Evaluating the impact of soft management policies on construction and demolition waste recycling efficiency: A hybrid simulation-machine learning approach
Zhikun Ding, Xinping Wen, Yue Teng, Huanyu Wu
Abstract
Construction and demolition waste (CDW) recycling plays a critical role in sustainable development, yet the sector faces a low recycling rate and ineffective management practices. While current CDW recycling management predominantly employs government-led rigid measures, the potential of soft management approaches remains critically underexplored. To address this, this study evaluates the impact of soft management policies and identifies optimal strategies through an empirical study. An agent-based model embedded with back-propagation neural networks was innovatively developed and validated using data from 1005 residential projects in Shenzhen, China. Results reveal that soft management policies significantly enhance recycling rates and generate greater environmental, economic, and social benefits compared to rigid policies. Among soft policies, guidance policy performs best, increasing recycling rates by 14.13 %, followed by voluntary (7.24 %) and incentive (4.32 %) policies, while mandatory policy shows minimal improvement (0.24 %). However, a hybrid policy combining soft and rigid measures delivers the highest benefits, nearly three times those of the baseline policy. This study provides empirical support for integrating system simulation and machine learning to address CDW recycling management. More importantly, these findings advance scholarly understanding of the effects of soft management on CDW recycling and offer valuable insights for governments to refine CDW recycling policies, promoting a more sustainable future in the built environment.