STATISTICAL ANALYSIS AND ANN-BASED PREDICTION OF WIND POWER GENERATION: A CASE STUDY OF PEMBA ISLAND, ZANZIBAR, TANZANIA.

Hamza Khamis Kombo
Ridho Irwansyah

Multidiciplinary Output Research for Actual and International Issue (Morfai Journal)
Vol. 6 No. 2 (2026): Multidiciplinary Output Research For Actual and International Issue

This research work proposes the integrated statistical and ANN approach for the evaluation and prediction of the wind power potential of Pemba Island, Zanzibar, Tanzania. Monthly wind speeds and meteorological data from NASA POWER were used to assess the wind power potential of the region using ten consecutive years of data (2014-2023). From the results, it was evident that the region has marginal to moderate wind power potential since the wind speed averages 6m/s. Wind speed in the region is high between May and August. Wind power density in the region was between 130.46 and 170.43 W/m2. This information proved that the region has the ability to implement small to medium-scale and hybrid wind power systems. An artificial neural network was also developed to predict the wind power density using various parameters such as wind speed, temperature, relative humidity, wind direction, and precipitation. From the results, the ANN has excellent predictive ability since the value of R2 was approximately 0.9998, and very low levels of error for the prediction of the wind energy potential of the islanded grid.