计及风电概率分布特征的鲁棒实时调度方法研究

发布时间：2018-06-21 22:10

本文选题：鲁棒优化 + 实时调度　；参考：《山东大学》2017年硕士论文

【摘要】：风力发电与常规电源相比具有明显的不可控性,大规模的风电并网将使电力系统运行中的不确定性进一步增强,加重电力系统备用负担。同时,考虑风电的有功调度问题在运用不确定性优化方式进行处理时,算法的实用性与结果的最优性之间存在矛盾,有待进一步深入研究。针对这一现状,当前阶段,大量研究将优化数学领域的先进方法应用于电力系统的优化调度问题,取得了较好的应用效果。其中,鲁棒优化是一种利用区间扰动信息,在最劣扰动条件下进行最优决策的不确定性优化方法,具有基础数据易得、计算效率高、适用于大规模系统求解等优点,对于多类扰动集合具有较好的计算性质。为实现对可接纳扰动集合边界的统筹优化,需要进一步解决鲁棒优化的调度算法利用不确定量的概率分布信息的关键问题。本文是电力系统鲁棒调度理论在计及风电概率分布信息条件下的完善与发展,进行的主要工作如下。首先,本文在阐明鲁棒优化自身特点的基础上,对鲁棒优化方法在电力系统机组组合问题中的应用进行了介绍,阐述了连续性、偶发性扰动模式下的鲁棒优化方法建模规律;然后,介绍了鲁棒优化方法在经济调度问题中的研究现状,介绍了三类典型方法,包括自适应鲁棒优化方法、含仿射矫正过程的实时调度鲁棒优化方法和最大化可接受扰动范围鲁棒优化方法,并对其各自的特点进行了阐述;最后,对该领域研究面临的关键问题和未来的发展方向进行了探讨和分析。然后,针对高比例风电接入的具体问题,分析了传统鲁棒调度中不确定集的构成方式及保守性控制方法,进而,分析了不确定集在描述系统高比例可再生能源接纳能力时的不足及改进方向,给出风电功率接纳的条件风险价值指标,研究提出了一种计及风电功率概率分布特征的鲁棒实时调度方法,依据系统运行成本与风电接纳条件风险价值(conditional value-at-risk,CVaR)双重指标,按照鲁棒优化的建模思路构建优化模型,对自动发电控制机组的运行基点及备用容量进行决策,获得具有统计意义的节点可接纳风电功率的范围。进而,为了使优化调度结果更加符合电力系统的运行实际,将模型中的参与因子同时作为变量进行决策,并分别利用启发式交替迭代算法和大M法对由此导致的非线性问题(双线性项)进行处理,实现了运行基点与参与因子快速有效地协调决策。方法体现了鲁棒优化和随机规划方法的融合,调度结果具有概率优性,能有效提高电网运行效益且计算效率较高。最后,对简单6节点系统及IEEE 118节点系统的计算与分析验证了所提出方法的有效性和计算效率。
[Abstract]:Wind power generation has obvious uncontrollability compared with conventional power supply. The large-scale wind power grid connection will further enhance the uncertainty in power system operation and increase the reserve burden of power system. At the same time, there is a contradiction between the practicability of the algorithm and the optimality of the result when considering the active power scheduling problem of wind power, which needs to be further studied. In view of this situation, at present, a large number of advanced methods in the field of optimization mathematics have been studied and applied to the optimal dispatching problem of power system, and good results have been obtained. Among them, robust optimization is an uncertain optimization method for optimal decision making under the condition of the worst disturbance and interval perturbation information. It has the advantages of easy access to basic data, high computational efficiency, and suitable for solving large-scale systems. It has good computational properties for multi-class perturbation sets. In order to realize the overall optimization of the boundary of admissible perturbation set, the key problem of robust optimal scheduling algorithm using the probability distribution information of uncertain variables needs to be solved further. This paper is the improvement and development of the robust dispatching theory of power system under the condition of taking into account the probability distribution information of wind power. The main work of this paper is as follows. Firstly, on the basis of clarifying the characteristics of robust optimization, this paper introduces the application of robust optimization method in power system unit commitment problem, and describes the modeling law of robust optimization method under the mode of continuous and accidental disturbance. Then, the research status of robust optimization method in economic scheduling problem is introduced, and three kinds of typical methods, including adaptive robust optimization method, are introduced. The robust optimization method for real-time scheduling and the robust optimization method for maximizing acceptable disturbance range including affine correction process are presented and their respective characteristics are described. The key problems in this field and the future development direction are discussed and analyzed. Then, aiming at the specific problem of high proportion wind power access, the paper analyzes the structure of uncertain sets and conservative control methods in traditional robust scheduling, and then, In this paper, the shortcomings and improvement direction of uncertain sets in describing the acceptance capacity of high proportion renewable energy of the system are analyzed, and the conditional risk value index of wind power admission is given. In this paper, a robust real-time scheduling method based on the characteristics of wind power probability distribution is proposed. According to the double indexes of system running cost and conditional value-at-risk value (Cvar), the optimization model is constructed according to the idea of robust optimization. The operating base point and reserve capacity of the automatic generation control unit are determined and the range of wind power acceptable to the node with statistical significance is obtained. Furthermore, in order to make the optimal dispatching results more in line with the actual operation of the power system, the participation factors in the model are taken as variables to make the decision at the same time. The heuristic alternating iteration algorithm and the large M method are used to deal with the nonlinear problem (bilinear term), which realizes the fast and effective coordination decision between the basic point and the participation factor. The method embodies the combination of robust optimization and stochastic programming. The scheduling results are probabilistic optimality, which can effectively improve the efficiency of power grid operation and the efficiency of calculation. Finally, the calculation and analysis of simple 6-bus system and IEEE 118-bus system verify the effectiveness and efficiency of the proposed method.
【学位授予单位】：山东大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TM73

【参考文献】