In the context of “Construction 4.0,” management and control failures occur due to the complex interdependencies of multidimensional objectives in large-scale projects. This study aims to resolve these bottlenecks by developing an improved decomposition multi-objective evolutionary algorithm-based construction configuration management and control model.
A two-layer coding system of “process topology-resource pattern” is utilized. Furthermore, adaptive neighborhood adjustment and differential evolution operators are employed to achieve efficient collaborative optimization of discrete construction sequences and heterogeneous resources.
Simulation results indicate that the model has good convergence, with a hypervolume index of 0.92 and a 12.4% reduction in average project duration. In massive examples, the peak resource demand is reduced to 72.7 units, and the cumulative cost input ratio on day 40 is controlled to 19.0%, mitigating resource fluctuations and realizing smooth capital flow.
The research findings provide Pareto optimal decision support for refined engineering management, balancing efficiency and safety by innovatively integrating a state-aware dynamic neighborhood mechanism with discrete resource scheduling logic.
