<p>The mining retractable belt conveyor is widely used in underground coal mines. However, due to the frequent advancement of the coal face and the repeated retraction and release of the belt, the traveling trolley often deviates and terrain undulations occur, leading to uneven force distribution on the conveyor belt within the storage silo and subsequently causing misalignment. To address this issue, this paper suggests an automatic deviation correction device and an intelligent control system specifically designed for the working conditions of storage belt silos. Firstly, the key structural components of the correction device are designed and theoretically analyzed, followed by finite element simulations, which confirm that the device meets strength requirements under extreme working conditions. Then, the deflection control strategy is developed by designing the correction process and optimizing speed parameters. To enhance control adaptability under dynamic conditions, a particle swarm optimization (PSO) algorithm is employed to fine-tune the parameters of a conventional PID controller. Based on this, a fuzzy control strategy is integrated to construct a PSO–fuzzy PID controller with real-time self-adjustment capabilities. Comparative simulations in MATLAB demonstrate that the proposed controller outperforms both traditional PID and PSO-optimized PID controllers in terms of accuracy, responsiveness, and robustness. Furthermore, the controller is applied to the structural control of the correction device through a SolidWorks–MATLAB–Adams co-simulation platform. Simulation results reveal that the system achieves fast and stable responses, with no overshoot, even under continuous angular adjustment commands and external disturbances. Therefore, this study provides a novel and effective solution for intelligent deviation correction in belt conveyors operating within storage silo environments.</p>