Research On The Permeability Test Of Tipping Paper Based On The Expert System Of Confidence Rule Base

The use of perforated tipping paper is one of the most simple and effective technology for reducing tar in tobacco enterprises.Because of the requirements of different tobacco products for tipping paper permeability are not the same,the research on detection technology of tipping paper permeability has become the difficulties that tobacco enterprises to overcome.So far the detection for tipping paper permeability is dependent on the hardware equipment,but the study of air permeability detection by the experts and scholars shows that there are many fundamental characteristics of tipping paper related to paper permeability,such as drilling,hole size,size of paper materials,such as the number of holes on tipping paper,the area of the holes,paper material and so on.It supply the the basis for build the a soft measurement model for detect the permeability for perforated tipping paper.This paper extracted two main factors influencing tipping paper permeability,pore area and paper image gray value,which is based on the image processing technology for the detection of tipping paper permeability.This paper presents a belief rule base based on(Belief Rule Base,BRB)soft measurement method for tipping paper permeability detection.The main work of this paper is as follows:(1)Using image processing technology to obtain the image informati-on as feature variables,which is related to permeability of tipping paper and relatively easy to be extracted,and then establish the detection model for tipping paper permeability based on BRB.(2)To solve the problem that the model is not accurate due to the inaccuracy of the parameters in the rule base,the differential evolution algorithm is introduced to optimize the parameters of the model.In order to reduce the dependence of the rule base on the experts,the rule base attributes and the reference values of the rules are added into the parameters to be optimized.(3)In order to improve the performance of the algorithm,a local search called NM simplex method is combined with differential evolution algorithm,and the Meta-Lamarckian learning strategy is introduced to adjust the mutation strategy with the evolution algebra.Experimental results show that the proposed algorithm achieves good accuracy.(4)A differential evolution algorithm based on co evolution is proposed to improve the efficiency of the BRB algorithm,and an improved mutation strategy is added to improve the efficiency of the algorithm.Through simulation and comparison algorithm,the detection accuracy of BRB tipping paper permeability detection model is verified.Experimental results show that the precision of tipping paper permeability detection model based on BRB is good,and it has reached the industry requirements.