| With the continuous consumption of traditional energy and people’s continuous pursuit of the concept of environmental protection,various countries begin to develop wind and solar energy and other clean energy for energy generation,so as to meet people’s needs.However,due to the instability and discontinuity of these natural resources,the power grid will be impacted,resulting in greater security risks.Energy storage technology can effectively avoid the above problems,among which the micro compressed air energy storage technology has attracted social attention due to its advantages such as not being limited by the system site selection,environmental friendliness,convenient energy acquisition and long service life.Since the speed of the system needs to be controlled before the compressed air energy storage is connected to the grid for power generation,in order to prevent the instability caused by the speed fluctuation from affecting the safety and stability of the power grid and the quality of electric energy.In this paper,the miniature compressed air energy storage system is taken as the research object.After proper modeling and simulation,the speed characteristics of the turbine unit are optimized and analyzed.This paper first based on the general process of system identification.The operation data obtained from the test stand are selected appropriately,and the original data are preprocessed by the wavelet analysis noise reduction method.Based on the working principle of high pressure air in turbine unit determines the mechanism modeling and system identification toolbox of MATLAB hybrid modeling method,has realized the compressed air energy storage system of the valve,nozzle indoor air volume and expander rotor model is established and the model validation,the results show that the modeling method can better fit the real system data.Then,the Simulink simulation module under the speed control system is established.Based on the basic PID control strategy,the speed of the turbine expander is controlled from the rising speed stage to the stable speed stage.The turbine speed was controlled by fuzzy PID control,BP neural network with S function and BP neural network with adaptive learning rate optimized by Adagrad.The PID parameters and the output response curve of the speed were optimized.The optimization performance is compared and analyzed based on the optimization results of the four control strategies.The results show that the general PID has a poor control effect on the speed,the overshoot is 7.07% and the system stability time is 612.20 s,while the overshoot of BP neural network and the improved BP neural network is 0%,and the system is more stable.The system stability time of the fuzzy PID control and the BP neural network PID with adaptive learning rate are both within 90 s.Compared with the 612.20 s system,the response time is greatly shortened and the optimization effect is good.Finally,the influence of the possible interference signals on the dynamic characteristics of the expander speed is analyzed through the model under the four control strategies.Transient and continuous perturbations are used to analyze the speed perturbations of the turbine during the speed rise and the steady stage.The results show that the anti-interference ability of the system speed under the control of fuzzy PID and improved BP neural network is relatively strong.Then,the model robustness analysis of the BP neural network control system with adaptive learning rate is carried out.The results show that the control system has good robustness and meets the basic requirements of speed control before grid-connection. |
PDF Full Text Request