元件可靠性参数是进行电力系统可靠性评估的基础,错误的元件可靠性参数必然带来错误的评估结果,进而影响电力系统规划的准确度。考虑到随着负荷侧智能电表等的普及和安装,可以获取相对准确的系统/节点可靠性指标,本文提出了一种利用系统/节点可靠性指标求取未知的元件可靠性参数的方法。首先,基于非序贯蒙特卡洛模拟(Non-sequential Monte Carlo simulation, NMCS)的可靠性评估,建立了可靠性指标关于可变可靠性参数的解析表达式,该解析表达式由给定的元件可靠性参数可以直接地获取可靠性指标值。基于该解析表达式,构建了求取未知元件可靠性参数的非线性方程组;其次,针对非线性方程组的常规求解算法普遍依赖未知变量的初值而未知参数的初值无法直接获取的难点,利用基于模拟退火算法的高阶多项式逼近来给出待求参数的初始值。最后,对于部分误差较大的参数采用具有大范围收敛特性的延拓法进行修正,最终求取准确的未知可靠性参数。分别采用RBTS, IEEE-RTS79和川渝电网测试系统进行分析,算例结果表明:所提方法能有效准确地求取未知的元件可靠性参数。
英文摘要:
Component reliability parameter is the basis of power system reliability evaluation. Wrong component reliability parameters will inevitably lead to wrong evaluation results, which will affect the accuracy of power system planning. Considering the popularity of the intelligent meter, accurate reliability index of the system can be easily obtained. This paper proposed a method to obtain component reliability parameters using reliability index of the system. Firstly, based on non-sequential Monte Carlo method, an analytical expression of the reliability index with respect to the variable component reliability parameters is established, which can directly obtain the reliability index value from the given component reliability parameters. According to analytical expression, the nonlinear equations for obtaining unknown reliability parameters of components are constructed. Secondly, in view of the difficulty that the conventional algorithms for solving the nonlinear equations generally rely on the initial values of unknown variables and the initial values cannot be obtained directly, the initial values of the parameters are given by high-order polynomial approximation based on simulated annealing algorithm. Finally, for some parameters with large errors, the extension method with wide range convergence is used to obtain accurate unknown reliability parameters. RBTS, IEEE-RTS and Sichuan-Chongqing power grid test systems are used to analyze the results. The cases show that the proposed method can effectively and accurately obtain the unknown reliability parameters of the components.