Data-driven Multivariable Adaptive Predictive Control of Municipal Solid Waste Incineration Process
摘要
针对城市固废焚烧(MSWI)过程炉膛温度、烟气含氧量等关键工艺参数难以精准控制的问题,提出一种数据驱动的多变量自适应预测控制方法.首先,通过引入操作变量约束和设计加权障碍函数,构造无约束形式的目标函数以降低优化求解的复杂度;其次,提出一种稀疏在线牛顿算法,通过稀疏分解海森矩阵,实现控制律的高效求解;同时,设计针对多变量控制的加权系数动态分配策略,能够根据实际运行状态自适应调整炉膛温度和烟气含氧量的权重,进一步提升控制性能;此外,对所提控制方法的可行性和稳定性进行分析,以确保其在实际应用中的可靠性;最后,采用 MSWI 厂实际运行数据验证了所提控制方法的可行性和有效性.
To address the issue of difficulty in precisely controlling key process parameters such as furnace temper-ature and flue gas oxygen content in the municipal solid waste incineration process,this paper proposes a data-driv-en multivariable adaptive predictive control method.Firstly,by introducing operational variable constraints and designing a weighted barrier function,an unconstrained form of the objective function is constructed to reduce the complexity of the optimization solution.Secondly,a sparsified online Newton algorithm is proposed,which achieves efficient solution of the control law through sparse decomposition of the Hessian matrix.Meanwhile,a dynamic weighting coefficient allocation strategy for multivariable control is designed,which can adaptively adjust the weights of furnace temperature and flue gas oxygen content according to the actual operating state,further enhan-cing the control performance.Additionally,the feasibility and stability of the proposed control method are ana-lyzed to ensure its reliability in practical applications.Finally,the feasibility and effectiveness of the proposed con-trol method are verified using actual operation data from a municipal solid waste incineration plant.