Local damages such as microcracks and corrosion in crane steel structures often exhibit strong localization and weak mode shape perturbation characteristics. Especially when the damage scale is smaller than the modal wavelength, it cannot be simply treated as an overall damage issue for identification. In response to the diverse damage modes of crane structures and the need for dense sensor placement for local identification, a sensor optimization placement method based on multi-level damage feature fusion is proposed. Firstly, structural sub-regions are divided according to the main beam diaphragms, and sensors are arranged with boundary points as key measurement points. The displacement frequency response amplitude changes before and after damage are utilized to identify damage and eliminate insensitive measurement points to complete preliminary optimization. Secondly, a displacement frequency response amplitude change matrix is constructed, and the damage signal is enhanced through cross-frequency weighted superposition to form a damage identification vector, accurately locating the damage occurrence area. Furthermore, node-level correction is performed in candidate areas based on displacement flexibility difference values, and precise localization of damage points is achieved through priority sorting of flexibility differences. Simulation results show that under the condition where corrosion damage is set in units 20739, 20762, and 20785, the maximum point of the flexibility difference damage index is located in unit 20762, which coincides with the preset damage location, verifying the effectiveness of the hierarchical placement strategy from initial damage screening to precise localization.
Liu et al. (Sat,) studied this question.