Innovative Solar Ironing System: Combining Renewable Energy with Advanced Deep Learning Techniques

This research study introduces an innovative solar-based ironing system that integrates renewable energy with advanced deep-learning techniques to optimize energy consumption in household ironing operations. The system harnesses solar panels to collect solar energy, ensuring efficient ironing during daylight hours. In cases of insufficient solar generation, the system automatically switches to a conventional power supply, guaranteeing uninterrupted functionality. A key feature of this system is incorporating a Long Short-Term Memory (LSTM) model, which predicts solar power output based on historic data and real-time climate information sourced from the OpenWeatherMap API. This predictive capability enhances energy management by providing timely insights into expected solar energy availability, allowing users to plan ironing tasks effectively. Data preprocessing techniques, including imputation of missing values using the median and normalization, ensure the quality and reliability of the input data for training the LSTM model. The model is assessed using various performance metrics and compared to traditional forecasting methods, including ARIMA, Gated Recurrent Units (GRU), and Feedforward Neural Networks (FNN). The results demonstrate that the LSTM model significantly outperforms these alternatives in predicting solar energy generation. This research contributes to sustainable living by promoting renewable energy usage in everyday tasks and illustrates the potential of combining deep learning with solar technologies for improved energy efficiency in household applications.