大滞后对象控制策略的研究.pdfVIP

大滞后对象控制策略的研究 大滞后对象控制策略的研究 , 摘 要 针对具有大滞后对象的控制问题，着重选取几种控制策略,考虑时滞大小,模型不确定性及外 部干扰等因素，利用 MATLAB 软件进行仿真，比较各种控制策略的控制性能，以期待对含有大 时滞的实际对象的控制起到方法的指导作用。仿真结果表明，对大滞后对象的控制，常规PID 控 制算法的控制效果差，只适合无时滞或时滞短的系统；大林（Dalin ）控制算法和 Smith 预估算法 对模型不确定对象控制的鲁棒性差，只能用于定常系统；模糊控制和神经网络控制不需要被控对 象的精确模型，具有很强的鲁棒性，被证明是控制大滞后对象的有效方法。 关键词：大滞后,PID 控制,Dalin 算法,Smith 预估,模糊控制,神经网络 1 大滞后对象控制策略的研究 The study of control strategies for plant with big time delay Chen Shaoqing (Department of automation of Information Engineering college ,Beijing Institute of Petrochemical Technology, Beijing, postcode 100617) Abstract The control methods for big time-delay plant are discussed. Several control strategies are mainly introduced.Considering the length of time delay ,the uncertainty of the model and the effect of external disturbance, different control strategies are compared by the means of MATLAB simulation, so that we can get the direction of the control method for big time delay plant. The result of simulation shows that the conventional pid algorithm has poor control performance, which is suitable for non-time-delay plant or small time delay plant ,that Dalin control algorithm and Smith predictor control algorithm can not guarantee the robustness to the uncertainty of the model ,which are fit for constant time-delay system, and that fuzzy control algorithm and neural network control algorithm, which do not need the precise model of the plant and have strong robustness to the uncertainty of the model, are proved to be effective methods for big time-delay plant. Key w