Identification Method of Building Engineering Structural System Based on Fuzzy Neural Network

There are complex problems such as nonlinearity, time-varying and environmental interference in the identification of building structures. Based on this, this paper introduces a system identification method based on fuzzy neural networks. This method uses the powerful nonlinear mapping and learning capabilities of fuzzy neural networks to accurately and quickly identify the dynamic characteristics of building structures. First, the paper uses three-dimensional modeling technology to construct a digital twin model of the building structure as the basis for system identification. Secondly, IoT devices are deployed at key locations of the building structure to collect structural response data. Then, the input data is converted into a fuzzy set through fuzzification processing, and the fuzzy rules are trained and optimized using the self-learning ability of the neural network. Then the model is used to identify the dynamic characteristics of the building structure. In the test of 20 sets of different building structure data, the average accuracy of this method reaches 99.17%, with the highest accuracy reaching a perfect 100% and the lowest reaching 98.2%. The identification time and identification accuracy are also better than the identification method based on radial basis function network. This paper enhances the understanding of structural dynamic behavior and provides new perspectives and tools for structural safety assessment and disaster prevention in the field of construction engineering.