Prediction and remediation of failed product identification based on manufacturing history data

Aiming at the problem that product identification was not read due to wear, pollution and other factors caused by production environment and process complexity of discrete manufacturing enterprise, a prediction and remediation method of failed product identification was presented based on manufacturing history data. The product manufacturing history data model based on Direct Part Marking (DPM) was given. The factors of failed product identification were analyzed and the history data was standardized by Z-score, and the extracted feature was optimized through Principle Component Analysis (PCA) method. The neural network model for prediction failed product identification was established, and product identification was remedied by using neural network prediction results with identification transfer and inheritance method. The experimental results showed that the proposed method could better predict or remedy the failed product identification.