SEP-NMPC: Safety Enhanced Passivity-Based Nonlinear Model Predictive Control for a UAV Slung Payload System

A research team from York University has introduced SEP-NMPC (Safety Enhanced Passivity-Based Nonlinear Model Predictive Control), a breakthrough control framework designed for quadrotor drones transporting suspended, swinging payloads. The core challenge—maintaining stability while a heavy payload swings wildly and ensuring neither the drone nor the payload hits an obstacle—has been a major barrier to reliable autonomous delivery and construction. SEP-NMPC solves this by embedding two key mathematical guarantees directly into its online optimization: a strict passivity inequality for stability and High-Order Control Barrier Functions (HOCBFs) for safety.

For stability, the framework uses a shaped energy storage function with adaptive damping, which actively dissipates the excess energy from payload swings, forcing the system to converge smoothly to its target. For safety, the HOCBFs act as a mathematical force field, rendering user-defined clearance zones around both the drone and the payload as "invariant." This means the optimization is physically prevented from calculating any trajectory that would breach these safe zones, even when avoiding moving obstacles. Critically, the entire optimization remains a Quadratic Program (QP), allowing it to be solved in real-time at each control cycle without any heuristic switching or manual tuning.

The research, accepted for presentation at the prestigious ICRA 2026 conference, was validated through extensive simulations and real-world experiments. The results confirm that drones using SEP-NMPC can achieve stable, collision-free transport of slung loads across all tested scenarios. This represents a significant unification of formal stability and safety proofs within a single, practical control scheme, moving beyond ad-hoc solutions to a certifiably reliable framework.