Research On Data Driven Control Methods For Clean Energy Power Generation Processes

Gas-steam combined cycle units and doubly-fed wind power generation units both use clean energy to generate electricity,which is of great significance to reduce fossil energy consumption and greenhouse gas emissions,and improve power generation efficiency.In order to ensure the safe and efficient operation of the generation units,it is necessary to design the control system for them.The mechanism of combined cycle power generation process and doubly-fed wind power generation process are complex,nonlinear,multivariate,strongly coupled,and easily perturbed by operating conditions during operation.It is difficult to establish accurate mathematical models,therefore,the traditional model-based control algorithms can not easily achieve satisfactory control performances.In this dissertation,the control strategies for combined cycle system and doubly-fed wind power generation system are studied under the framework of data driven control and the main work is as follows:(1)For the gas-steam combined cycle system without known model,an equivalent dynamic linearized data model of the system is established by using the input and output data.The system nonlinearities and uncertainties are compressed into a time-varying parameter.Then the parameter adaptive law and model-free adaptive control law are given by designing two criterion functions,respectively.Finally,simulation experiments are conducted and the results show that the control strategy can effectively improve the load tracking capability,anti-disturbance capability and robustness of the combined cycle system without the need of the mathematical model.(2)To address the maximum power point tracking problem of doubly-fed wind power generation system,stator flux directional vector control technology is used to decouple the power of the generator.The double closed-loop generator control system with the inner loop as the current loop and the outer loop as the speed loop was designed.A maximum power tracking method based on model free adaptive learning control(MFALC)algorithm is presented.The model free adaptive learning controller is designed based on the input and output data of the wind power generation system.Using this controller,the optimal reference speed at different wind speeds is calculated in real time.Then maximum wind energy capture can be achieved by regulating the generator speed according to the optimal reference speed value.The proposed control method does not rely on the accurate mathematical model of the wind power generation system,and avoids mechanism modeling and system identification.Through the comparative simulation experiments with step stable wind speeds and natural wind speeds,the effectiveness and superiority of the proposed method are verified.