This article focuses on the issues of significant quality fluctuations and difficulty in quality traceability in sand casting products due to complex production processes, long manufacturing processes, and large production volumes. It proposes a data mining method based on the Huazhu ERP system. Through in-depth excavation and analysis of collected data, the mining results are presented in various forms. Additionally, a neural network is utilized to predict the quality of products with different process parameters, aiming to enhance the quality control ability and efficiency of enterprises.
1. Introduction
In the 21st century, with the continuous development of computer technology, digital, networked, and intelligent technologies have become the main driving forces in the manufacturing industry, providing new technical paths for quality control in manufacturing. Many enterprises have introduced systems such as ERP and MES for information management. Sand casting, an important branch of casting, plays a crucial role in the production of engine cylinder heads in the automotive industry. However, due to its characteristics, there are problems with product quality fluctuations and difficult quality traceability. Although sand casting enterprises are transforming towards automation and informatization, the existing data has not been effectively utilized, resulting in a situation of “data explosion but knowledge deficiency”. Data mining, as a process of revealing new relationships, trends, and patterns from large amounts of existing data, is of great significance for sand casting enterprises. ERP and MES systems can provide data support for data mining. Therefore, introducing the ERP system and conducting data mining based on it is an important development direction for quality control and product quality improvement in sand casting enterprises.
2. Huazhu ERP Main Framework
The Huazhu ERP system is developed by the Huazhu Software Center of Huazhong University of Science and Technology. It aims to standardize enterprise management, improve efficiency, reduce costs, accelerate the informatization process, and enhance market response ability. The system’s business process is centered around customers, driven by tasks, and production is pulled by orders, enabling comprehensive management of various aspects such as procurement, production, sales, and inventory. Y Company, a casting subsidiary of a state-owned enterprise, mainly produces diesel engine components and has introduced the Huazhu ERP system for piece-by-piece management of castings. The application process of the Huazhu ERP system in Y Company includes order entry, process route setting, BOM formulation, casting process allocation, production preparation, planned production, production processing, and sales and delivery.
3. Data Mining
3.1 Data Collection
Data collection in the Huazhu ERP system is achieved through both automatic device collection and manual entry. Automatic collection includes data from various production processes such as sand mixing, core making, and casting. Manual entry covers order information, process cards, quality registration, and shipping information.
3.2 Data Mining Model
The proposed data mining model is mainly based on association analysis and neural network methods. Using the casting number as the input, it can associate with various aspects such as orders, quality records, and production records. The model can analyze customer needs, process parameters, and defect causes. The key data tables in the database include the order details table, planned production table, production quality inspection table, production parameter record table, and shipping order, which are related through primary and foreign keys.
3.3 Data Display
The final analysis results of data mining are presented in the form of data tables and statistical charts, including quality statistics data tables, scatter plots of daily casting production, bar charts of defect distribution, and pie charts of monthly casting production. At the same time, by training a neural network model with process parameters as input and quality results as output, the quality of products with different process parameters can be predicted.
4. Conclusion
This article studies the problem of the lack of data mining and analysis in casting enterprises after the introduction of information management systems. A data mining model based on the Huazhu ERP system is proposed, which can associate various business processes and provide in-depth information on customers, defect distribution, and production capacity. The model can also trace the quality problems of casting products and predict the quality of different process parameters, thereby enhancing the enterprise’s quality control level and promoting its better development.
In the future, with the continuous development of data mining technology and the increasing application of ERP systems in the casting industry, more in-depth research and application can be carried out to further improve the quality and efficiency of sand casting products.
