摘要
慢变过程量控制的开工启动阶段或设定值变化的中间过渡阶段，设置一小幅度的阶跃输入是允许的，由此得到的阶跃响应可作为对象预测模型。在线建模情况下，模型失配性质应分类为随机性失配、系统性失配及截断过早失配。文章分别讨论了各类失配的特点，提出修正处理的方法，并为加快优化过程作了考虑。因为免除了离线整定过程，有条件将动态矩阵算法应用于廉价的实时控制器，可望改善丝绸工业中常见的许多加工设备的控制质量，具有实用意义。
In start working stage and transiting stages of initialization a small step input was allowed for slower industrial processes. The step response obtained from this can look on as the predictive model of project. In the case of creating model on line, the nature of model distortion should assort random, systematic and interceptive prematurily. This pepar discussed all kinds of distortion characteristics respectively, presented modificative method and advisement for accelerating optimize couse. The application study of the dynamic matrix control algorithm for using in instrumental controller avoid the preparatory steps line off, which possess practicability meaning. It is hopeful, that control quality of devices is improved in silk industrial.
出处
《苏州丝绸工学院学报》
1997年第4期52-59,共8页
Journal of Suzhou Institute of Silk Textile Technology
关键词
预测控制
动态矩阵算法
控制系统
仿真
predictive control, dynamic matrix control, simulation of control system
