Study Of Dynamic Model Of Boiler Furnace Combustion And Multiobjective Model Predictive Control

As more attention has been paid to the improvement of air quality and ecological environment, thermal power enterprises have to increase the operating efficiency of the thermal power unit on the one hand, and manage to decrease the pollutant emission on the other, so as to meet the increasingly strict emission standard. Hence, to optimize the boiler for energy conservation and emission reduction is of value for both practical application and engineering application. The main research content includes:1. One-dimensional modeling of the furnace. The furnace is firstly divided into several levels along the height direction. Then the lumped parameter method is employed to capture the flowing, heat transfer, coal combustion and NOx reaction of each level. Then, a one-dimensional model can be got in a software called Dymola with Modelica. The dynamic characteristics of temperature and concentration of NOx in each level can be got as well as the statical distribution of temperature and concentration of NOx along the height of furnace.The dynamic characteristic of the furnace is also be studied during simulations for further study of combustion optimization.2. Build the space state models of the boiler furnace at 100%,80%, and 60% load efficiency respectively based on the subspace identification method. Compared with the traditional transfer function model, this model makes it more convenient to describe the multivariable system and to design a more advanced control algorithm. The result of application of the subspace identification method in boiler furnace combustion system is studied by simulation.3. Combine the multi-objective model predictive control with the multi-model strategy and the disturbance observer, and establish the optimization system of the boiler furnace. The efficiency and the concentration of NOx are the two objectives of the boiler furnace optimization; the multi-objective model predictive control helps to solve the multi-objective optimization; the multi-model strategy helps to deal with the nonlinear problem caused by extensively variable working condition. And application of disturbance observer can control the influence of coal changing. The effectiveness of MOMPC in boiler furnace combusion optimization system is studied in the situation of load changing and coal changing.