摘要
传统的预测控制在处理积分对象时,往往会因为模型与过程的失配,导致过程在受到持续干扰时会达到一个异于设定值的稳态。首先将积分对象的传递函数模型转换为扩展非最小化状态空间模型,然后结合预测控制完成对积分对象的控制。基于扩展非最小化状态空间模型的预测控制能有效地处理模型失配情况下的各种干扰,最后的仿真例子验证算法的有效性。
Caused by the model/plant mismatch in practice, the integrating process achieves the new steady state which is different from the set-pointwhen it encounters a continuous disturbance under the control of the conventional predictive control. Firstly the transfer function model ofthe integrating process is transformed into the extended non-minimal state space（ENMSS） model. Then the predictive control is introducedto deal with the integrating process based on such model. The proposed approach can handle various disturbances under model/plant mis-matched cases for the integrating process, and the case study illustrates its effectiveness finally.
基金
精细化工气体反应和精馏的先进控制和优化项目
国家国际科技合作专项项目:工业回转窑能源管理和控制一体化关键技术(No.015DFA60200)
关键词
积分对象
扩展非最小化状态空间模型
预测控制
干扰
Integrating Process
Extended Non-Minimal State Space Model
Predictive Control
Disturbance
