基于含时序项代理模型的电力系统时域仿真不确定性分析

Uncertainty Analysis of Time-domain Simulation for Power SystemBased on Surrogate Model Containing Time Series Terms

随着电力系统中不确定性因素的不断增加,传统确定性的时域仿真分析已难以满足不确定性环境下系统暂态分析的需要。为此,提出了一种基于含时序项代理模型的时域仿真不确定性分析方法。该方法采用非线性有源自回归(NARX)模型表征系统动态特性,利用多项式混沌展开(PCE)方法刻画系统内随机因素的影响,在保证计算精度的同时,提高了时域仿真不确定性分析过程的效率。同时,引入了最小角回归(LAR)策略以选取代理模型的基函数,克服了高维随机变量所引起的维数灾问题。采用标准IEEE 9节点和118节点系统进行仿真,并与传统方法对比,验证了所提方法的准确性和高效性。

With the continuous increase of uncertain factors in power systems,the traditional deterministic time-domain simulation analysis has difficulty in meeting the system transient analysis needs of the power system under uncertain situations.To this end,an uncertainty analysis method of time-domain simulation based on a surrogate model containing time series terms is proposed.This method uses the nonlinear autoregressive with exogenous input(NARX)model to characterize the dynamic characteristics of the power system,as well as using the polynomial chaos expansion(PCE)method to describe the influence of random variables in the power system.While ensuring the calculation accuracy,the efficiency of the uncertainty analysis process of the time domain simulation is increased.At the same time,the least angle regression(LAR)strategy is introduced to select the basis function of the surrogate model,which overcomes the dimensionality disaster problem caused by high-dimensional random variables.The standard IEEE 9-bus system and IEEE 118-bus system are used for simulation,and the accuracy and efficiency of the proposed method are verified compared with the traditional methods.