模型预测控制在钠氨合成工艺上的应用研究

Model Predictive Control Strategy for Sodium Amide Synthesis Technology

基于钠氨合成连续化生产工艺,从化学反应机理出发,建立钠氨塔式反应器动态机理模型。并对模型进行动态分析,对多变量非线性模型进行集总化、线性化处理。推广单变量模型预测控制到多变量模型预测控制,对反应器温度进行控制。通过仿真比较,验证了MPC控制算法比PID控制具有更好的控制效果。

The dynamic mechanism model of sodium amide tower reactor was build,based on the chemical reaction mechanism of sodium amide synthesis process,and linear process model was also got from this multi-variable nonlinear model.Meanwhile,multi-variable model predictive control was conducted by extending single-variable model predictive control,and used in the temperature control of reactor.Research shows that MPC algorithm has better control effect than PID control.