【摘要】 为了给木片质量在线分级,实现木片的优化管理,本文设计了木片管理系统CMS(Chip ManagementSystem).该系统采用前馈神经网络模型测得木片新鲜度和种类等级,利用近红外传感器测量木片含水率,根据多次粒度测量法获取木片尺寸信息,基于贝叶斯分类器原理利用虚拟传感器实现树皮含量的检测,最后根据以上测得的木片属性参数定义木片的质量等级.测量结果表明该系统可以准确地在线测量木片的属性参数,实现木片等级的在线预测.因此本文提出的CMS,可以根据木片质量的等级信息管理木片、监控木片的供给,从而达到制浆过程中优化精炼控制.更多还原

【Abstract】 A chip management system was proposed in this paper in order to realize the online measurement of wood chip quality and the optimization of wood chip management.In this system,a feedforward neural network was developed to estimate the grade of chip freshness and species;a near-infrared sensor was used to evaluate the average moisture content;a granulometry theory was used to calculate the distribution of chip size;a virtual sensor based on Bayesian color classifier theory was applied to the detection of bark content;and then the wood chip quality grade was defined according to aforementioned parameters.The results show that the CMS can measure the wood chip property parameters accurately and realize the online prediction of its quality grade.Thus CMS can be applied to chip yard management and chip feeding monitoring to optimize the pulping process.更多还原