Immune Clonal Selection Network Clustering Algorithm and its Application to Fault Diagnosis

Aiming at the clustering problem of complex shape data, a novel immune network clustering algorithm based on the clonal selection is proposed. Firstly, network approach is employed to make the model of antibody and construct the similarity matrix of the antibodies, and then a modularity-based clustering criterion function is designed to realize the self-adaptive compression for the antibody network. Secondly, an immune clonal selection network clustering algorithm (ICSNCA) is developed. And simulation results of the university of california, irvine (UCI) standard data sets show that the proposed algorithm, compared with the aiNet algorithm and the general immune network clustering algorithm, is efficient for the data, and has advantages on the high compression ratio and the high accuracy ratio. Finally, the method is applied to the fault diagnosis of the Tennessee Eastman Process (TEP) data. Compared with other clustering algorithms, the proposed method is superior in the classification accuracy and the compression ratio, and it is very important for fault diagnosis.