Spatial multi-vector and multi-rigid-body obstacle avoidance planning for multi-robot coordinated suspension system

Abstract

Multi-robot coordinated suspension systems (MCSS) face challenges such as difficulties in decoupling due to strong coupling in multi-body dynamics, computational complexity explosion caused by high-dimensional state spaces, and a scarcity of obstacle avoidance planning methods for multi-rigid-body systems. To address these issues, the telescopic pyramidal configuration (TPC) and the multi-strategy geyser-inspired algorithm (MGEA) are proposed. These methods enable multi-vector and multi-rigid-body obstacle avoidance planning via a hierarchical-search and step-optimization (HSSO) framework. The MGEA enhances global search capabilities through chaotic mapping initialization, Lévy flights, differential evolution, and stability constraints, while hierarchical cooperative planning ensures the prevention of cable entanglement and decouples multi-robot motion. Simulation and experimental results show that MGEA outperforms benchmark algorithms, achieving a 16.35 % reduction in trajectory length, a 13.74 % increase in computational speed, and an 18.60 % improvement in minimum fitness, while maintaining zero collisions in cluttered 3D environments. This method provides an efficient solution for industrial lifting tasks and demonstrates scalability potential. These advancements establish a theoretical foundation for real-time planning in dynamic non-convex environments, addressing critical challenges in the application of multi-robot suspension systems.