HomeResearch and Technology DevelopmentHigh-precision quality management system based on XRD/Rietveld analysisFureture Outlook

High-precision quality management system
based on XRD/Rietveld analysis


>Introduction >Quality Management Video >Overview and Features >Measurement and Analysis >Future Outlook

Construction of a new quality management system

The introduction of the XRD/Rietveld analysis system in our production sites has made it possible to perform precise mineral composition measurement 24 hours a day. This has created the new challenge of how best to use this measurement data to optimize production.

To promote the effective use of this data, we developed our new quality management system, the Taiheiyo cement Quality Predictive System (TQPS). The TQPS uses various test data such as XRD/Rietveld analysis data. It utilizes the "neural network" (see bottom of page) information processing method to predict quality properties with a high degree of precision.
This system makes it possible to use only the data obtained during the cement production process to predict test values, such as cement compressive strength values, which could normally only be measured after a certain amount of material aging. By predicting product qualities during production, Taiheiyo Cement can rapidly predict the quality changes that would result from the use of new fuel and raw materials, and swiftly implement any necessary response measures.
Furthermore, optimal operation conditions determined by the TQPS can be established to help reduce production costs. This system has undergone testing in our central R&D lab, and since 2012 has been rolled out to Taiheiyo Cement plants using the XRD/Rietveld analysis approach. Systems are being established and operated on a trial basis in these sites.

This is an example of our innovative and ambicious development, helping further improve quality while reducing costs, creating customer satisfaction.

What are neural networks?

Neural networks are mathematical models based on the information relay mechanisms of the central nervous system. They are used in various fields, such as 2D barcode reading and other pattern recognition applications.

We focused on the fact that it continually learns from presented data to produce optimized solutions, applying this ability to the analysis of the relationships between cement characteristics, production operation conditions, and quality properties. The analysis model is updated whenever new data is added, maintaining a consistently high level of analysis precision.
Schematic view of analysis using a neural network


Copyright© TAIHEIYO CEMENT CORPORATION All Rights Reserved.
back to top