A Vector Splicing-Based Multi-Source Heterogeneous Data Fusion For Quality Fluctuation Characterization

Industrial process data often presents a characteristic with a wide range of sources and diverse structures, which leads to the problem of quality fluctuation characterizing and accurately modeling. This paper proposes a vector splicing-based multi-source heterogeneous data fusion for quality fluctuation characterization method. Firstly, the gray level co-occurrence matrix (GLCM) is used to calculate the feature parameters of image data information to construct the image feature vector. Then, vector concatenation technology is used to combine structured text data with image data to achieve heterogeneous data feature fusion and quality fluctuation characterization. Finally, a case study is conducted to evaluate the classification accuracy of the proposed heterogeneous data fusion method for different operating conditions, result reveals that proposed method shows a better performance compared to single structured-quality analysis method.