基于ARM-Linux平台的显式模型预测控制算法实现

Realization of explicit model predictive control system based on ARM-Linux platform

为了对显式模型预测控制算法进行实验研究,建立了基于ARM-Linux平台的显式模型预测控制实物仿真控制系统.在ARM-Linux平台上实现了显式模型预测控制算法的在线计算过程,把传统的隐式模型预测控制的反复在线优化计算问题转变为在线数据搜索过程,简化了在线计算过程,提高了在线计算速度,减少了在线计算时间.对显式模型预测控制算法的离线计算步骤、实物仿真控制系统设计的整体结构、离线计算得到的被控系统状态分区及其对应分区上的显式线性控制规律库结构、接口函数的结构和ARM-Linux环境下的控制系统结构进行了讨论.最后给出了基于ARM-Linux平台的显式模型预测控制算法控制效果.

In order to conduct experimental research of the explicit model predictive control algorithm, an ARM-Linux platform-based physical simulation control system of the explicit model predictive control algorithm is developed. This control system realizes the online computation process of the explicit model predictive control algorithm on the ARM-Linux platform, and transforms the online optimiza tion programming problem of the traditional model predictive control system into online data search, simplifying the online computation procedure, speeding the online computation and reducing the online computation time. The procedures of off-line computation of the explicit model predictive control algorithm, the architecture of the physical simulation control system, the regions of the explicit control- ler and its corresponding optimal control law, the architecture of interface function and the architecture of the control system on the ARM-Linux platform are also presented. Finally, control results of the physical simulation control system of the explicit model predictive control algorithm are provided.