OPTIMISATION OF PRECISION INJECTION MOULDING PROCESSING PARAMETERS AND IMPLEMENTATION OF QUALITY PREDICTION SYSTEM FOR LCD LIGHT GUIDE PLATE

摘要

The quality of the light guide plate determines the luminance uniformity of liquid crystal display panels. Onlyrnthrough proper control of the injection moulding processing parameters for the light guide plate can make thernprecision of the light guide plate’s dimensions to be improved, thus enhancing the light guide plate’s opticalrnproperties. The objective of this research was fi rst to discover the optimal combination for the light guide plate’srninjection moulding processing parameters. The Taguchi method was used to conduct the parameter design andrnselect the control factors within the processing parameters that affect light guide plate quality.rnA confi rmation experiment with a 95% confi dence interval was then adopted to verify the reproducibility. Atrnthe same time, a back propagation neural network was applied to set up a quality prediction system for the lightrnguide plate moulding. Its input layer had eight neurons, representing eight control factors. The hidden layerrnhad fi ve neurons. The output layer had one neuron, which was the signal-to-noise ratio of the light guide plate’srnluminance uniformity. By comparing the neural network’s predicted values with the confi rmation experiment’srnactual values, the system’s ability to make accurate predictions could be verifi ed.