64 J. Zeng et al. within 20% of the threshold, as indicated by the triangle markers. This suggests a possible false positive or minor sensitivity in the model for Bridge-9-SS. Nevertheless, all other bridges remain well below their respective thresholds, reinforcing the localized nature of the damage in Bridge-4-SS. Similarly, for Bridge-17-TS (Figure 4), representing a two-span bridge, the RDI results indicate structural damage. The RDI values for Bridge-17-TS exceed the threshold, with values around 14 to 16. However, unlike the single-span bridge, the magnitude of the difference is slightly lower, reflecting the distribution of damage across the multiple spans. Despite this, the method effectively identifies the presence of damage in a more complex structural configuration, while other bridges maintain RDI values below their thresholds. These observations demonstrate the effectiveness of the signal-level approach in detecting structural anomalies across both single-span and multi-span bridges, with the RDI values serving as a robust indicator of damage. The method’s ability to consistently identify damage in various bridge types underlines its applicability in diverse structural health monitoring scenarios. Fig. 3 Distribution of RDI and TRDI values for the damaged state: Bridge-4-SS. Fig. 4 Distribution of RDI and TRDI values for the damaged state: Bridge-17-TS.
RkJQdWJsaXNoZXIy MTMzNzEzMQ==