Research On Advanced Control Of Coal Gasification Process Based On Mutiple Models Switching And Model Self-adaption Control

Coal has always taken the leading position among the three major fossil fuels in China.At present,with the goal of efficient and clean energy,higher demands are being placed on the processing,conversion and utilization of coal.Coal gasification technology is the core technology for the clean conversion and utilization of coal.Coal gasification is a process in which coal and oxygen are used as raw materials and react under high temperature and pressure to produce fuel gas such as carbon monoxide and hydrogen.The coal gasification process is accompanied by the change of coal quality,which leads to the instability of working conditions and large fluctuations in the production process.Compared with the petrochemical industry,the application of advanced control in the coal chemical industry is more challenging.The reason is that the traditional advanced control method based on a single model is difficult to adapt the changing conditions of coal quality.Aiming at the coal quality variation problem of coal gasification process,combined with the production reality of coal-water slurry gasifier,this thesis carries out the research of multi-model switching control method and model self-adaption control method.Firstly,the multi-model controller structure of coal gasifier is designed,and the model set of multi-coal conditions is established offline.During the online operation of the controller,the integral squared error-total squared variation is used as the performance index of the controller to characterize the change of coal quality.An advanced controller model switching criterion is defined.The convergence of the model switching process is also demonstrated.Secondly,a self-adaption method for model predictive controllers is proposed,which is based on the principle of Zhang Neural Network(ZNN).The essence of the method is to adjust the parameters of the controller model by the ZNN algorithm to re-match the model between the controlled object and the manipulated variables when the coal quality fluctuates or changes.The convergence of the ZNN algorithm for the model self-adaption solution process is also demonstrated,ensuring the feasibility of the model self-adaption algorithm.Finally,the dynamic matrix control and performance monitoring software is developed.According to the dynamic matrix controller principle and multimodel switching criterion,the software designs the controller for the coal gasification process;according to the advanced controller evaluation principle,the controller performance is monitored and evaluated.The DCS configuration and operation screen of the advanced control system were designed and industrially applied on a large coal-water slurry gasifier of a chemical company.The application results show that,compared with the traditional manual operation,the advanced control method reduces the control deviation of key variables,while increasing the output of gasification products,reducing the raw material consumption of coal gasification industrial process and improving the economic efficiency of coal gasification plant,thus providing an effective way to realize the advanced control of coal gasification industrial process.