GA-MPC based multi-objective optimal control of overhead cranes for position tracking and sway suppression

As core intelligent equipment in modern automated warehouses and high-duty industrial production, the tracking accuracy and swing angle suppression of overhead cranes directly determine logistics throughput, energy efficiency, and operational safety. In practical engineering applications, traditional methods like Linear Quadratic Regulators (LQR) often fail to balance these objectives under uncertain payload variations, while conventional Model Predictive Control (MPC) suffers from a heavy reliance on manual trial-and-error for weight tuning. The switching term in Adaptive Sliding Mode Control(ASMC)’s control law causes high-frequency jitter. To address this, this paper proposes an intelligent cooperative control strategy based on a Genetic Algorithm (GA)-optimized MPC. By incorporating real-world physical constraints (such as actuator saturation) and industrial duty-cycle requirements, a nonlinear dynamic model with load-swing coupling is established. A dual-objective MPC framework is then constructed, leveraging GA to eliminate manual tuning dependency and achieve global optimization. For computational efficiency, the controller’s internal prediction model employs small-angle linearization, while its robustness against plant-model mismatch and nonlinear sway dynamics is strictly validated using a fully coupled nonlinear physical plant. The strategy’s applicability to common industrial scenarios is validated through multi-scenario simulations under stochastic variations in rope lengths and payload masses. Results demonstrate that the strategy significantly enhances performance—reducing tracking errors by 32.74% and swing angles by 14.44%, while lowering control energy consumption, offering a reliable, and generalizable solution for fully autonomous material handling systems.