Optimization of plastic injection moulding process using data mining: A case study
Suhaimi, M. A.
MetadataShow full item record
Plastic injection moulding (PIM) process is used for converting plastics from its raw material into a kind of a semi-finish or finish product. PIM is a complex process due to the non-linear behaviour of controllable parameters available in producing high quality product. PIM usually used in a mass production line to support high demand and wide range of products, from as simple as electronic devices to as complicated as aerospace devices. Therefore, it is very important to control products from defects and to gain knowledge about parameters, which will influence to the whole PIM process. For this purpose, this paper presents an optimization of PIM process parameters in a fishing reel production via Data Mining methods. Our previous research done to optimize the parameters using Design of Experiments (DOE) methods proves that the PIM process can be optimized by running 16 experiments by two levels of fractional factorial design. In this paper, Data Mining will be utilized to provide optimal parameters of the PIM machine with desired accuracy. First, the important PIM machine parameter data in fishing reel production were collected. Then, the collected data are analysed using the REPTree Decision Tree method for classification. It is found that, this approach brings out the important decision of PIM machine parameters that are useful in obtaining the desired output results. The results from Data Mining method not only provides the same results as statistical method, but also introduces more efficient quality improvement activities with minimal cost and time consumption.
Paper presented at the Proceedings of the 24th International Conference on Flexible Automation & Intelligent Manufacturing, held May 20-23, 2014 in San Antonio, Texas, and organized by the Center for Advanced Manufacturing and Lean Systems, University of Texas at San AntonioIncludes bibliographical references