基于SMP状态空间等值的智能变电站保护系统短期可靠性评估

Short-term reliability assessment of smart substation protection system using SMP state space equivalent model

电气设备短期可靠性为电力系统风险评估提供数据支持,状态空间法中传统转移频率守恒等值算法基于稳态指标定义,不适用于短期可靠性评估,故需进行改进.本文研究等值条件,首次证明满足等值条件的系统等值后仍为马尔可夫过程,暂态状态概率可由矩阵相乘得到.不满足等值条件的系统,提出用半马尔可夫过程描述.确立等值前后转移率及分布函数的关系,推导半马尔可夫核矩阵与等值后转移率、分布函数的关系,以矩阵描述形式求得暂态状态概率和灵敏度指标.运用所提算法对一智能变电站保护系统进行短期可靠性评估,结果表明传统算法低估系统可用率,暂态灵敏度的峰值和出现时间也存在误差.初始状态影响系统暂态可用率以及暂态灵敏度的收敛速度.

The risk assessment of power systems is supported by the short-term reliability of electrical equipments. Due to its steady definition, the traditional frequency equilibrium in state space method is not suitable for short-term reliability assessment, which should be improved. In this paper, the equivalent condition is studied. It is proved that for the systems satisfying the equivalent condition, the instantaneous state probabilities are multiplied by matrices. Otherwise, an equivalent algorithm is newly proposed. The equivalent systems are described by the semi-Markov process. The relationships of the transition rates and distribution functions before and after equivalence are found. The kernel elements are derived from them. The instantaneous state probabilities and sensitivities are derived by matrix form from the kernel matrix. An example of short-term reliability assessment for smart substation protection system is presented. Numerical results show that the traditional method underestimates the availability. Errors also occur in the peak value and time of the instantaneous sensitivity. The state probabilities and convergence rate of sensitivities are significantly influenced by the initial state.