Design Of Main Steam Temperature Control System For Supercritical Once-through Boiler Based On Multi-model

The main steam temperature of superheater outlet is one of the important parameters in the boiler operation process of thermal power plant.Its quality will directly affect the boiler thermal efficiency and coal consumption rate of power supply,and also related to the safe and stable operation of the whole generator set.Therefore,accurate control of steam temperature is the key to safe production,energy saving and cost control of power enterprises.With the reform of national energy policy,many small and medium-sized generating units with high emission and high coal consumption have been phased out,replaced by large unit generating units represented by supercritical technology.These units have high parameters,large capacity,strong nonlinearity,and tight coupling between various parameters,which makes its main steam temperature more difficult to control.In addition,large capacity thermal power units often have to undertake the task of peak regulation of the power grid,which means the load changes are drastic and frequent.Traditional cascade PID control can no longer meet the requirement of accurate control of the main steam temperature of supercritical thermal power units under the task of peak regulation.In order to solve the above problems,the steam temperature control system of supercritical once-through boiler is designed based on the historical data of plant-level Supervisory Information System(SIS)of a Shanxi power plant.Firstly,through the analysis and preprocessing of a large number of historical operating data,the improved particle swarm optimization algorithm with linear decreasing inertia weight and adaptive variation was used to identify the parameters of the steam temperature controlled object model under the three typical working conditions of 100% load,75% load and 50% load,by which the transfer function model of the steam temperature controlled object under each typical load was obtained.Secondly,the model predictive control theory is applied into the traditional steam temperature control.The main steam temperature cascade DMC control system of supercritical direct current boiler is constructed by the dynamic matrix control(DMC)as the main controller and the proportional control as the auxiliary controller.The control quality is better than the traditional cascade PID control under a single load.Finally,the multi-model strategy is utilized into the cascade DMC control system.By weighting the output signals of the sub-controllers under typical loads,the temperature of the main steam can be well controlled under a wide range of operating conditions.Simulation experiments show that the multi-model cascade predictive control system designed in this paper can not only effectively overcome the characteristics of large inertia and large lag of the steam temperature object of supercritical once-through boiler,and significantly improve the quality of steam temperature control,but also adapt to the task of peak regulation of large capacity units,and still achieve the safe operation index of units under a wide range of rapidly changing working conditions.It has certain engineering application value to realize automatic control of main steam temperature.