| Circulating fluidised bed boiler combustion systems are now used in large numbers in the power and heating industries due to their excellent fuel adaptability,low pollutant emissions and high combustion efficiency.However,there is a strong coupling and long lag between the combustion parameters of circulating fluidised bed boilers,and each parameter is strongly influenced by other factors in the system,showing strong non-smoothness.In the face of the complex combustion process of circulating fluidised bed boilers,conventional PID control has not been able to better realise the automatic control of boilers to achieve the goal of unattended operation,therefore,it is important to study the combustion system of circulating fluidised bed boilers to realise the automatic control of boiler combustion system.This topic is based on the actual demand of a heat source plant,combined with advanced control algorithms and software and hardware,to achieve automatic control of circulating fluidized bed hot water boiler combustion system,the main work of this thesis is as follows:(1)A simplified model of circulating fluidized bed hot water boiler was established.Taking a circulating fluidized bed hot water boiler of a heat source plant as the research object,the process flow of the hot water boiler combustion system is studied and the internal structure of the hot water boiler combustion system is analyzed.In view of the complexity of the combustion system structure of circulating fluidized bed hot water boiler,the multivariable complex combustion system is simplified into a model with three inputs and three outputs by analyzing the main sub-modules of boiler combustion,whose input characteristic variables are coal feed,primary air volume and induced air volume,and the corresponding output characteristic parameters are discharge water temperature,flue gas oxygen content and material layer temperature.(2)Study the model identification of the combustion system of circulating fluidized bed hot water boiler.In view of the fact that there are many factors affecting the boiler combustion system and the traditional mechanistic modeling method cannot establish an accurate mathematical model,a long-short time memory neural network based on the northern hawk optimization algorithm is designed to identify the simplified model of the boiler combustion system.The simulation results show that the long-short time memory neural network optimized based on the northern hawk algorithm can identify the model more accurately compared with the long-short time memory neural network.(3)To study the control method of circulating fluidized bed hot water boiler combustion system.In response to the poor control effect of traditional PID control in the boiler combustion system,generalized predictive control is chosen to be applied in the circulating fluidized bed hot water boiler combustion control system.An improved implicit multivariate generalized predictive control model is established by combining the long-short time memory neural network prediction model based on the optimization of the Northern Cthulhu algorithm.Based on the simulation platform to compare the conventional PID control and the improved implicit multivariate generalized predictive control,the superiority of the implicit multivariate generalized predictive controller designed in this paper is verified.(4)Realize the overall design of the combustion control system of circulating fluidized bed hot water boiler.Siemens S7-410 is selected as the central processor of the distributed control system to build the hardware system,and the schematic design and lower computer program writing are carried out.C# language is used to design the upper computer client,and the client interface is divided into different functional modules to complete the overall development of the boiler combustion system and realize the system unattended. |
