A Path Similarity Based Multimodal Multi-Objective Evolutionary Algorithm for Path Planning Problem

The multimodal multi-objective path planning problem aims to identify a set of equivalent Pareto-optimal routes between a given start and end points that simultaneously minimize traffic congestion, the number of intersections, and path length. This problem is particularly important for handling emergency traffic incidents and disaster relief scenarios. However, existing algorithms generally struggle to maintain the complete set of equivalent optimal paths. To overcome this challenge, a path similarity driven multimodal multi-objective evolutionary algorithm (MMOEA) is proposed for path planning. The algorithm measures the dissimilarity among candidate paths in the decision space to effectively identify and preserve promising multimodal solutions. Extensive comparisons with four representative approaches on the IEEE CEC 2021 multimodal path planning benchmark verify that the proposed method exhibits superior performance in both detecting and maintaining multiple equivalent solutions.