Multi-agent Path Planning Problem Under a Multi-objective Optimization Framework

In this study, a mixed-integer programming formulation is developed for a team of homogeneous sensing agents under a bi-objective optimization framework to solve a discrete open-loop centralized multi-agent search and rescue path planning problem. The first objective represents the maximization of probability of target detection to ensure the success of mission planning and the second objective represents minimization of the cumulative path length of all the agents to ensure resource utilization and ensure adequate area coverage. A two-phase fuzzy programming technique is used to find the Pareto optimal solution. Numerical experiments are conducted with CPLEX to evaluate the effectiveness of the solution procedure with varying number of agents, and the impact of the size of a grid-based rectangular map with a sparsely distributed non-cooperative finite number of stationary targets.