PBWOA: a performance-based multi-strategy whale optimization algorithm for multi-robot cooperative path planning

Abstract Efficient and collision-free trajectory generation for multiple robots in cluttered environments remains a central challenge, as path optimality, inter-agent coordination, and obstacle avoidance often impose competing demands. This paper proposes a performance-based multi-strategy whale optimization algorithm (PBWOA) to address multi-robot cooperative path planning, where a time-stamp segmentation (TSS) framework is adopted as the underlying problem formulation. PBWOA embeds a performance-based adaptive mechanism among the canonical behaviors of the whale optimization algorithm (WOA)—encircling, spiral updating, and random exploration—where each behavior’s selection probability evolves adaptively through a shifted-softmax performance evaluation mechanism. This adaptive behavioral regulation enables a self-balancing interplay between exploration and exploitation throughout the optimization process. The accompanying TSS model establishes a unified temporal discretization that synchronizes multi-robot trajectories, streamlining collision avoidance and communication while enhancing cooperative efficiency. Extensive evaluations on the CEC2017 benchmark suite, together with multi-robot planning experiments formulated under the TSS model, demonstrate that PBWOA achieves more stable and reliable convergence than existing metaheuristic and WOA-derived methods.