Smart Grid

Conversion loss modeling plays a crucial role in hybrid AC/DC microgrid (MG) energy management (EM). However, accurate calculation of the conversion losses is often very costly. Additionally, existing surrogate models typically rely on fixed-voltage DC buses, leading to excessive voltage magnitudes. To overcome these limitations, we propose surrogate models based on piecewise linear neural networks (NNs) that estimate conversion losses using converter power and variable-voltage DC buses.

This dataset is created by an experimental setup of a DC-PV -Battery-based grid-connected distributed generation system. This dataset is split into four parts such as irradiance, and temperature, which were measured by a meteorological station, and lastly, PV output current and voltage acquired by an inverter. Furthermore, we can have a chance to obtain the output PV power by multiplying current and voltage. The dataset has 288 elements for one day as a time series since the station obtains the data within five minutes. However, the whole dataset has three days of data with 864 elements.

This dataset is created for neural network-based surrogate modeling of the power conversion losses. The dataset includes four sets of data (for AC/DC conversion losses under inversion/rectification moes and DC/DC conversion losses during battery charging/discharging, respectively) for the neural network. The raw data is generated using high fidelity analytical models. 

This is a dataset of the 2-area, 3-area, and 5-area interconnected power systems built in our article. By downloading them, you can use them to test and analyze the distributed optimization method presented in our article and other literature.