In this paper, Nonlinear Model Predictive Control (NMPC) and Nonlinear Moving Horizon Estimator (NMHE) are combined to control, in a distributed way, a heterogeneous fleet composed of a steering car and a quadcopter. In particular, the ground vehicle in the role of the leader communicates its one-step future position to the drone, which keeps the formation along the desired trajectory. Inequality constraints are introduced in a switching control fashion to the leader’s NMPC formulation to avoid obstacles. In the literature, few works using NMPC and NMHE deal with these two vehicles together. Moreover, the presented scheme can tackle noisy, partial, and missing measurements of the agents’ state. Results show that the ground car can avoid detected obstacles, keeping the tracking errors of both robots in the order of a few centimeters, thanks to trustworthy NMHE estimates and NMPC predictions.

Motivated by a traffic management application, this work proposes a novel robust optimal control framework for networks subject to disruptions. It minimizes deviations from a target vehicle configuration while enforcing bounds on the expected values of quadratic constraints including states and controls, and fulfilling balance equations constraints. The problem uses a recent Linear-Quadratic (LQ) framework with stochastic quadratic constraints. Validation via two case studies shows that the method, despite disruptions, maintains proximity to references and satisfies all constraints, outperforming a traditional LQ controller and a heuristic controller, both of which lack the stochastic constraints. Furthermore, offline verification checks the covariance matrices' asymptotic limits to guarantee bound compliance.

Tight formation pursuit is a common problem for small multi-robot systems navigating narrow, cluttered environments. In this context, this work proposes centralized and decentralized Nonlinear Model Predictive Control (NMPC). The highly nonlinear cost function and the predictive capabilities of the NMPCs allow the agents to react promptly and avoid collisions while maintaining a tight formation with the leader. In the centralized scheme, followers utilize a Kalman Filter (KF) to estimate the leader's speed. This estimate is then used by the NMPC to predict the leader's motion, assuming a constant velocity over the prediction horizon. Similarly, in the decentralized scheme, each agent solves its own optimization problem, predicting the other robots' future behavior starting from the speeds estimated by a local KF. Both schemes have been extensively validated through indoor experiments involving various leader trajectories and data loss scenarios. The results show that the ground vehicles effectively adapt their formation to maintain a close distance to the leader without collisions.