摘要:
本文针对一个大型炼油蒸馏塔的控制,设计了一种适用于具有多个确定性扰动对象的多
输入预报自校正控制器,其特点是综合了在线辨识、多输入前馈、最小方差和PI规律各自的优
点,并采用变遗忘因子的最小二乘估计算法在线修正参数,使控制器的收敛性、稳定性与自适应能力较一般自校正控制器大为提高.实践表明,该控制器很好地克服了生产中的干扰和波
动,平稳了操作,成功地实现了对蒸馏塔的产品质量控制.
Abstract:
In solving the control problem of a large crude distillation column, a multi-input prediction
self-tuning controller which can be used to a process with multi-disturbance, has been
designed. The advantages of on-line identification, multi-input feedforward, minimum variance
and PI control law are considered in the design of the controller. A least-square estimator
with variable forgetting factor is used for the revision of its parameters. Hence, the
controller is better than an ordinary self-tuning one in convergency, stability and adaptivity.
The quality control of the distillation column is successfully realized. Results show
that the controller has effectively overcome the disturbances in production, smoothed the operation
and brought in a notable economic profit.
下载:
