Probabilistic Power Flow Analysis Based on Low Rank Approximation and Principle Component Analysis

Abstract

Probabilistic power flow (PPF) analysis is usually applied for evaluating the effects of uncertain parameters on power system performances. This paper presents a technique to enhance the arbitrary polynomial chaos expansion (aPCE) based PPF analysis technique when applying to system with many uncertain parameters. The proposed method represents a power system response as low rank approximation (LRA). In addition, the principle component analysis (PCA) is applied to reduce the number of uncertain parameters and also de-correlate them. This combination enables the proposed method to perform PPF of the power system having large number of uncertain parameters. Based on preliminary numerical results on the modified IEEE 57-bus system, it can be noticed that the proposed modified method is able to find accurate statistical characteristics of the responses but uses less computation time compared to the MCS based PPF analysis.