An Interior Point Algorithm for Optimal Coordination of Automated Vehicles at Intersections
In this paper, we consider the optimal coordination of automated vehicles at intersections under fixed crossingorders. We state the problem as a Direct Optimal Control problem, and propose a line-search Primal-Dual Interior Point algorithm with which it can be solved. We show that the problem structure is such that most computations required to construct the search- direction and step-size can be performed in parallel on-board the vehicles. This is realized through the Schur-complement of blocks in the KKT-matrix in two steps and a merit-function with separa- ble components. We analyze the communication requirements of the algorithm, and propose a conservative approximation scheme which can reduce the data exchange. We demonstrate that in hard but realistic scenarios, reductions of almost 99% are achieved, at the expense of less than 1% sub-optimality.
Model Predictive Control