阶梯式预测控制在SCR脱硝控制中的应用

Application of Step Predictive Control in Denitration Control of SCR

SCR脱硝系统是一个非线性、大惯性、大时滞的系统,采用传统的PID控制不易得到理想效果,而传统的DMC控制算法需要进行复杂的矩阵求解,其实时性和计算强度无法满足热工在线控制要求。因此,在利用改进最小二乘法辨识得到SCR对象模型后,提出了一种基于阶梯式求解策略的改进DMC算法(SDMC),通过阶梯式策略避免了DMC的矩阵求逆过程。同时将矩阵计算简化为向量计算,在不降低DMC本身优越控制性能基础上,大大提升了计算效率,更利于工程实际应用。通过在SCR系统上与PID设定值扰动、输入NO_(x)扰动以及模型失配等对比仿真可知,SDMC能良好地响应NO_(x)出口浓度变化,可快速克服系统扰动,在对象模型不匹配时仍有不错的鲁棒性,可作为火电厂复杂热工对象的一种适宜先进控制策略。

The SCR denitrification system is a nonlinear,large inertia and large time delay operation system.It is not easy to obtain ideal results by using traditional PID control.However,the traditional DMC control algorithm cannot meet the requirements of thermal on-line control because of its real-time and computational performance due to the complex matrix solution.Therefore,after using the improved least square method to identify the SCR object model,an improved DMC algorithm(SDMC)based on the step-by-step solution strategy is proposed.It avoids the matrix inversion process of DMC through the step-by-step strategy,and simplifies the matrix calculation into vector calculation.On the basis of not reducing the superior control performance of DMC itself,the calculation efficiency is greatly improved,which is more conducive to the practical application of engineering.By comparing the SCR system with PID set value disturbance,input NO_(x) disturbance and model mismatch,it can be seen that SDMC can respond well to the change of NO_(x) outlet concentration,overcome the system disturbance quickly,and still have good robustness when the object model does not match.It can be used as a suitable advanced control strategy for complex thermal objects in thermal power plants.