ABSTRACT: -The growing integration of renewable energy sources such as solar and wind into smart grids presents a range of operational challenges due to their variability and intermittency. These challenges include voltage and frequency instability, power quality issues, and reduced system efficiency. This research proposes a hybrid forecast-control framework optimized using Multi-Objective Particle Swarm Optimization (MOPSO) to enhance grid stability, efficiency, and renewable energy utilization. The proposed system integrates three key components: (1) machine learning-based short-term forecasting of renewable generation using LSTM and XGBoost models; (2) intelligent real-time control using reinforcement learning to dynamically manage power flows and grid parameters; and (3) MOPSO for optimizing multiple grid objectives, including voltage stability, power loss minimization, and maximized RES penetration. Simulations conducted on modified IEEE 33-bus and 69-bus test systems with varying levels of renewable penetration demonstrate the framework’s effectiveness. Results show improved voltage profiles, increased renewable utilization (up to 28%), and reduced power losses (up to 35%) compared to conventional control strategies. The MOPSO algorithm successfully generates Pareto-optimal solutions, enabling flexible trade-offs for decision-makers. The framework also incorporates a second-assessment review layer for enhanced reliability under uncertain conditions and offers pathways for further integration of EVs, demand-side management, and adaptive market models. This study demonstrates how a hybrid MOPSO-optimization approach can significantly advance smart grid performance, providing a scalable solution for modern, resilient, and sustainable energy systems.

Key words: Smart Grid Optimization, Renewable Energy Integration, Multi-Objective Particle Swarm Optimization (MOPSO), Machine Learning Forecasting, reinforcement Learning Control, Grid Stability and Efficiency