This study aims to construct a graded evaluation method for the construction quality of Highway Prefabricated Beam Bridges, in order to support refined quality control beyond the traditional “qualified/unqualified” judgment. Existing evaluation methods have difficulty effectively distinguishing quality differences within the qualified range, may cause redundant calculations for indicator information with non-linear correlations, are susceptible to the influence of outlier expert judgments, and may mask local critical defects when adopting linear weighted aggregation. This paper uses historical quality inspection and expert evaluation records from 50 prefabricated beam bridges to validate the proposed method. Experimental results show that the proposed method achieves an RMSE of 1.24, outperforming the traditional AHP method (1.79) and the equal-weight method (2.52). Five-fold grouped cross-validation at the bridge level gave an RMSE of 1.2180 ± 0.0802 and an R2 of 0.9536. It can provide accurate intelligent decision support for quality and safety control during the construction of Highway Prefabricated Beam Bridges, and promotes the digital upgrade of quality management in the construction industry. The proposed framework can be integrated into a digital construction quality management platform to support bridge quality ranking and early warning of safety-critical defects. The originality of this paper lies in the integration, within a unified evaluation framework, of continuous score conversion based on probability distributions, manifold-learning-based redundancy correction for indicators, geometrically median-based robust fusion of expert weights, and a threshold-triggered risk-sensitive weight adjustment mechanism.
To reduce dependence on expert judgement and improve the consistency and resolution of construction quality evaluation for highway prefabricated girder bridges (HPGBs), this study proposes an intelligent evaluation method integrating key parameter control and weighted neural network optimization. First, a three-level evaluation system covering the finished bridge, first-level indicators, and secondary indicators is established for the full construction process from prefabrication to installation. Threshold-type indicators are scored using a two-point rule, whereas finite interval indicators are converted into continuous scores using a two-sided piecewise-normalized truncated normal model. Grey relational analysis is then used to identify key quality parameters and quantify inter-indicator relational strength, and an adaptive weight decay mechanism attenuates redundant non-key indicators while retaining the key parameters. Data-driven indicator importance is subsequently learned using a weighted neural network, and hierarchical weighted aggregation is used to obtain the final bridge score. The model development database contains 10 HPGBs and nine expert evaluations per bridge, yielding 90 labelled expert-bridge records; Bridges 1–7 were used for training and Bridges 8–10 for bridge-wise testing. On the test set, the proposed method achieved R2 = 0.9147, RMSE = 1.294, MAE = 1.0987, and MRE = 1.36%. Compared with the traditional AHP method, MAE was reduced by approximately 22.01%, and R2 increased by approximately 5.86%. An engineering application yielded a comprehensive score of 84.4, corresponding to Grade B. These results demonstrate the feasibility of the proposed framework for refined quality evaluation within the HPGB class considered in this study.