Predictive Analysis of Market Clearing Price in the Indian Electricity Market Using Time Series Algorithms

His paper investigates the effectiveness of four advanced time series forecasting models-LSTM., SARIMA, XGBoost, and Facebook Prophet-in predicting market-clearing prices in the dynamic Indian electricity market. Addressing the challenge of forecasting with voluminous and nonlinear data, this study employs diverse methodologies, from statistical to deep learning techniques, to capture temporal dependencies. Utilizing a comprehensive dataset from the IEX (India Energy Exchange Limited), which encompasses approximately 98% of the traded volume, these models are applied to weekly, monthly, and yearly data. Through extensive hyperparameter tuning and data preprocessing, each model's performance is assessed, highlighting specific strengths and limitations. This comparative analysis offers insights into selecting the most suitable model based on data characteristics, with implications for enhancing forecasting accuracy in volatile markets.