北京市电动出租车充电设施选址优化

发布时间：2018-07-22 14:51

【摘要】：随着近年来环境和能源短缺问题日益严峻,世界各国对于能源结构调整和环保技术研究的热情都持续高涨。由于传统的燃油燃气汽车是环境污染和能源消耗的主要来源,所以清洁的电力能源驱动的电动汽车受到了广泛的关注和青睐,国内外都有大量电动汽车被投入市场作为传统交通工具的替代品。然而由于电动汽车技术的局限性,其续航里程较短、充电时间长,所以电动汽车充电站无法使用传统加油站的建设模式和同等的建站密度,导致了现有充电站难以满足大量增长的电动汽车的充电需求,限制了电动汽车的进一步应用和发展。本文针对北京市电动出租车充电站选址的问题,以乘客出行和电动出租车运行的特征参数进行蒙特卡洛仿真,模拟了电动出租车充电需求产生的过程,识别了在目前情况下能够保障乘客正常出行的电动出租车充电需求的时空分布;采用voronoi图法划分了充电站的服务范围和容量,并在已知充电需求分布的基础上,构成了选址模型的约束条件,而充电站建设和运行的成本函数则构成了选址模型的目标函数;使用基本粒子群算法求出了低约束选址模型的解,与P中值模型的解对比验证了低约束选址模型的有效性,并引入禁忌粒子群算法提高了算法求解的精度;对影响选址模型结果的参数进行了敏感性分析、求出了选址模型的最优解。本文使用乘客出行距离和时间参数与电动出租车电池电量状态参数的概率分布,使模拟过程符合真实的事件发生概率,从而避免了对复杂电动出租车运行轨迹的研究,减少了数据获取的难度和数据冗余,同时提高了充电需求预测的精度;构建了基于粒子群算法求解的低约束选址模型,在保证模型解的精度能够满足需求的情况下,降低了模型约束的要求,简化了模型的复杂度,从而使得外部条件发生变化时模型需要调节的参数数量较少,模型能够较好地应对条件变化速度快的选址环境,避免了约束条件重新构建的繁杂过程,增强了模型的适用性和易用性;使用禁忌粒子群算法,在保证算法求解速度的前提下,提高了算法的精度,较好地避免了算法陷入局部最优的情况,为低约束条件下的选址模型提供了能够满足求解精度的理论依据。
[Abstract]:With the problem of environment and energy shortage becoming more and more serious in recent years, the enthusiasm for energy structure adjustment and environmental protection technology research in the world is continuously rising. Because the traditional fuel gas vehicle is the main source of environmental pollution and energy consumption, the clean electric vehicle driven by electric energy has received wide attention and favor. At home and abroad, a large number of electric vehicles have been put into the market as a substitute for traditional vehicles. However, due to the limitation of the electric vehicle technology, the electric vehicle charging station can not use the traditional gas station construction mode and the same station density because of its short mileage and long charging time. As a result, the existing charging stations are difficult to meet the increasing demand for electric vehicles, which limits the further application and development of electric vehicles. Aiming at the problem of location of electric taxi charging station in Beijing, this paper simulates the process of electric taxi charging demand by Monte Carlo simulation based on the characteristic parameters of passenger travel and electric taxi operation. The spatiotemporal distribution of charging demand of electric taxi which can guarantee the normal travel of passengers is identified, the service range and capacity of charging station are divided by voronoi diagram method, and the distribution of charging demand is known. The cost function of the charging station construction and operation constitutes the objective function of the location model, and the solution of the low constraint location model is obtained by using the basic particle swarm optimization algorithm. The comparison with the solution of P median model verifies the validity of the low constraint location model, and introduces Tabu Particle Swarm Optimization (Tabu) algorithm to improve the accuracy of the algorithm, and analyzes the sensitivity of the parameters that affect the result of the location model. The optimal solution of the location model is obtained. In this paper, the probability distribution of passenger travel distance and time parameters and electric taxi battery state parameters is used to make the simulation process accord with the true probability of occurrence of events, thus avoiding the study of complex electric taxi running trajectory. The difficulty of data acquisition and data redundancy are reduced, and the accuracy of charge demand prediction is improved. A low-constraint location model based on particle swarm optimization algorithm is constructed to ensure that the accuracy of the model solution can meet the requirements. The requirements of model constraints are reduced, and the complexity of the model is simplified, so that the number of parameters that the model needs to adjust when the external conditions change is less, and the model can better deal with the location environment where the conditions change quickly. It avoids the complicated process of reconstructing the constraint conditions, enhances the applicability and ease of use of the model, and improves the accuracy of the algorithm by using Tabu Particle Swarm Optimization (Tabu) algorithm under the premise of ensuring the speed of solving the algorithm. The algorithm can avoid falling into the local optimal condition and provide a theoretical basis for the location model under low constraint conditions to meet the accuracy of the solution.
【学位授予单位】：北京交通大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：U491.8

【参考文献】