能源互联网环境下电动汽车充电站选址优化模型研究

发布时间：2019-01-02 13:13

【摘要】：在全球能源互联网的大环境下,作为新能源、新材料、新方向、新技术等现代高科技技术综合集成品的电动汽车将在能源、交通、电力、通信等行业掀起巨大变革,这一变革或将成为新能源革命的重要推动力量。各国为鼓励电动汽车行业的发展相继出台了大量优惠政策。电动汽车相关优惠、支持政策的落地,势必会带动电动汽车行业迅速发展,所以充电站等基础配套充电设施的发展和规划是否科学、合理将成为影响电动汽车行业能否平衡、快速发展的关键因素。而目前作为基础配套设施的充电站建设存在着认识不统一、政策配套不完善、协调推进难度大、标准规范不健全等问题,严重制约着电动汽车行业的规模型发展。为了更科学、规范化的建设充电站,本文在充分考虑电动汽车与能源互联网融合的基础上针对目前充电站规划存在的问题从理论和实践意义论述了充电站科学优化选址的必要性和紧迫性。首先从电网安全、经济发展、交通便利性、环境影响、技术发展和规划协调性六个方面分析了影响充电站选址的因素,构建了基于能源互联网的充电站选址评价指标体系并根据系统论的观点从系统的角度探讨了经济、交通、环境、技术、电网安全等各规划因素之间的逻辑关系及相互作用,构建了充电站选址DPSIR模型,并将各系统指标分别划分在驱动力系统、压力系统、状态系统、影响系统和响应系统五个不同的系统中,通过系统间的相互关联作用分析了各要素之间的逻辑影响。其次,提出了基于遗传算法优化BP神经网络的充电站选址模型并使用熵值法对各指标的权重进行了计算。再次,为避免传统物元可拓模型出现指标值超出节域导致关联函数无法计算的情况,利用改进的物元可拓模型对充电站进行了选址规划。另外,本文基于遗传变异改进蚁群聚类算法的RBF神经网络对充电站进行了选址评价。使用基于遗传变异改进的蚁群聚类算法来确定RBF神经网络隐含层的个数,从而解决RBF神经网络初始参数没有科学方法难以精确选取的问题并使用实例证明了此方法的科学性和有效性。最后比较了各方法的优缺点,为充电站选址规划问题提供了新的求解方法和思路。
[Abstract]:In the global energy Internet environment, as new energy, new materials, new directions, new technologies and other modern high-tech integrated products of electric vehicles in energy, transportation, electricity, communications and other industries will bring a huge change. This change may become an important driving force of the new energy revolution. In order to encourage the development of electric vehicle industry, countries have introduced a large number of preferential policies. The development and planning of the basic charging facilities such as charging stations and other basic supporting facilities will inevitably lead to the rapid development of the electric vehicle industry. Whether the development and planning of the basic charging facilities, such as charging stations, is scientific or not, will affect the balance of the electric vehicle industry. Key factors for rapid development. At present, the construction of charging station as the basic supporting facilities has some problems, such as the lack of unity of understanding, the imperfect policy, the difficulty of coordination and promotion, the unsound standard and so on, which seriously restrict the scale development of the electric vehicle industry. In order to build a more scientific and standardized charging station, On the basis of considering the fusion of electric vehicle and energy internet, this paper discusses the necessity and urgency of the scientific optimization of charging station location from the theoretical and practical significance in view of the problems existing in the current charging station planning. First of all, the factors influencing the location of charging stations are analyzed from six aspects: power grid safety, economic development, convenience of transportation, environmental impact, technological development and coordination of planning. The evaluation index system of charging station location based on energy Internet is constructed, and the logical relationship and interaction among various planning factors, such as economy, traffic, environment, technology, power grid security, etc., are discussed from the viewpoint of system theory. The DPSIR model of charging station location is constructed, and the system indexes are divided into five different systems, namely, driving force system, pressure system, state system, influence system and response system. The logical influence of each element is analyzed by the interrelation between systems. Secondly, a charging station location model based on genetic algorithm to optimize BP neural network is proposed and the weight of each index is calculated by entropy method. Thirdly, in order to avoid the problem that the index value is beyond the domain and the correlation function can not be calculated in the traditional matter-element extension model, an improved matter-element extension model is used to plan the location of charging station. In addition, RBF neural network based on genetic variation improved ant colony clustering algorithm is used to evaluate the location of charging station. Ant colony clustering algorithm based on genetic variation is used to determine the number of hidden layers in RBF neural network. In order to solve the problem that the initial parameters of RBF neural network are difficult to select accurately without scientific method, the scientific and effective method is proved by an example. Finally, the advantages and disadvantages of each method are compared, which provides a new method and train of thought for the location planning of charging station.
【学位授予单位】：华北电力大学(北京)
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP18;F426.471;F426.61

【参考文献】