Research on Building Quality Evaluation Method Based on Association Rule Algorithm and Neural Network

Based on the fusion of association rule algorithm and neural network, a modal parameter damage detection method for building engineering structures is proposed.The frequent item sets in the historical operation data are mined through the decision tree model, and the damage identifiers are constructed by combining the vibration mode analysis as the feature input of the neural network to achieve the accurate identification of the location and degree of structural damage.A modal parameter extraction and recognition model was designed, and a collaborative correction mechanism of local substructure association rules and neural networks was proposed, significantly improving the efficiency of damage detection.Experiments show that when the amount of alarm data is small, the speedup ratio of this method differs less from that of traditional methods.With the increase of data volume, the speedup ratio under the collaborative correction mechanism has increased to 3.89/5, and the recognition error is less than 3.36%, verifying its efficiency and robustness in large-scale scenarios.However, the performance limitation problem in small data scenarios still needs to be optimized.In the future, the applicability can be expanded through transfer learning or data augmentation techniques.This research provides a technical path with high precision and high response speed for the health monitoring of building engineering structures.However, in practical applications, it is necessary to balance the data scale and computational efficiency.