Regression-Based Prediction of Power Generation at Samanalawewa Hydropower Plant in Sri Lanka Using Machine Learning

Table 9

Comparison of previous related studies.

Ref	Country of study	Input variables	Modeling technique	Performance of the models
[3]	Ghana	Temperature and rainfall	Statistical analysis	—
[6]	Sri Lanka	Rainfall	ANN (LM)	R = 0.86; MSE = 1.03 × 10^6
			ANN (BR)	R = 0.73; MSE = 8.9 × 10^3
			ANN (SCG)	R = 0.76; MSE = 7.42 × 10^5
[7]	Nigeria	Evaporation losses, reservoir inflow, storage, reservoir elevation, turbine release, net generating head, plant use coefficient, tail race level	ANN	R = 0.89
[8]	Brazil	Rainfall at seven subbasins	Group method of data handling (GMDH)	R = 0.90; MAE = 443; MAPE = 12.34%
			ANN (BR)	R = 0.88; MAE = 450; MAPE = 12.41%
			ANN (LM)	R = 0.83; MAE = 593; MAPE = 17%
[9]	Ghana	Rainfall, ENSO, lake level elevation, and net lake inflow	Stepwise multiple regression	R^2 = 0.753; Adjusted R^2 = 0 .742