| With the adjustment of energy structure in China’s "14th Five-Year Plan",combined cycle units using natural gas as the main fuel have been more widely applied.Compared to the power generation units using coal as the main fuel,combined cycle units have significant advantages such as fast start-up,high flexibility,and clean environmental performance.The operation of combined cycle units is susceptible to disturbances,in order to ensure the safe and stable operation of combined cycle units,it is necessary to solve the problem of susceptibility to disturbances,therefore,research on disturbance rejection is conducted using intelligent control methods targeting the typical control loops of combined cycle units.First,to meet the needs of key technologies for intelligent control of combined cycle units,analysis is conducted on the typical control loops of combined cycle units,and a framework for intelligent control key technologies is designed.Then,based on historical data,nonlinear auto-regressive model with exogenous inputs is established,and the results prove that the model can effectively fit the delay characteristics of the typical control loops of combined cycle units.Then,the impact of disturbances on combined cycle units is explored by conducting model-based disturbance analysis,obtaining the distribution form of the output under input disturbances in the typical control loop of the gas turbine,this provides a theoretical basis for solving the disturbance rejection problem.On this basis,the control strategy design of typical control loop of combined cycle units is carried out,To achieve good control capability and disturbance rejection effect,a phase advance error-based active disturbance rejection controller is designed.On the basis of the error-based active disturbance rejection controller,the controller improves the disturbance estimation ability of the extended state observer by introducing the phase advance,thus improving the control ability and disturbance rejection effect of the whole controller.By comparing the Bode plots of different controllers affected by parameters,the advantages of the new controller in observing disturbances are verified.The new controller is tested on the typical control loop of the combined cycle units,which verifies that the new controller can still maintain good performance under model disturbance.Finality,a multi-class student psychology based optimization is designed to tune the parameters of the controller,reduce the difficulty of tuning,and further improve the control effect.Based on the "rocket class-ordinary class" topology,the student psychology based optimization is integrated with the social network search algorithm.The local search ability and global convergence ability are balanced by the variable parameter of class number,and the sine cosine algorithm and the initialization of chaotic map are combined to improve the performance of the algorithm.The benchmark function test shows that the multi-class student psychology based optimization has both global and local performance advantages.The multi-class student psychology based optimization is used to tune the phase advance error-based active disturbance rejection controller to form intelligent control strategy,then apply it to the control process of the combined cycle units.The simulation results show that the intelligent control strategy designed in this study has a good effect in setting value tracking and disturbance rejection. |
