Sampling-Based Safety-Critical Control for Planar Quadrotors with Suspended Payloads

Safe trajectory tracking for quadrotors with suspended payloads remains challenging due to nonlinear dynamics, safety constraints, and real-time computation requirements. We present a novel safety-aware control framework for underactuated quadrotor-payload systems that concurrently addresses safety-critical control and real-time computation. Our approach integrates Control Barrier Functions (CBFs), chance constraints, and Model Predictive Path Integral (MPPI) control into a unified framework. By reformulating safety constraints as differentiable cost terms within MPPI's sampling-based optimization, our method avoids the computational overhead of traditional CBF-QP approaches. Moreover, incorporating chance constraints provides probabilistic safety guarantees, ensuring robust performance under real-world disturbances. We validate our framework through extensive simulations of planar quadrotors with suspended payloads in obstacle avoidance scenarios. Experimental comparisons reveal up to 2.5x fewer safety violations than standard MPPI and up to 2x faster time-to-finish than the MPPI with CBF only controllers, even under lateral and stochastic wind gust disturbances. To our knowledge, this is the first implementation combining CBFs, chance constraints, and MPPI for quadrotor-payload systems, offering significant promise for safe and efficient aerial transport in practical deployments.