization problem, which considers dynamic demands of charging electric vehicle, building and maintenance costs of charging stations. Hereafter, a multi-objective genetic algorithm (MOGA) with a seeding inheritance mechanism is used to solve the investigated problem, so that the minimization of transportation costs of vehicles, minimization of building and maintenance costs can be optimized simultaneously. In order to compare the performance of the seeding inheritance mechanism, a general multi-objective genetic algorithm without the seeding inheritance mechanism is also conducted for solving the investigated problem. The experimental results indicate that both multi-objective genetic algorithms are capable of providing a number of non-dominated solutions for decision makers to build charging stations. The results also indicate that with the uses of the seeding inheritance mechanism, the proposed multi-objective genetic algorithm can converge to better non-dominated solutions within the same computation costs than MOGAs without the seeding inheritance mechanism.