电力系统仿真不确定性评估中主导参数的选择

Selection of Key Parameters in Uncertainty Evaluation for Power System Simulation

电力系统模型参数的不确定将给仿真结果带来不确定性,概率分配法(probabilistic collocation method,PCM)通过少数几次仿真建立起仿真结果与不确定性参数之间的多项式关系,为仿真不确定性评估提供了有力的工具。但随着不确定性参数个数的增加,PCM所需仿真次数将呈指数增长,为减小计算量,有必要识别出主导不确定性参数。提出了一种基于一阶PCM拟合的主导不确定性参数的选择指标,克服了传统灵敏度指标依赖于参数取值,以及无法考虑参数概率分布和取值范围等缺点,能有效地选出主导不确定性参数。进而仅对主导不确定性参数进行分析,提高PCM的适应性。算例结果验证了所提指标的有效性。

Uncertain parameters that exist in power system will bring uncertainty to simulation result. Probabilistic collocation method （PCM） can establish a polynomial relationship between the simulation result and uncertain parameters through a few simulations on smartly chosen parameter values, and provides a powerful tool for uncertainty evaluation. Although PCM is touted as a very economic technique, computational complexity nevertheless grows exponentially with the number of uncertain parameters. Key uncertain parameters which influence the uncertainty of result most should be recognized to reduce computational complexity. An index for key parameter selection based on first-order PCM fitting is proposed, which can overcome the shortcomings brought by traditional sensitivity method. For example, the sensitivity method relies on the value of parameter. Besides, it can＇t take into account the probability distribution and vary range of parameter. Polynomial relationship can be established with a few key parameters through PCM. Thus the adaptability of PCM is improved. Testing results of case studies demonstrate the validity of the proposed method.