Abstract
In the United States, there are over 600,000 bridges with a complex built-in system. Data from the Federal Highway Administration show that nearly 7% of bridges were reported to be in poor condition in 2023. The factors causing bridge degradation are a combination of external environmental effects and internal material deterioration. Adequate and on-time resources are fundamental for optimal bridge management. In this context, an effective, highly reliable prediction approach is required to forecast performance changes precisely. This chapter proposed a new prediction method that hybrid stochastic paint optimizer and grey wolf optimizer algorithm with extreme learning machine algorithm (SPOGWO-ELM) to forecast the deterioration of a bridge. The factors affecting bridge degradation were screened using correlation analysis on a dataset comprising 539 bridge inspection records. Subsequently, the SPOGWO-ELM algorithm was used to establish a nonlinear mapping relationship between the influential factors and the bridge condition deterioration index (BCDI). The results verify the method's accuracy in predicting bridge conditions. In addition, statistical indices such as coefficient of determination (R2), mean squared error, root mean squared error, and mean absolute error were used to assess the performance of the proposed method. The results show that the proposed SPOGWO-ELM model exhibits significantly high R2 values, indicating improved prediction capability. As a result, the ELM-based model with a graphical user interface is demonstrated to be an effective and useful tool for estimating BCDI and to support engineers in making optimal decisions. The achievements of this study make a novel contribution to improving the bridge infrastructure’s state of knowledge and help build a safer and more sustainable transportation network with the assistance of artificial intelligence.