Time Study of Automatic Edgebanding Machine Using BP Artificial Network

摘要

The edgebanding time is measured in one furniture factory made for wardrobe in Guangdong province,the BP network model has been built between operation time and length of gluing edge,cross edge,thickness of edgeband,then the data is calculated,trained and simulated by BP network tool box of Matlab R2009b.The result shows that BP network is a good method to build the correlation between factors and operation time,and the simulation accuracy is high by training of detailed data,MSE(Mean Squared Error) fulfills the target error of 0.001 when the net training get 1470 cycles,and the maximum error of output is less than 0.3 second,the relative error was less than 5％,the actual requirement of practice can be fulfilled.