Skip to main content

Table 6 Analysis of random variable dimensions and simulation results in practical engineering

From: Planning of distributed renewable energy systems under uncertainty based on statistical machine learning

Method

Reference and Journal Title

Renewable energy dimension and data sources

Analysis of simulation results

Probability Theory

[156] Electrical Power and Energy Systems

4 wind farms in China

The PDF of WP scenarios calculated by the RVM-copula method are very similar to the empirical copula, and the correlation of small-scale WP scenarios is more accurately simulated.

Probability Theory

[32] Journal of Modern Power Systems and Clean Energy

26 wind farms in East China

The R-vine copula model is introduced to deal with the high-dimensional characteristics and correlation of WP scenarios, which is more accurate and flexible than the Gaussian copula calculation results. When the scenario dimension increases, the accuracy of this method decreases due to the limitation of computational ability.

SML

[55] IEEE Transactions on Power Systems

24 wind farms, 32 solar power plants located in the Washington from NREL

The marginal distribution of generated scenario by the model-free GAN network is closer to the actual distribution than the Gaussian copula. When the power fluctuation of renewable energy is large and the spatial correlation is enhanced, the calculation accuracy of traditional probability theory is greatly reduced.