论文部分内容阅读
针对一类系统模型的输入输出函数不满足连续可导条件的非线性系统,设计了一种基于线性近似的模型预测控制方案。运用Stirling插值公式对非线性函数进行线性近似处理,并将模型预测控制性能指标重构为一个二次型最优化问题,通过对二次型最优化问题的求解得到了模型预测控制的最优控制序列。为了减少计算量,忽略了线性近似过程中产生的非线性高阶项。仿真结果表明,该模型预测控制方案控制效果满意,且具有降低控制能量消耗和缩短控制时间的优点。
Aiming at the nonlinear system which the input and output functions of a kind of system model do not satisfy the continuous derivable condition, a model predictive control scheme based on linear approximation is designed. The Stirling interpolation formula is used to linearly approximate the nonlinear function, and the model predictive control performance index is reconstructed into a quadratic optimization problem. The optimal control of the model predictive control is obtained by solving the quadratic optimization problem sequence. In order to reduce the computational complexity, the nonlinear high-order terms generated in the linear approximation are ignored. The simulation results show that the model predictive control scheme has satisfactory control effect, and has the advantages of reducing the control energy consumption and shortening the control time.