摘要
氧化槽进气量是生物氧化预处理过程中重要的操作调控参数,由于氧化槽内成分复杂,预处理过程具有强非线性、耦合性和大时滞性的特点。针对这一问题,提出一种以在线支持向量回归机(OS-VR)为模型的非线性预测控制方法。采用粒子群算法与梯度下降原理滚动优化非线性模型预测控制的目标函数,求得最优控制量。仿真结果表明,所提算法能够有效地对氧化槽进气量进行预测和控制,为生物氧化预处理过程提供新的方法。
Air input of biological oxidation pretreatment process is important control parameter.Togetherwith many other complicated factors of oxidation tank,pretreatment is strong nonlinearity,coupling andlarge time delay.A nonlinear model predictive control based on online support vector regression(OSVR)is proposed.After that,the optimal control law is obtained by the particle swarm optimization algorithmand rolling optimization.Simulation results show that the proposed algorithm can predict air input ofeach oxidation tan and control efficiently,and provide a new way for biological oxidation pretreatmentprocess.
作者
孟宪强
南新元
曾庆凯
Meng Xianqiang;Nan Xinyuan;Zeng Qingkai(College of Electrical Engineering,Xinjiang University,Xinjiang Urumqi 830047,China)
出处
《科技通报》
北大核心
2017年第7期56-60,共5页
Bulletin of Science and Technology
基金
国家自然科学基金(61463047)
关键词
在线支持向量回归机
滚动优化
进气量
预测控制
online support vector regression (OSVR)
receding horizon optimization
air input
predictive control