Path Planning for Surface Agent Formations Based on the Dynamic Programming Value Iteration Algorithm

Addressing the critical requirement for agent formations to successfully complete engineering operations in complex maritime environments, this paper proposes a path planning method based on the Value Iteration algorithm from Dynamic Programming. Firstly, a gridded model of the water surface is established, incorporating depth constraints and interception probability data. Subsequently, the probability product optimisation problem is transformed into a weight sum minimisation problem via a negative logarithmic transformation. Finally, a Dynamic Programming Value Iteration algorithm is designed to facilitate path search under 4-connectivity movement constraints. Simulations conducted on a 25 × 41 complex aquatic grid validate the effectiveness of the strategy through comparison of results from three simulation-based estimation methods, providing a reference for path planning in formation movement.