SwarmCCO: Probabilistic Reactive Collision Avoidance for Quadrotor Swarms Under Uncertainty

We present decentralized collision avoidance algorithms for quadrotor swarms operating under uncertain state estimation. Our approach exploits the differential flatness property and feedforward linearization to approximate dynamics performs reciprocal avoidance. account uncertainty in position velocity by formulating constraints as chance constraints, which describe a set of velocities that avoid collisions with specified confidence level. two different methods solving constraints: our first method assumes Gaussian noise distribution. second is its extension non-Gaussian case using Mixture Model (GMM). reformulate linear into equivalent deterministic are used an MPC framework compute local collision-free trajectory each quadrotor. evaluate proposed algorithm simulations on benchmark scenarios highlight benefits over prior methods. observe both provide improved performance method. On average, requires ~ 5 ms trajectory, while computationally more expensive 9 average 4 agents.