基于脉冲微分动力系统理论的风电功率预测的极大误差评估方法研究

This project evaluates the errors of power generation prediction during the development of large-scale and highly-concentrated wind power, especially the effect on the power system frequency stability when a very large error pulse appears during switching. The achievement of the research can be used to optimize scheduling of power system which contains wind power. It can also be used to improve acceptance of the power grid to large-scale wind power plant. In order to analyze the error characteristics of typical wind power prediction model, we did massive data mining, analysis and processing. Then we revealed the rules and characteristics of very large error pulse in wind power prediction. The error pulse of wind power prediction can be random, uncertain and time-vary for many reasons, like ramp meteorology. And the time characteristics as well as the capacity of backup power source also vary from time to time. Thanks to the pulse differential dynamical systems theory, we created a mathematical model of the dynamic process of power system frequency, which takes all those factors into consideration. This model can be used to analyze the power system frequency dynamic behavior caused by very large error pulse of wind power prediction, and be used to evaluate and organize the power system planning and scheduling. The model can also be used in providing the boundary conditions of the stable operation of wind power system, ensuring large-scale wind power works efficient and reliable. The projects plays an active role in improving wind power forecasting models, as well as management, development and improve the wind power prediction error evaluation system.

评估大规模高集中风电开发应用中风电功率预测误差，尤其是极大误差脉冲对接入电力系统频率稳定性的影响，有助于优化含风电电力系统的运行调度，提高电网对大规模风电的接纳能力。项目通过风电场海量数据挖掘、分析与处理，分析典型风电功率预测模型的误差特征，揭示风电功率预测出现极大误差脉冲的一般规律和特征；由于爬坡气象等因素造成的风电功率预测极大误差脉冲具有随机性、不确定性和时间性，考虑电力系统备用电源投切的时间特性和容量大小，基于脉冲微分动力系统理论和方法，建立描述电力系统频率动态过程的数学模型，分析风电功率预测极大误差脉冲引起的电力系统频率动态行为，用于评价和指导风电接入电力系统规划和调度运行，提供含风电电力系统的稳定运行的相应边界条件，实现大规模风电的高效可靠利用。项目研究也对改进风电功率预测模式和管理方法、发展和完善风电功率预测误差评价体系起到积极作用。

随着大规模风电基地的不断开发、并网风电装机容量日益增大，大规模高集中度的风电场群功率预测误差对电力系统的影响变得更为显著。.作为风电场群有功功率变化的一种表现形式，极大误差是指风电场群的功率预测误差在短时间内突然发生的大幅变化。本文为合理高效评估极大误差对电力系统运行的影响，做了以下工作：.第一，对实际风电功率数据，研究了功率极大误差总体统计特性，针对预测误差短时骤变这一有别于常规统计的特殊问题，采用极值统计方法，研究了不同风电并网容量、不同时间尺度下风电功率预测误差的统计特性，指出了在多数情况下风电功率预测误差有限，在极少部分情况下的极大误差会远远超出正常范围。基于此，将风电场群功率预测误差分为日常误差和极大误差，并分别给出了定义方式。.第二，指出应当重点分析极大误差对系统充裕性的影响，基于极大误差中风电功率变化速率大、常规机组爬坡率易构成约束、前后时刻关联性强的特性，提出了基于多时刻系统潮流计算的仿真方法，通过模拟日前风电预测、发电计划制作、自动发电控制、系统削减部分负荷等环节，模拟出系统对极大误差的反应，从而评估极大误差对系统充裕性的影响。仿真结果表明，所用的方法能充分计及风电功率预测误差的时间序列特性和线路传输容量约束，能比传统方法更细致地反映系统特性。.第三，指出应当重点分析日常误差对系统经济性的影响，基于日常爬坡中风电功率变化速率小、风电功率变化幅值大、样本数量较多的特性，提出了基于误差初始时刻系统潮流和常规机组调节能力计算各时刻在安全/经济约束下的可接受的风电功率预测误差极值的方法，通过求取削减负荷电量、弃风电量、常规机组增发高成本电量、常规机组减发低成本电量以及基于以上各电量的综合成本作为日常误差对系统经济性影响的指标量，从而评估日常误差对系统经济性的影响。仿真结果表明，提出的方法避免了复杂的动态计算，同时保留了系统的网络参数信息。