The emergence of IoT-based sensor networks has revolutionized industries, enabling data-driven decision-making. However, the escalating demand for energy-efficient solutions poses a pressing challenge. This paper proposes integrating Particle Swarm Optimization (PSO) with Deep Q-Learning (DQL) to enhance energy efficiency in IoT systems. The methodology involves optimizing DQL model parameters to predict energy consumption and improve residual energy management. Experimental results demonstrate significant enhancements in energy efficiency metrics, validating the effectiveness of the proposed approach. Future research will focus on real-world implementations and additional optimization techniques for further improvements.