主动配电网源—储容量优化配置研究

发布时间：2018-05-12 02:25

本文选题：主动配电网 + 储能系统　；参考：《南京理工大学》2017年硕士论文

【摘要】：随着新能源发电技术的不断发展和大规模应用,在缓解传统能源的同时造成了新能源利用率低、成本浪费等一系列问题,如何针对主动配电网中提高分布式能源利用率、保证供电可靠性来对储能系统进行优化配置是亟待解决的问题。本文针对主动配电网中分布式电源和储能系统的容量配置问题进行研究,主要进行以下几个方面的研究:首先,对分布式电源、可控能效负荷以及储能系统进行研究,建立了风力发电和光伏发电系统的数学模型,建立了可控能效负荷包括空调、热水器和照明负荷的能耗响应数学模型以及储能系统和逆变器的数学模型。针对需求侧能效负荷管理,结合分时电价和用户满意度的评价指标研究了单体用户可控能效负荷优化调度策略,为了对储能系统进行合理利用制定了储能系统充放电管理策略。其次,为了选取典型的自然资源样本,采用KMO和Bartlett球度检验选取相关性最佳的样本数据用于主成分分析,并且对小波-BP神经网络预测模型中的权值、伸缩因子和平移因子采用动量-自适应学习速率进行修正。针对多目标分布式电源和储能系统容量优化配置问题,采用和声搜索算法进行求解,对于算法所存在的收敛性能和容易陷入局部最优等问题进行了改进,对搜索过程中的记忆保留概率、和声微调幅度和微调扰动概率采用动态参数模式进行更新调整,与遗传算法进行融合提高其收敛性能。而后,以夏季典型日和冬季典型日的风速为例进行短期风功率预测,并与传统BP神经网络预测和WNN网络预测进行对比,结果表明了基于主成分分析的改进小波-BP神经网络预测方法的快速性和准确性。在分布式发电预测的基础上,以主动配电网中某地区夏季典型日和冬季典型日的风速、太阳光照和负荷需求数据为例,以分布式电源和储能系统的配置成本、新能源弃电率和负荷缺电率为多目标进行优化配置,通过算例验证了该模型和改进的和声搜索算法的正确性和有效性。最后,为了进一步降低配置容量和成本,通过需求侧能效负荷优化管理对分布式电源和储能进行容量优化配置,其配置结果与能效负荷优化管理前相比,其配置成本下降了 7.1%,负荷缺电率下降了 2.8%,新能源弃电率下降了 2.1%,验证了所制定的单体用户可控能效负荷优化管理策略的可行性和正确性,表明了在对需求侧可控能效负荷进行优化管理后,在降低用户用电成本的同时,能够进一步减少分布式电源和储能系统的配置成本,而且能够有效提高主动配电网的供电可靠性。
[Abstract]:With the continuous development and large-scale application of new energy generation technology, a series of problems, such as low utilization rate of new energy and cost waste, have been caused while alleviating traditional energy sources. How to improve distributed energy utilization efficiency in active distribution network? It is an urgent problem to ensure the reliability of power supply to optimize the configuration of energy storage system. In this paper, the capacity configuration of distributed generation and energy storage system in active distribution network is studied. The following aspects are studied: firstly, distributed power generation, controllable energy efficiency load and energy storage system are studied. The mathematical models of wind power generation and photovoltaic power generation system are established. The mathematical model of energy consumption response of controllable energy efficiency load including air conditioning, water heater and lighting load, and the mathematical model of energy storage system and inverter are established. According to the demand side energy efficiency load management, combined with the evaluation index of time-sharing price and customer satisfaction, the optimal scheduling strategy of single user controllable energy efficiency load was studied, and the charge and discharge management strategy of energy storage system was established in order to make rational use of energy storage system. Secondly, in order to select a typical natural resource sample, KMO and Bartlett sphericity test are used to select the best correlation sample data for principal component analysis, and the weights in the wavelet BP neural network model are predicted. The scaling factor and the translation factor are modified by the momentum-adaptive learning rate. In order to solve the problem of capacity optimization of multi-objective distributed power supply and energy storage system, the harmonic search algorithm is used to solve the problem. The convergence performance of the algorithm and the problem of falling into local optimum are improved. The memory retention probability, the amplitude of harmonic fine tuning and the probability of fine tuning disturbance are updated and adjusted by dynamic parameter mode, and the convergence performance is improved by fusion with genetic algorithm. Then, taking the wind speed of typical days in summer and winter as an example, the short-term wind power prediction is carried out, and compared with the traditional BP neural network and WNN neural network prediction. The results show that the improved wavelet BP neural network prediction method based on principal component analysis is fast and accurate. On the basis of distributed generation prediction, the wind speed, solar illumination and load demand data of a typical day in summer and typical day in winter in an active distribution network are taken as an example, and the configuration cost of distributed generation and energy storage system is taken as an example. The new energy loss rate and load power shortage rate are optimized for multi-objective configuration. The correctness and effectiveness of the model and the improved harmonic search algorithm are verified by an example. Finally, in order to further reduce the configuration capacity and cost, the energy efficiency load optimization management on the demand side is used to optimize the configuration of distributed power generation and energy storage, and the configuration results are compared with those before the energy efficiency load optimization management. The allocation cost has decreased by 7.1, the load power shortage rate has decreased by 2.8 percent, the new energy consumption rate has dropped by 2.1 percent, and the feasibility and correctness of the optimized management strategy of controllable energy efficiency load for individual users has been verified. It is shown that after optimized management of demand-side controllable energy efficiency load, it can further reduce the configuration cost of distributed power generation and energy storage system while reducing the cost of consumer electricity consumption. And it can effectively improve the power supply reliability of the active distribution network.
【学位授予单位】：南京理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TM73

【参考文献】