微电网系统控制技术研究

发布时间：2019-05-29 05:18

【摘要】：独立微电网将分布式电源、储能装置、区域负荷及控制装置有效整合,形成单一可控单元,在海岛、沙漠等偏远地区的开发上有着广泛的应用前景。由于大量间歇性可再生能源发电的接入及系统容量较小,独立微电网中的随机性问题愈发凸显。制定有效的系统控制策略,实现独立微电网稳定高效的运行,在实际开发建设中有着重要的意义和迫切的需求。本文考虑到负荷和可再生能源发电在实际运行中的波动导致提前制定的调度计划无法有效执行,在对负荷可再生能源发电超短时预测的基础上,提出了一种针对独立微电网可控电源功率输出的协调控制策略。论文首先总结了当前独立微电网系统发展建设历程,介绍了典型独立微电网结构,建立了常见分布式电源的数学模型；研究了集中控制模式下可控电源的控制原理,考虑到尽可能利用可再生能源,研究了其最大功率跟踪发电模式的原理及实现方法。其次,探讨了独立微电网内负荷可再生能源发电超短时预测的必要性,通过总结分析常规电力系统中负荷超短期预测方法,采用基于最小二乘支持向量机的负荷实时数据序列直接预测方法；在考虑可再生能源间歇性的实际情况下,采用与负荷预测相同方法实现对风力发电的超短时预测,通过相似日选择实现对光伏发电功率的直接预测。再次,提出了包含日前调度计划和在线控制策略的独立微电网多时间尺度控制策略,建立控制系统网络架构；在满足负荷需求和电源约束条件的基础上建立了集中控制式的独立微电网日前调度模型,综合考虑了燃料成本、维护成本和排放成本等多种目标,并引入粒子群算法对优化调度模型进行求解；引入了微电网净负荷的概念,在对净负荷超短时预测的基础上,提出了一种在线控制策略,以保证日前调度计划有效性为目标,控制系统内可控电源功率输出,提前响应净负荷变化。最后,针对某个海岛微电网算例,制定了其日前调度计划；验证了负荷和可再生能源发电超短时预测方法的有效性；并得到在线控制策略结果；与不含在线控制策略仅按日前计划运行对比分析,验证了本文的多时间尺度调度策略的效果。
[Abstract]:Independent microgrid effectively integrates distributed power supply, energy storage device, regional load and control device to form a single controllable unit, which has a wide range of application prospects in the development of islands, deserts and other remote areas. Due to the small access and system capacity of a large number of intermittent renewable energy power generation, the randomness problem in independent microgrid is becoming more and more prominent. It is of great significance and urgent need to formulate effective system control strategy to realize the stable and efficient operation of independent microgrid. In this paper, it is considered that the fluctuation of load and renewable energy power generation in actual operation leads to the failure of the scheduling plan made in advance, on the basis of ultra-short-term prediction of load renewable energy power generation. A coordinated control strategy for controllable power output of independent microgrid is proposed. Firstly, the paper summarizes the development and construction process of the current independent microgrid system, introduces the typical structure of the independent microgrid, and establishes the mathematical model of the common distributed power supply. The control principle of controllable power supply in centralized control mode is studied. Considering the possible use of renewable energy, the principle and implementation method of maximum power tracking power generation mode are studied. Secondly, the necessity of ultra-short-term forecasting of load renewable energy generation in independent microgrid is discussed, and the ultra-short-term forecasting method of load in conventional power system is summarized and analyzed. The direct prediction method of load real-time data sequence based on least square support vector machine is adopted. Considering the intermittent of renewable energy, the same method as load forecasting is used to realize the ultra-short time prediction of wind power generation, and the direct prediction of photovoltaic power generation power is realized by similar day selection. Thirdly, a multi-time scale control strategy for independent microgrid including pre-day dispatching plan and on-line control strategy is proposed, and the network architecture of the control system is established. On the basis of satisfying load demand and power supply constraint, a centralized control model of independent microgrid is established, which takes into account many objectives, such as fuel cost, maintenance cost and emission cost. Particle swarm optimization algorithm is introduced to solve the optimal scheduling model. In this paper, the concept of net load of microgrid is introduced, and on the basis of ultra-short time prediction of net load, an on-line control strategy is proposed to control the power output of controllable power supply in the system in order to ensure the effectiveness of the dispatching plan before the day. Respond to net load changes in advance. Finally, for an example of an island microgrid, the pre-day dispatching plan is worked out, the effectiveness of the ultra-short-term forecasting method for load and renewable energy generation is verified, and the results of on-line control strategy are obtained. Compared with that without online control strategy, the effect of the multi-time scale scheduling strategy in this paper is verified by comparing and analyzing the results of the multi-time scale scheduling strategy without online control strategy.
【学位授予单位】：浙江大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TM73

【参考文献】