A Foothold Selection Framework for Hexapod Robots Integrating Multi-Constraint Pruning and Adaptive Evaluation

Foothold selection for hexapod robots in complex terrain presents a significant challenge, demanding both high computational efficiency and robust adaptability. To address this issue, this paper introduces a foothold selection framework integrating multi-constraint pruning with adaptive evaluation. The framework employs a hierarchical decision-making mechanism. Initially, a Multi-dimensional Hard-Constraint Pruning Architecture (MHCPA) systematically incorporates kinematic, terrain, collision, and stability constraints. This stage efficiently eliminates physically infeasible solutions, thereby drastically reducing the decision space. Subsequently, an Adaptive Weighted Foothold Selection (AWFS) algorithm is applied to the pruned set of candidates. Leveraging a Foot Candidate Evaluation Function (FCEF), the AWFS algorithm dynamically adjusts evaluation weights based on the specific mission context, enabling adaptive decision-making. Simulation results demonstrate that the proposed framework improves decision-making efficiency by up to 90.8% while consistently securing high-quality footholds. Furthermore, in comparative analyses against state-of-the-art methods, our framework exhibits a favorable trade-off between solution quality and computational performance.