分割多目标风险分析框架下不同风速模拟方法对电力系统可靠性评估的影响

Impacts of Wind Speed Simulation Methods on Adequacy Evaluation of Power Systems in the Partitioned Multi-objective Risk Analysis Framework

比较不同风速模拟方法对可靠性评估的影响,为风速模拟方法的选择提供依据。首先,根据实测风速数据研究了风速的季节特性、日特性、自相关性等宏观特性,分析并比较了实测风速法、实测风速均值法、概率抽样法、多状态等效法和时间序列模拟法在模拟风速宏观特性上的差异。然后,针对无条件期望可靠性指标的不足,提出了充裕度评估中条件可靠性指标体系,并给出了分割多目标风险分析框架下基于蒙特卡洛方法的电力系统充裕度评估过程。最后,对加入风电的IEEE-RTS测试系统进行了可靠性评估。算例结果表明:时间序列模型计算出的无条件及条件可靠性指标与实测风速法相比差别较小;概率抽样法计算出的系统状态指标呈高频率、低持续时间的特点;风速均值法计算出的健康状态条件概率偏大,而临界状态和风险状态的条件概率偏小,且健康状态呈现明显的低频率、长持续时间的特点。

The impacts of wind speed simulation methods on reliability assessment were discussed to provide useful reference for selection of those methods. First, the seasonal and diurnal characteristics, autocorrelation were investigated based on the observed wind speed data. The advantages and dis- advantages of the observed wind method, the mean observed wind speed method, the probability sampling method, the multi-state equivalent method and the time series modelling method were analyzed. Second, aiming at the disadvantages of the unconditional expected values of the reliability indices, the conditional reliability indices for adequacy evaluation were proposed, and the reliability assessment process of power systems based on the partitioned multi-objective risk method （PMRM） was discussed. Finally, the modified IEEE-RTS including wind generation was evaluated based on the PMRM. Numerical results show that there is little difference in the unconditional and conditional reliability indices between time series methods and the observed wind speed method; the conditional reliability indices of system states using the probability sampling method have the characteristic of high frequency and short duration; the conditional probability of the healthy state using the mean observed wind speed method is larger whereas the conditional probabilities of the marginal and at-risk states are smaller compared with using the observed wind speed method, and the healthy state presents the pro- perties of low frequency and long duration.