大滞后系统数字内模控制器的设计及仿真研究

对于带有严重纯滞后的工业对象，目前的控制策略是史密斯预估控制，或者是内模控制，他们都要使用连续域的数学模型。而在目前的对象辨识中，辨识出来的对象往往是离散形式的数学模型，这就需要连续到离散模型的转换。在转换的过程中，难免会出现一些偏差，使得辨识出来本来就不是很精确的数学模型变得更加偏离实际系统，进而使得控制效果不尽理想。该文针对这种情况，提出一种基于离散对象模型的数字内模控制（DIMC）算法。该算法直接利用对象辨识给出的离散模型，导出离散的内模控制器，直接用于工业生产的计算机控制。仿真结果表明，该方法能克服被控对象参数变化和时滞变化对控制性能的影响，具有很强的鲁棒性能。

Design and Simulation of Digital IMC for Great Lag System

For an industrial plant with high great lag, the current control strategy is Smith predictive control, or internal model control, both of them are using continuous models. However, most of the results of current plant identifications are given by discrete models. This needs some kind of transform from continuous to discrete models. During this process of transform, some errors are inevitable, thus making the models far from the actual systems. Then the control effect is bad by doing so. For the fact given above, a digital internal model control (DIMC) algorithm based on discrete model is developed. The algorithm makes use of discrete model given by plant identification directly, deduces discrete internal model controller, and is used in computer control of industry manufacture directly. Simulation shows that this method can overcome the influence on control performance come from the parameter variation and the time delay variation of the controlled object, and has stronger robustness.