Quantitative Evaluation Method for the Significance of Worsted Forespinning Parameters Based on BP Network
Gui Liu, Weidong Yu · 2007
The BP neural network characteristic has been summarily analyzed. Based on its error back propagation method, the peculiarity of modifying its weightiness and threshold value to make the calculated error come down along the negative gradient direction, the article proposed a new approach that used the weightiness distribution between the input and output layer to appraise the input parameters' significant degree. Take the worsted craft as the example, each input parameter's contribution rate has been calculated to the roving unevenness (R1) and roving weight (R2) respectively, and the remarkable and effective parameters are excavated out. Meanwhile contrasting to the multivariate regression significance analysis (MRSA), the BP neural network method is more exact than MRSA and also can be used in the forecast and control of the actual produce and manufacture.