In common manufacturing lines, engineers usually adjust machine parameters according to product defects. It takes many tests to make the product stable, which not only takes time but also wastes additional material costs. Therefore, Intelligent manufacturing has been widely used in manufacturing in recent years, and it is an effective solution for shortening time and manpower. In this study, injection molding was used as an example. The product used was a flat plate model with uneven thickness, and the pressure was monitored on the mold. Observe the changes of the melt during the forming process by online monitoring, and analyze the correlation between different parameters and quality. The results show that melt temperature, mold temperature, injection speed, holding pressure, holding time and cooling time are analyzed to be relatively important for product quality. Through the changes of these indicators and feature analysis, a complete data set can be summarized. In further research, these data sets will develop into important materials for deep learning (machine learning) in the future.