Study On Economic Model Predictive Control Based Cold-end System Optimization Of Direct Air-cooling Generating Unit

The direct air-cooling generating unit,where the ambient air is directly utilized instead of water to cool down the turbine exhaust steam,has the advantages of high water saving rate,flexible operation and simple structure.Therefore,it has been the first candidate for the thermal power plants in the northern China with abundant coal and rare water resources.The direct air-cooling condenser pressure is an important parameter for the cold-end system,and it is a comprehensive indicator of the unit's operating economy and stability.Therefore,it is of great significance to optimize the economic operation of the direct air-cooling generating unit cold-end system to reduce the coal consumption and improve the unit control performance and economy.To this end,this paper first studies the cold-end system dynamic modeling of the direct air-cooling generating unit,and applies advanced algorithms that incorporate economic optimization and control to the coldend system and boiler-turbine coordinated system to improve the unit operating economy.The main contributions of this thesis are as follows:(1)After taking into consideration the complicated couplings between the cold-end system and its connected equipment,a nonlinear dynamic mathematical model of the cold-end system is established,which can accurately reflect the relation among unit back pressure and main steam parameters,fan array rotation speed and ambient temperature,and the influence of the unit back pressure on the key thermal parameters of the turbine and regenerative system.This model is proved to be suitable for both simulation research and controller design.(2)A nonlinear optimization problem is built to maximize the net power output(the difference between power output and fan array power consumption)by optimizing the fan array rotation speed,therefore the optimal fan rotation speed and corresponding optimal unit back pressure can be obtained under given main steam parameters and ambient temperature.Besides,a tracking model predictive controller with guaranteed stability and an economic model predictive controller with guaranteed stability are designed for the unit back pressure control system to improve the system economic performance.The simulation results have verified the effectiveness of the proposed method.(3)Due to the vulnerability of unit back pressure to the ambient temperature,a zone model predictive controller and a zone economic model predictive controller are designed for the unit back pressure control system to reduce the fluctuation of the control quantity,and enhance the system robustness.Moreover,an event-triggered zone economic model predictive controller is proposed for the unit back pressure control,in which a trigger condition of judging whether unit back pressure is within the economic zone is introduced to determine whether to conduct the online economic optimization.The simulation results have proved that the proposed method has similar economic performance with economic model predictive control,and the controller online computation time has obviously decreased.(4)The boiler-turbine coordinated system model with coupled cold-end system model is built for a supercritical direct air-cooling generating unit.The unit main steam pressure and condenser back pressure are optimized simultaneously at given unit load and ambient temperature to minimize the unit heat rate.On the basis of this,the unit back pressure control is introduced into the traditional boiler-turbine coordinated control system.Since the fan rotation speed can be used to quickly adjust the unit back pressure and power output,the load response rate can be improved at the early stage in the load-changing condition.The simulation results have indicated that this new control structure can improve the control performance and reduce the coal consumption,and therefore improve the system economic performance.