Traffic Aware Many-Objective Dynamic Route Planning

We study 4-objective dynamic route planning based on real road network and real traffic data from Aarhus, Denmark. Our 4 objectives include total vehicular emission cost (TEC), travel time, number of turns, and distance. The first two objectives are time-dependent and traffic-dependent. The third objective is a novel objective for simpler routing and better user experience. The fourth objective is standard for all route planning problems. We provide a thorough explanation of how to design NSGA-II to solve the 4-objective dynamic route planning problem. Our experiments produce a diverse set of solutions for this problem and provide user the flexibility of selecting a path based on their preferences of the four objectives. Our work provides a good framework for many-objective dynamic route planning and can be extended to other techniques such as many-objective Particle Swarm Optimzation (PSO) and many-objective Ant Colony Optimization (ACO).