【作者】 蒋伟；

【导师】 叶思敏；

【作者基本信息】 四川大学 ， 控制理论及控制工程， 2001， 硕士

【摘要】 广义预测控制是80年代以来发展起来的一种新型计算机控制方法，它适用于开环不稳定系统和非最小相位系统，具有优良的控制性能和鲁棒性。但是，一般的广义预测控制算法存在一定的不足，很少考虑到算法在在线辩识和滚动优化时对系统稳定性和鲁棒性的影响。近年来，人们一直在研究鲁棒性强的预测控制算法，虽然取得了一定的成果，但是还很不完善。本文的算法正是在这种情况下提出的。在这篇论文中，提出了一种具有预测误差校正的ρ增量型广义预测控制算法，它能够通过预测误差增强系统在模型失配时的鲁棒性，同时，通过选择适当的增益因子ρ，可以改善系统的控制效果、提高系统的稳定性和鲁棒性。 本文首先介绍了几种具有代表性的广义预测控制算法，然后详细推导了具有预测误差校正的ρ增量型广义预测控制算法，得出了算法的公式。仿真实验表明：本文的算法能够改善系统的控制效果，增强了系统的稳定性和鲁棒性，从而验证了算法的有效性和可行性。更多还原

【Abstract】 Generalized Predictive Control (GPC) is a new computer control algorithm, which has made great progresses from its introduction since 1980’s. It is capable of stable control progresses with nonminimum_phase and open_loop unstable plants. Though GPC has excellent control performance and robustness, the common GPC algorithm has some drawback, it rarely considers of the influence on system stability and robustness caused by online identification and rolling optimization. Some achievements have been got in GPC algorithm with good robustness in recent year, but they are not good enough. The algorithm presented in the paper is on the basis of this point. In this paper a GPC algorithm proposed with predictive error correcting and increasing factor, which is capable of enhancing the system robustness by predictive error, meanwhile, it can also improve the control effect by selecting proper increasing factor, as well as the system stability and robustness.This paper firstly introduces several typical GPC algorithm, then carefully deduces the GPC algorithm with predictive error and increasing factor and get some formula. The results of simulation indicate that the algorithm proposed inthis paper is capale of improving the control performance, promoting thesystem stability and robusiness, and they also show that the promoting algorithmis effective and feasible.更多还原