Topologically adaptive ant colony optimization for scheduling mega-structure pile foundations with endogenous spatio-temporal constraints

Pile foundation scheduling of super-large engineering structure presents significant concerns and challenges due to the dynamically triggered and evolved vibration exclusion zones generated by impact boring. However, conventional re­source-constrained project scheduling theories often exhibit spatial blindness, resulting in cascading construction critical path delays. Therefore, this study proposes a Topologically Adaptive Ant Colony Optimization (TA-ACO) algorithm embedded within a state-dependent simulation environment. The algorithm employs a dual-colony cooperative mechanism to effectively de­couple the adversarial resource coupling between piling crews and grouting crews. Specifically, a novel look-ahead heuristic strategy is introduced, guiding crews to maximize spatial sparsity, which is analogous to solving the Maximum Independent Set problem and avoiding topological deadlocks. Computational experiments demonstrate that TA-ACO effectively mitigates spatial fragmentation and reduces the project makespan by 21.5% to 28.3% across different large-scale instances, compared to the industry-standard greedy strategy. The engineering application results of large airports show that the proposed algo­rithm compressed the construction period of a critical zone, which includes 97 pile foundations, from 47 days to 34 days, se­curing a valuable 13-day lead time for subsequent superstructure assembly.