| description abstract | Upgrading existing neighborhoods is crucial for promoting sustainable development of the cities. To propose effective maintenance measures to mitigate the deterioration rate of existing neighborhoods, it is necessary to accurately assess their accelerated deterioration risk. Considering the probability, uncertainty, and dynamics of the risk assessment process, this study proposes a risk assessment model that combines Bayesian networks with Dempster–Shafer (D-S) evidence theory. Firstly, risk factors were identified by analyzing the deterioration scenario of existing neighborhoods and mapped to create the model’s topology structure. Expert knowledge was then utilized to address the problem of insufficient data. The D-S evidence theory and variable fuzzy set were applied to obtain prior knowledge. The parameters of the model were determined by employing parameter learning. Finally, forward prior reasoning and backward diagnostic reasoning were utilized to simulate the evolution process of the whole risk system. Together with the strength of influence and the sensitivity analysis, it is revealed that the living environment and municipal infrastructure subsystems were critical subsystems. Poor residential suitability, poor comfort of built environment, and inadequate transportation networks were important risk characterization elements. Additionally, the critical risk factors were inadequate operation and management systems, insufficient information management, inadequate structural reliability, higher incidence of vandalism, poor comfort in indoor environment, and inadequate characteristic and cultural facilities. The results of high probability evolution paths suggest that establishing an efficient management system and constructing new and upgraded service facilities to create a livable living environment is a priority to slow down the deterioration speed and reduce the probability of being regenerated earlier. This study provides valuable insights into assessing the accelerated deterioration risk of existing neighborhoods. It successfully shifts risk management from lagging pro-reactive treatments to early warning and control. The model facilitates the proposal of effective maintenance measures and risk avoidance recommendations to reduce the occurrence probability of accelerated deterioration accidents of the existing neighborhoods. | |
