POTENSI PLTB BERDASARKAN PERAMALAN CUACA MENGGUNAKAN METODE B-WEMA OPTIMASI LEVENBERG-MARQUARDT

The increasing demand for electrical energy in Indonesia, driven by population growth and energy consumption, necessitates a shift from fossil fuels, which contribute to greenhouse gas emissions and environmental damage. Renewable energy sources, such as solar, hydro, wind, and geothermal, offer environmentally friendly solutions, with wind energy being a key alternative. Bogor Regency, particularly Leuwiliang District, has significant potential for the development of Wind Power Plants (PLTB). This study proposes using Brown’s Weighted Exponential Moving Average (B-WEMA) optimized with Levenberg- Marquardt (LM) to predict the potential of PLTB based on weather forecasting. B- WEMA is selected for its high accuracy, while LM is employed to enhance prediction precision. The results indicate that B-WEMA with LM optimization provides better accuracy compared to B-WEMA without optimization, with the minimum difference between actual and predicted data being 5.946%, the smallest Mean Squared Error (MSE) of 0.0636, and a Mean Absolute Percentage Error (MAPE) of 0.109%. For wind speed, B-WEMA with LM optimization shows an average deviation of 10.92%, compared to 8.01% with B-WEMA without optimization. Parameters such as air pressure, temperature, and humidity also show improved results with B-WEMA optimized by LM. However, the variability in weather predictions still presents challenges, indicating a need for further attention in the forecasting methodology.
Keywords: Weather, PLTB, Brown's Weighted Exponential Moving Average (B- WEMA), Levenberg-Marquardt Optimization.