Assembly Quality Detection Based on Class-Imbalanced Semi-Supervised Learning

Due to the imperfect assembly process, the unqualified assembly of a missing gasket or lead seal will affect the product’s performance and possibly cause safety accidents. Machine vision method based on deep learning has been widely used in quality inspection. Semi-supervised learning (SSL) has been applied in training deep learning models to reduce the burden of data annotation. The dataset obtained from the production line tends to be class-imbalanced because the assemblies are qualified in most cases. However, most SSL methods suffer from lower performance in class-imbalanced datasets. Therefore, we propose a new semi-supervised algorithm that achieves high classification accuracy on the class-imbalanced assembly dataset with limited labeled data. Based on the mean teacher algorithm, the proposed algorithm uses certainty to select reliable teacher predictions for student learning dynamically, and loss functions are modified to improve the model’s robustness against class imbalance. Results show that when only 10% of the total data are labeled, and the imbalance rate is 5.3, the proposed method can improve the accuracy from 85.34% to 93.67% compared to supervised learning. When the amount of annotated data accounts for 20%, the accuracy can reach 98.83%.

Quality Inspection of Digital Archives of Application and Installation in Power Business Expanding based on Artificial Intelligence Technology

The existing quality detection methods of business expansion digital archives have the problem of fuzzy evaluation standard, which leads to low classification accuracy. This paper designs a quality detection method of business expansion Digital Archives based on artificial intelligence technology. The business characteristics of business development are extracted, the minimum business data unit is described, the digital archive catalogue database is established, the digital archive evaluation standard is defined, the text similarity is calculated, the user model is established, and the quality inspection mode is established by using artificial intelligence technology. Experimental results: the average classification accuracy of the designed method based on artificial intelligence technology and the other two quality detection methods is 55.763, 43.560 and 42.605, which proves that the quality detection method based on artificial intelligence technology has higher use value.

Research Progress on Quality Control Methods for Xiaochaihu Preparations

Xiaochaihu (XCH) is a classic Chinese medicine formula. XCH tablet, XCH granule, XCH capsule, and XCH effervescent tablet are included in the Chinese Pharmacopoeia. In this review, the formula and quality standards of XCH preparations at home and abroad were compared. The differences in manufacturing process of XCH preparations are discussed. The progress of research on the qualitative identification, quantitative detection and fingerprint chromatogram/specific chromatogram of XCH preparations was reviewed. The characteristic components of Pinelliae Rhizoma Praeparatum Cum Zingibere Et Alumine and Jujubae Fructus was rarely analyzed for XCH preparations. It is suggested that the specificity of drug quality detection methods should be improved. Considering drug safety and drug efficacy, it is suggested to set the upper and lower limits of the content of saikosaponins. The standards for heavy metals and other limited items for XCH preparations are also suggested to be set.