基于置信规则库专家系统的水松纸透气度检测研究

发布时间：2018-04-16 07:50

本文选题：打孔水松纸 + 透气度　；参考：《昆明理工大学》2017年硕士论文

【摘要】：打孔水松纸的使用是烟草企业降低消费者对焦油的摄入量使用的最为简单有效的降焦技术之一。由于不同的香烟产品对打孔水松纸的透气度的要求不一样,因此对水松纸透气度的检测技术的研究成为了烟草企业竞相攻克的难点。国内烟草企业目前对水松纸透气度的检测仍然基本依赖硬件仪器,但专家学者透气度检测的研究显示,水松纸的透气度与纸张的许多基本特性有关,如打孔数量的大小、孔面积大小、纸张材料等,这为打孔水松纸透气度的检测建立软测量模型提供了基础。本文根据已有的基于图像处理技术进行透气度检测的仪器,结合打孔水松纸的图像信息与透气度的关系,提取出影响打孔水松纸透气度的两个主要因素,孔面积和纸张图像灰度值。本文提出了一种基于置信规则库(Belief Rule Base,BRB)的水松纸透气度检测的软测量方法。本文的主要工作如此下:(1)利用图像处理技术获取水松纸的图像信息,提取影响水松纸透气度的特征变量,建立基于BRB的透气度检测模型。(2)针对置信规则库依赖专家知识,参数设置不准确会导致的模型不准确的问题,将差分进化算法引入对模型部分参数进行优化。为了减少规则库对专家的依赖,将规则库前提属性和输出的参考值作为参数加入了待优化参数中。(3)为提高算法性能,对差分进化算法加入单纯形局部搜索,且引入Meta-Lamarckian学习策略随进化代数的叠加调整变异策略,以提高算法的性能。实验证明了所提算法的取得了较好的精度。(4)提出基于协进化的差分进化算法训练BRB的结构参数,并加入了一种改进的变异策略来提高算法效率。通过实验仿真和算法比对,验证了BRB水松纸透气度检测模型的检测精度。实验表明,本论文提出的基于BRB的水松纸透气度检测模型精度较高,已达到工业要求。
[Abstract]:The use of perforated pine paper is one of the most simple and effective techniques to reduce the consumption of tar in tobacco enterprises.Because different cigarette products have different requirements for the air permeability of perforated pine paper, so the research on the testing technology of the air permeability of pine paper has become a difficult point for tobacco enterprises to conquer.At present, the domestic tobacco enterprises still rely on hardware instruments for the air permeability detection of water pine paper. However, the research of the air permeability test by experts and scholars shows that the air permeability of the paper is related to many basic characteristics of the paper, such as the size of the perforation number.The size of hole area, paper material and so on, which provides the basis for the establishment of soft sensing model for the air permeability of perforated pine paper.In this paper, according to the existing instruments for air permeability detection based on image processing technology and the relationship between the image information of perforated pine paper and the air permeability, two main factors affecting the air permeability of perforated water pine paper are extracted.Hole area and paper image gray value.In this paper, a soft sensing method for air permeability detection of water pine paper based on belief Rule base (BRBB) is presented.The main work of this paper is as follows: (1) using image processing technology to obtain the image information of water pine paper, to extract the characteristic variables that affect the air permeability of water pine paper, and to establish a model of air permeability detection based on BRB.Because of the inaccurate setting of the model, the differential evolution algorithm is introduced to optimize some parameters of the model.In order to reduce the dependence of the rule base on experts, the premise attribute and the output reference value of the rule base are added as parameters in the parameters to be optimized. In order to improve the performance of the algorithm, simplex local search is added to the differential evolution algorithm.The Meta-Lamarckian learning strategy is introduced to adjust the mutation strategy with the superposition of evolutionary algebra to improve the performance of the algorithm.Experimental results show that the proposed algorithm achieves good accuracy.) A coevolutionary differential evolutionary algorithm is proposed to train the structural parameters of BRB and an improved mutation strategy is added to improve the efficiency of the algorithm.The accuracy of the model for air permeability detection of BRB water pine paper is verified by experimental simulation and algorithm comparison.The experimental results show that the model proposed in this paper based on BRB has high accuracy and meets the requirements of industry.
【学位授予单位】：昆明理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP391.41

【参考文献】