【摘要】 针对快速动态过程与低维非方多变量系统,提出了一种多变量约束快速动态矩阵控制(DMC)算法.采用动态优化和经济优化两层结构,自动实现非方系统的控制结构转换.在满足基本控制目标的基础上,利用多余的自由度实现经济目标优化.算法采用阶跃响应模型与参数模型相结合的方法改进了预测模型,消除了截断误差,仅取每个输出未来的一点进行优化并计算单步控制作用.与常规多变量约束DMC算法比较,该算法有效地减小了计算负荷,而性能损失很少,可以直接在现有集散控制系统(DCS)上实现.工程应用实例与仿真证明了该控制算法的有效性.更多还原

【Abstract】 For fast dynamic processes and non-square multivariable systems, a multivariable constrained fast dynamic matrix control (DMC) algorithm was presented. A two-layer architecture including dynamic and economic optimization induced an auto control structure transformation for non-square multivariable systems. The extra degrees of freedom could be utilized to calculate the optimal values of inputs when the basic control objects of outputs were met. The approach of coordinating the step response model with the parametric model was employed to improve output prediction and remove truncation error. The one-step control move was calculated by only optimizing one point in the future horizon for each output. Compared to the traditional DMC algorithm, the proposed algorithm greatly reduces the calculation burden and has better control performance, and it can be embedded in distributed control system (DCS). The effectiveness of the algorithm was validated by an industrial project and a simulated example.更多还原