输入受限的广义预测控制快速算法

Fast Algorithm of Generalized Predictive Control with Input Constraints

针对传统约束广义预测控制算法的计算量大这一缺陷,通过只精确计算当前时刻的控制作用,而对未来时刻的控制序列进行离线近似计算,提出了一种输入受限的广义预测控制快速算法。该算法简单,不必求解Diophantine方程和逆矩阵,大大减小了在线计算量。并且,能够将输入很好地控制在约束范围之内,并具有良好的控制性能。仿真结果表明了该算法的可行性和有效性。

To avoid large amounts of computation of the traditional constrained generalized predictive control （CGPC） algorithms, a fast algorithm of GPC with input constraints is presented by exactly computing the current control action and approximately computing the future control sequence off line. The algorithm is more simple and dose not need solving Diophantine equation and the inverse matrix so that the on-line computation burden is strongly reduced. More importantly, good performance of control is guarantees ,while putting inputs and input increments in the constraints required. The simulation results show that the proposed algorithm is practicable and effective.