The areas near interchanges are often the most congested parts of a highway. In this work, we formulate the interchange management problem as an scheduling, decision-making, and optimization problem. We convert the problem onto the graph domain and adopt simulated annealing to optimize. Experimental result shows that our algorithm outperforms the first-come-first-serve strategy in terms of the total time required for all the vehicles passing the interchange area.
Please visit https://kevinchang73.github.io/ for detailed information.