为探究纺纱过程中质量波动的规律,预测产生系统误差的不确定性因素的理论依据,分析纺纱质量特征值波动的成因、规律,以及影响因素的产生机理及其关系;利用纺纱过程的实时数据,从纺纱质量波动规律表达、人—机—环境脆性模型构建,以及TARCH(1,1)模型对影响因素异常行为辨识三个方面对纺纱质量特征值波动的内在机理进行了建模与设计,进而提出了基于实时数据的纺纱质量波动预测方法。通过实验仿真与对比分析,结果表明:预测方法实现了纺纱质量特征值波动过程的可视化,做到了影响因素异常行为的事前预警以及成纱质量的实时在线监测。
In order to find the regular pattern of spinning quality fluctuation rules, to predict un- certain factors of system error theory basis, analysis is done to the cause resulting in fluctuation of spinning quality eigenvalue,the regularity and the influence factors of the generation mecha- nism and the relationship~ the spinning process of real time data, expression from the fluctua- tion of spinning yarn quality and construction of man-machine-environment brittleness model and TARCH (1,1) model of influencing factors of abnormal behavior identification intrinsic mechanism of three aspects on spinning quality characteristics value fluctuation modeling and design,and puts forward the prediction method based on real time data of spinning quality fluc- tuation. The simulation and comparative analysis finds that the prediction method realizes the spinning quality eigenvalue variation in the process of visualization, so that the influence factors of abnormal behavior of the advance warning and real time monitoring of the yarn quality.