<p>The rapid growth in global energy demand and the increasing penetration of renewable energy sources (RESs) have accelerated the deployment of microgrids (MGs) as decentralized, resilient, and sustainable energy systems. Despite their advantages, the intermittent nature of renewables and load variability introduce significant operational and control challenges. Conventional heuristic optimization techniques have been employed to address these complexities; however, they often exhibit limitations in handling nonlinear, non-convex, and multi-objective optimization problems. MHOAs have gained significant attention as efficient optimization tools due to their strong global search ability and adaptability in addressing complex optimization problems in MGs. This review provides a comprehensive analysis of recent developments in the application of MHOAs for techno-economic optimization, energy management, resilience improvement, and fault detection within MG systems. Moreover, this study incorporates machine learning (ML) methodologies into the MG management framework. Various ML paradigms—including supervised, unsupervised, reinforcement, and deep learning (DL) techniques—are systematically classified and comparatively assessed in terms of predictive capability, scalability, data dependency, and suitability for real-time implementation. The integration of MHOAs with ML techniques forms an intelligent and adaptive framework that strengthens the sustainability, reliability, and operational performance of next-generation MG systems, while also highlighting existing challenges and potential avenues for future research.</p><p><b>Clinical trial number:</b> Not applicable.</p>