Table 1 Summary of related work in PV power forecasting.

Maneesha P. et al.⁷

Particle Swarm Optimization

PV forecasting data

4 different resolutions/horizons

Assouline et al.⁸

Support Vector Machine (SVM) with kernel technique

Urban Switzerland rooftop PV data

Not specified

Preda et al.⁹

SVM algorithm

Real sensor-generated data for hybrid renewable systems

Not specified

Muhammad W. et al.¹⁰

Extra Trees (ET) and Random Forest (RF)

Hourly PV power data

Hourly

Costa¹¹

LSTM, CNN, and hybrid CNN-LSTM

Residential PV system data

Multiple horizons

Tovar et al.¹²

5-layer CNN-LSTM

Real PV data from Mexico

Not specified

Grzebyk et al.¹³

XGBoost

1,102 distributed PV systems

Hourly (daily resolution)

Ferlito et al.¹⁴

Various ML methods (complexity comparison)

PV forecasting data

Not specified

Ibtihal A.A. et al.¹⁵

Ensemble stacked ML (RF, GB, MLR)

Two PV systems (different sizes/ages)

Hourly

Asiedu et al.¹⁶

Single, ensemble, and hybrid ML models

Solar PV data

4 horizons (1 day, 1 week, 2 weeks, 1 month)

Present study

Multi-label ML (LR, PR, ANN, DL, RF, GBT, DT, k-NN, SVM)

Real data from Cairo, Egypt - BAPV plant: 1-year, 5-minute intervals, meteorological + power data

Multiple (1 day, 1 week, 1 month)