大规模冰灾影响下输电系统风险评估与融冰策略的研究

Power systems suffer severely operational risk and enormous economic losses in large-scale ice disaster event. A serious icing event occurred in China 2008 caused the direct economic losses over 100 billion RMB. In this project, taking predictability and defensibility of the icing event as guiding ideology, the aim is to build moderate security of power grid during ice storms. Firstly, reliability modeling of transmission line is performed, which provides a bridge between the severity of meteorological characteristic variables and the reliability of power lines. Based on the model, the reliability of power line could be tracked dynamically during ice storm according to the forecasted meteorological information. Futhermore, a short-term risk assessment of transmission grid is carried out by combining Geography Information System (defining the physical exposure of transmission grid) and the design standard (determining the vulnerability resistance for ice disaster). Simultaneously, key lines in the grid could be dynamically identified, and a timely warning mechanism and quickly de-icing strategy were proposed. According to the cost benefit analysis of risk assessment, the investment and economic cost for the security defense scenario could be determined. Aiming at the problem of large-scale towers collapse and lines breakage, the structural reliability theory was applied in analyzing the loading risk, and a optimal ice-melting strategy was made based on grid level by nonlinear programming theory..Based on the research described in this project, operators could timely obtain the risk measurement and warning information of power grid by the forecasted meteorological information during ice storms. Futhermore, a standardizing and programming de-icing measure could be conducted, and the adverse effects of the icing event may be reduced to its minimum.

大规模冰灾事件给电网带来巨大的运行风险与经济损失(2008年超过1000亿)。本项目以电网冰灾事件可预测性与可防御性为指导思想，旨在对冰灾天气下电网的适度安全性问题进行研究。首先建立气象参数特征变量的恶劣程度与元件停运风险之间的桥梁- - 元件停运风险模型。根据气象预报信息，动态跟踪元件停运风险水平，进而结合地理信息系统(界定元件物理暴露量)与设计标准(确定元件抗冰脆弱性)，实施整个电网的风险评估，动态识别电网中关键线路，并建立及时预警机制与重点防御策略。依据风险评估的成本效益分析，确定电网防御决策的工程投入与经济代价。针对大规模倒塔、断线问题，应用结构可靠性理论进行风险分析，采用非线性规划理论制定最优融冰策略。.基于本项目的研究，在应对冰灾事件时，根据气象预报信息，运行人员能够及时获取电网风险测度与预警信息，采取标准化、程序化安全防御措施，实时掌控电网风险水平，最大限度减少冰灾负面影响。

电力系统的冰灾事件会造成巨大的电网运行风险和经济损失。本项目以控制电网的运行风险为目标，旨在为构建冰灾天气下输电系统风险评估与线路融冰管理工具奠定理论基础。.1）项目对架空线路覆冰预测模型展开了研究，提出了气象过程信息的概念，基于气象过程信息和支撑向量机算法建立了覆冰回归预测模型，通过算例分析发现：与传统的预测模型相比，本项目提出的预测模型具有更高的覆冰预测精度。.2）项目对覆冰架空线路故障概率模型展开了研究。在研究线路故障时考虑模倒塔断线和冰闪跳闸这两种主要的故障类型，最后以多失效竞争风险模型对线路综合故障概率进行描述。该模型可以动态预测覆冰线路的时变强迫停运率，是进一步实施输电系统风险评估的基础。.3）项目对输电系统风险评估算法展开了研究。将拟蒙特卡洛方法应用于风险评估中，并提出系统元件的维度重要性概念，解决了该方法改进性随问题维度增大而退化的现象，使得拟蒙特卡洛方法适用于系统维数相对较大的输电系统。通过算例可知：与传统蒙特卡洛方法相比，该方法在计算精度和计算效率上都具有一定优势。.4）项目对冰灾天气下输电系统风险指标体系展开了研究。首先提出了供电失效指标，从系统运行层面和负荷点供电层面分别对停电概率、缺失电量、供电可用率等多个角度的风险进行量化。其次，提出了覆冰线路的全局灵敏度指标，对覆冰线路的重要程度进行量化。前者明确系统的风险薄弱环节，便于运行人员及时发布运行风险测度；后者量化不同覆冰线路对系统风险影响的重要程度，便于运行人员动态识别关键线路。.5）项目对直流除冰优化策略展开了研究，从降低输电系统在除冰周期中线路覆冰量和运行风险这两个角度建立了直流除冰多目标优化决策数学模型。该模型可以帮助电力运行人员提前制定合理的除冰计划。.基于本文的研究，电网的运行人员在应对冰灾事件时能够根据气象预报信息快速有效的获取电网的运行风险水平，从而采取科学的安全防御措施，减小冰灾事件的负面影响。