Risk Assessment of Bridge Construction Safety Considering Multi-Information Fusion

Abstract

Due to the rapid development in the number of bridges worldwide, risk assessment has gradually been an important step to evaluate the safety of bridge construction. However, the assessment results are usually affected by the subjectivity and individual preference of the bridge experts. In this paper, an innovative bridge construction risk assessment algorithm considering multi-information fusion is proposed based on cluster analysis, set-valued statistic, and entropy method. First, the weight of each individual expert is obtained by adopting cluster analysis to reduce the influence of individual bias from experts. Second, the index weight considering importance-based and condition-based information fusion is achieved based on the minimum cross-entropy principle. The importance-based weight of each index is achieved by the fall-shadow method based on the set-valued statistic. The condition-based weight of each index is computed using the entropy method with respect to the evaluation results of the indexes. Finally, the fuzzy comprehensive evaluation process is utilized to conduct the assessment from the bottom layer to the target layer inside the hierarchical risk index system; the risk level of the target bridge construction safety and corresponding control principle are achieved based on the risk evaluation criteria. An example risk assessment of the construction stage for a long-span suspension bridge is applied to illustrate the validity of the proposed approach.