论文部分内容阅读
针对在系统关联严重的情况下,通过选择合适的权矩阵,预测控制可以实现间接解耦控制,但是系统的跟踪性能将大大降低的问题,提出了一种带设定值观测器的多变量广义预测解耦控制算法,结合前馈解耦控制策略与广义预测控制律,将其他通道对当前通道的影响作为扰动,从而实现近似解耦。在目标函数中用控制量的增量代替控制量,使控制增量变化不至于过于剧烈。并且增加了一种新的设定值观测器,根据设定值与实际值的偏差实时改变误差权矩阵,进一步减弱了耦合。最后,磨煤机的仿真结果验证了该控制算法的优越性。
Aiming at the problem of indirect decoupling control, but the tracking performance of the system will be greatly reduced by choosing the appropriate weight matrix in the case of serious system correlation, a multivariable generalized system with a set-value observer is proposed Predictive decoupling control algorithm, combined with feedforward decoupling control strategy and generalized predictive control law, takes the influence of other channels on the current channel as a perturbation to achieve approximate decoupling. In the objective function, the amount of control is used instead of the amount of control so that the control increment does not change too drastically. In addition, a new set-value observer is added. According to the discrepancy between the set value and the actual value, the error weight matrix is changed in real time to further weaken the coupling. Finally, the simulation results of the mill validate the superiority of the control algorithm.