A Control Strategy with Improved Frequency Response Characteristics of Variable Speed DFIM Pumped Storage

With the gradual integration of renewable energy into the power grid, the demand for flexible regulation methods is increasing. To solve this problem, the variable speed pumped storage power station adopts a doubly fed induction machine (DFIM) and a variable speed pump turbine as its main components to achieve adjustable speed and open water diversion gates, thus realizing flexible power regulation under any conditions. However, research on the frequency response characteristics of DFIM in pumped storage is still in the preliminary stage. The commonly used frequency control methods focus on converting grid frequency deviation into active power correction, leading to suboptimal frequency response characteristics. Therefore, this paper proposes a method of frequency response strategy and dynamic adjustment of control parameters, aiming at enhancing the frequency response characteristics of DFIM, and making it more effective in integrating renewable energy into the power grid. Firstly, a topology of variable speed pumped storage unit is introduced. The speed priority control strategy and its advantages are further discussed to improve overall efficiency. Then, a frequency response module is proposed to convert frequency deviation and its rate of change into speed correction added to the speed controller. Meanwhile, a dynamic adjustment method of controller parameters, determined by the frequency characteristic during deviation, is proposed, resulting in a more reasonable release or obtainment of rotor kinetic energy. Subsequently, DFIM and synchronous machines with the same capacity are compared. The results show that DFIM can release or absorb more kinetic energy than synchronous machines, making it a crucial means to provide short-term power and inertia support to the grid. This paper conducts two simulations using Matlab/Simulink. The first one is about rotor speed and reactive power regulation, also known as the speed priority control of the DFIM. The second one is to verify the effectiveness of the proposed frequency response strategy. The first simulation results show that the machine can accurately adjust its rotor speed, torque, and reactive power when the reference values change. Due to the reference value being changed in the form of a step function, there is an overshoot in a short period, which must be avoided when used in an actual machine. Then, the proposed frequency response strategy is simulated. This strategy is compared to a commonly used frequency-active power control strategy and a constant-speed control strategy. At the beginning of the simulation, a load was added. The results show that DFIM can lower its speed to release kinetic energy, provide short-term active power, and provide inertia support. When the load is disconnected from the power grid, the power grid frequency increases, and DFIM accelerates, converting the extra active power into rotor kinetic energy. The proposed parameter dynamic adjustment method adjusts the parameters based on the rate of frequency change, shortening the frequency and speed recovery time. Based on theoretical analysis and simulation results, the following conclusions can be drawn: (1) When there is no frequency disturbance, the speed priority control can effectively control the motor speed, torque, and reactive power of DFIM. (2) The proposed frequency response module converts frequency deviation into speed correction, which can directly release or absorb the rotor kinetic energy. Compared with the frequency-active power control strategy, the strategy reduces frequency variation and steady-state error, and suppresses fluctuations during frequency variation. The frequency response characteristics in both generating and pumping modes are improved. (3) The parameter adjustment method allows the controller parameters to be dynamically adjusted, which can release or absorb the rotor kinetic energy more reasonably during the process. By setting the controller parameters appropriately, the inertia support capability of DFIM can be improved. This method requires little computing capability and can be handily promoted in practical engineering applications. © 2023 Chinese Machine Press. All rights reserved.