Research on Optimal Design of Civil Sensors Based on Agglomerative Hierarchical Clustering Algorithm

In practical engineering, the sensor measurement points will have obvious clustering characteristics. Therefore, the cohesive hierarchical clustering algorithm is introduced in this paper. Firstly, the degrees of freedom are classified according to the similarity of vibration features, and all degrees of freedom are divided into multiple clusters, and disjoint subsets are formed between the clusters to avoid the concentration of measurement points. Secondly, the hierarchical clustering algorithm is improved, and a single objective function sensor location search method is established with the minimum discomfort sensor location selection criterion and MAC pattern guarantee criterion as objective functions. Different methods are used to search the best position of acceleration sensor. Finally, the empirical performance and scalability of the proposed algorithm are verified by an example analysis. In this paper, a new branch direction of hierarchical clustering is studied, which provides a meaningful exploration for the empirical performance and scalability of balanced hierarchical clustering, and provides new possibilities for structured data analysis and mining tools.