大电力系统可靠性评估的蒙特卡洛仿真及概率密度估计

Bulk Power System Reliability Assessment Using Monte Carlo Simulation and Probability Density Estimation

详细阐述了用于大电力系统可靠性指标预测的两种不同蒙特卡洛仿真方法,即序贯蒙特卡洛仿真(状态持续时间抽样)和非序贯蒙特卡洛仿真(状态抽样)的基本原理,对这两种方法的收敛特性及其计算精度和样本容量的概率不确定性关系进行了深入研究。此外,基于核密度估计技术,实现了可靠性指标的概率密度估计,探索了从可靠性指标内在分布规律和结构特征出发深刻揭示电网风险特性的新思路。基于RBTS、IEEE-RTS79和IEEE-RTS96可靠性测试系统的计算分析表明了所提方法的有效性。

The fundamental principles of two different Monte Carlo simulation techniques are studied to predict reliability indices of bulk electric power systems. The two Monte Carlo simulation techniques designated as the sequential and non-sequential methods are utilized. The convergence performance of the. two methods and the probabilistic uncertainty relation between sample sizes and calculation accuracy are explored. By utilizing the kernel density estimation technique the pictorial representation of probability density distribution is realizcd for reliability indices. The probability density estimations of reliability indices facilitates us more complete understanding of system risk from the internal distribution laws and their structural features. The proposed method is verified using RBTS, IEEE-RTS79 and IEEE-RTS96 test systems.