Tehnički vjesnik, Vol. 33 No. 2, 2026.
Izvorni znanstveni članak
https://doi.org/10.17559/TV-20250719002846
Prediction Algorithm for Aerospace Product Quality Classification
Deyu Shen
; College of Computer Science and Technology, Nanjing University of Aeronautics and Astronautics, Nanjing, P.R. China Beijing System Design Institute of the Electro-mechanic Engineering, Beijing, P.R. China; No. 51 Yongding Road, Haidian District, Beijing, P.R. China
Ningzhong Liu
; College of Computer Science and Technology, Nanjing University of Aeronautics and Astronautics, Nanjing, P.R. China No. 29, Jiangjun Avenue, Jiangning District, Nanjing City, Jiangsu Province, P.R. China
*
Wei Liu
; Beijing System Design Institute of the Electro-mechanic Engineering, Beijing, P.R. China No. 51 Yongding Road, Haidian District, Beijing, P.R. China
Zhe Wang
; Beijing System Design Institute of the Electro-mechanic Engineering, Beijing, P.R. China No. 51 Yongding Road, Haidian District, Beijing, P.R. China
* Dopisni autor.
Sažetak
Productive activities in aerospace industry are characterized by their use of multiple species, small batches, discrete types etc. Therefore, the quality classification of aerospace products is a typical small-sample classification problem. The existing general algorithms are inadequate for an effective classification as well as prediction of aerospace production quality. To fill the gap, the research presents a specialized algorithm to the classification prediction of aerospace production quality by integrating isometric feature mapping (ISOMAP) with support vector machine (SVM). The accuracy of each main kernel function is compared, followed by a determination of the algorithm model using a radial basis function (RBF). Experiments show that the proposed algorithm significantly improves classification and prediction accuracy for aerospace product quality, thereby enabling the timely identification of potential quality issues and the subsequent reduction of quality issues in aerospace products.
Ključne riječi
aerospace products; classification prediction; production quality; SVM
Hrčak ID:
345001
URI
Datum izdavanja:
28.2.2026.
Posjeta: 149 *