Distributed Robust Receding Horizon Control for Multivehicle Guidance

Model Predictive Control for Parallel Planning of Conflict-free Trajectories for Multiple Aircraft

“…Protocol-based [2][3][4] or optimization-based [5][6][7][8][9][10][11][12][13][14][15] : the advantage of protocol-based algorithms is their simplicity and real-time applicability, whereas optimizationbased algorithms offer desirable trajectory or direction, which minimizes specified criterion at the cost of computational burden.…”

“…Centralized [13][14][15] or decentralized [2][3][4][5][6][7][8][9][10][11][12] : centralized algorithms are well-suited for conventional groundbased air traffic controls, while decentralized algorithms are suited for autonomous operations because of their distributed nature. Considering the recent advancements in information and communication technologies, we address optimizationbased decentralized algorithms for planning conflict-free trajectories of multiple aircraft.…”

“…Considering the recent advancements in information and communication technologies, we address optimizationbased decentralized algorithms for planning conflict-free trajectories of multiple aircraft. This class of algorithm can be further distinguished between sequential algorithms [5][6][7] and parallel algorithms. [8][9][10][11][12] In the sequential algorithms, each aircraft's trajectory is updated in a specified order.…”

“…To this end, we assume a terminal controller to perform circling at the horizon time of the MPC. Although similar approaches to guarantee the recursive feasibility are used in existing studies, 5,6) the novelty of our approach is its applicability to synchronous calculations. Moreover, the constraints on the velocity and acceleration, which are often overly simplified in existing studies, are tightly formulated in the framework of mixed-integer linear constraints (MILCs).…”

Optimisation based control framework for autonomous vehicles: Algorithm and experiment

“…In this setting, the vehicle is usually treated as a point mass and routines are generated mainly to maximize the mission success rate by taking into account the location of targets, threats and geographic information [1]. Hence trajectories generated by the high-level mission planner may not fully satisfy vehicle kinematic constraints and may also need alterations when encountering newly developed threats or obstacles.…”