关键词: 压缩空气储能, 电力系统, 先进绝热压缩空气储能(AA-CAES), 模型预测控制(MPC), 功率跟踪, 动态模型, 状态空间, 质量流量

Abstract: [Objectives]　The discharge process of compressed air energy storage (CAES) systems involves complex nonlinearities, multi-control-variable coupling effects, and state constraints, posing significant challenges for power tracking control. To enhance both the dynamic response performance and the tracking accuracy of power under safe operational constraints of the turbine, a power tracking optimization control strategy based on model predictive control is proposed. [Methods]　First, a multi-time-scale state-space dynamic model integrating the aerodynamic and thermodynamic processes for the multi-stage turbine on the discharge side of the CAES. Second, the model is transformed into a discretized state-space equation to serve as the power prediction model. Then, using the throttle valve opening and heat-transfer fluid flow rate of the heat exchanger as control variables, and taking maximum power tracking accuracy and minimum control cost as optimization objectives, a model predictive control-based power tracking strategy for the discharge side is proposed, considering operational safety constraints of components. Finally, leveraging actual power plant operational data, a simulation model for the discharge side is developed using MATLAB/Simulink to investigate the dynamic response characteristics and multi-variable coupling relationships under typical system disturbances. [Results]　Compared to PID decoupling control, the proposed method achieves a 56% reduction in the settling time under constrained control inputs, along with improved dynamic response performance and tracking accuracy. [Conclusions]　The study provides a theoretical basis and technical support for the flexible regulation of CAES power plants.

Key words: compressed air energy storage, power system, advanced adiabatic compressed air energy storage (AA-CAES), model predictive control (MPC), power tracking, dynamic model, state space, mass flow rate