Grey Neural Network And Its Application In MCMQC

The biggest problem in MCMQC (mass customization manufacturing quality control) is the lack of quality data. It results in the difficulty to predict quality defects. To solve the problem, we can make use of a new theory (the grey theory). In grey theory, grey means the lack of information and the bodies of information are considered as hierarchical, poly-dimensional and time-variant. Based on grey system, we studied the model building methods of it. To simplified the computation, the author also gave out three simplified model building methods (crosswise modeling method, lengthwise modeling method and hybrid modeling method), which can be intimated into model building of neural network. Finally, the author gave out an example for a typical grey BP neural network in which the crosswise modeling method was used and applied to mass customization manufacturing quality control.