Monitoring Data Factorization of High Renewable Energy Penetrated Grids for Probabilistic Static Voltage Stability Assessment

The increasing integration of large number of renewable energy sources (RESs) generates multiple kinds of uncertainties in power grids, which heavily challenges traditional deterministic approaches for voltage stability assessment (VSA). Under this circumstance, this paper proposes a high-dimensional spatial-temporal matrix factorization (STMF) approach for monitoring data to separate the deterministic part and uncertain elements caused by RESs, and presents a probabilistic index to online evaluate voltage stability. Specifically, this method explores the eigenvalue distribution of measurement data when penetrating multiple distributed wind and solar generations, as well as a learning-based STMF model is accordingly used to iteratively extract the random elements due to RESs. Unlike traditional matrix factorization approaches, this STMF method uses an adaptive and analytical model to depict the eigenvalue distribution of random components. Subsequently, a probabilistic index is designed to measure the voltage stability degree as well as provide quantiles to indicate its potential fluctuations. Numerous cases verify the effectiveness and advantages of the proposed method using different load changing models and various RES penetration levels, as well as other probabilistic and deterministic indicators are compared.

Automated summary

The increasing integration of renewable energy sources in power grids has created uncertainties that challenge traditional deterministic approaches for voltage stability assessment. This paper proposes a high-dimensional spatial-temporal matrix factorization approach to separate the deterministic part and uncertain elements caused by renewable energy sources, and presents a probabilistic index for online evaluation of voltage stability. The proposed method uses an adaptive and analytical model to describe the eigenvalue distribution of random components, and has been proven effective in various scenarios.