The Home Energy Management System (HEMS) has, as one of its primary objectives, the reduction of energy consumption to achieve greater energy efficiency. This study presents a novel method that integrates neural network-based weather forecasting with the hybrid optimization algorithm Particle Swarm Optimization and Differential Evolution (PSODE) to enhance decision-making processes influenced by environmental variables. The weather forecasting component relies on a neural network model trained on historical data from the NASA Power API. The resulting data streams provide a weather outlook that is not only reliable but also continuously updated. This forecast is then incorporated into the PSODE algorithm, which directly influences the decisions made by swarm particles based on predicted weather conditions. Moreover, the integration of weather forecasting techniques into a dynamic real-time system has been validated. The next step involves incorporating these forecasts into the PSODE algorithm, where the swarm particles’ positions are adjusted based on predicted weather patterns. By modifying its parameters in response to specific weather predictions, the proposed method enhances flexibility and effectiveness in highly dynamic environments. Simulation results demonstrate that this PSODE-based approach significantly improves optimization performance, providing enhanced efficiency and adaptability.