1 Introduction
Currently, although doubly-fed wind power systems still dominate the entire wind power generation system, direct-drive generator sets are beginning to attract more and more attention due to their inherent advantages [1]. Direct-drive wind power systems use wind turbines to directly drive multi-pole low-speed permanent magnet synchronous generators (PMSG) to generate electricity, and then convert the electrical energy through a power conversion circuit before connecting it to the grid. This eliminates the gearbox component, which has a high failure rate in traditional doubly-fed wind power systems, greatly improving system efficiency, effectively suppressing noise, and improving the reliability of system operation, thus gaining market favor.
2. Direct-drive wind turbine converter topology
Figure 1 shows the simplest forms of two topologies for a full-power converter in a low-voltage system.
In an active rectification topology, a three-phase voltage-source inverter replaces the uncontrolled rectifier and boost chopper units to control the generator load torque, thereby regulating the motor speed. This topology employs dual PWM (pulsewidth modulation) full-power converters, enabling high-performance generator control while avoiding the added complexity of the two-stage uncontrolled rectifier and boost chopper structure. It also reduces generator copper and iron losses and can adjust the generator power factor to 1, showing promising development prospects. Given the different focuses of the control strategies for the motor-side converter and the grid-side converter, this paper proposes a separate control method for the motor-side converter and the grid-side converter (system control block diagram shown in Figure 2), which can achieve effective control and thus generate high-performance dynamic characteristics.
3. Motor-side converter control strategy
This paper achieves maximum wind power tracking by controlling the generator set's rotational speed, enabling the generator speed to follow the constantly changing wind speed and extract more energy from the wind. When the wind speed is below the rated wind speed, the purpose of the system's speed control is to ensure that the unit operates in maximum wind power tracking mode. When the actual wind speed is higher than the rated wind speed, due to limitations such as mechanical strength, generator capacity, and inverter capacity, it is necessary to reduce the energy captured by the wind turbine to keep the power near the rated value. At this time, the pitch angle control needs to take effect to ensure that the unit remains near the rated power.
3.1 Maximum Power Point Tracking Algorithm for Wind Turbines Below Rated Wind Speed (mppt)
The power output of a wind turbine varies with its rotational speed. For any given wind speed, there is an optimal rotational speed that maximizes power. Therefore, the goal of wind turbine control is to control the rotational speed so that the turbine always operates at its maximum power output point. When the blade pitch angle is constant, there exists an optimal tip speed ratio λ that maximizes the wind energy utilization coefficient cp, which in turn maximizes the output power. According to the formula, to achieve maximum power point tracking (MPPT), the generator rotational speed must be adjusted according to the wind speed to maintain the optimal tip speed ratio.
The electromagnetic torque of a permanent magnet synchronous generator (PMSG) depends on the stator current of the motor. For direct-drive wind power systems using PMSG, there is no speed-increasing mechanism. Therefore, the wind turbine's speed at various wind speeds corresponds to the generator's speed, i.e., ω = ωg (ω is the turbine speed, ωg is the generator speed). Thus, to ensure the wind turbine's speed always follows the wind speed and remains at the optimal speed for that wind speed, the generator's rotor speed must follow the wind speed and maintain that optimal speed. Generator speed control requires first detecting the wind speed signal, then using the wind speed-optimal speed relationship to find the optimal speed. This optimal speed is then used as a reference speed and input to the motor driver. A speed closed-loop system helps the generator reach its optimal operating point. Since the generator's speed and electromagnetic torque are directly related, the torque element can be designed as the inner loop of the speed element. For permanent magnet motors, no excitation current is needed; the stator current only generates torque. In a rotating coordinate system, the electromagnetic torque of a permanent magnet motor, te = 1.5pψfiq, is only related to the q-axis current and not the d-axis current. Therefore, torque element control can be converted into current element control. Therefore, generator torque and speed control can be achieved simply by controlling the q-axis current. The speed control method is a closed-loop control system with current control as the inner loop and speed control as the outer loop. The main function of the generator-side converter is to adjust the output voltage signal ug and the electrical frequency fe according to changes in actual wind speed. Based on the vector control principle of permanent magnet motors, speed regulation can be achieved by controlling the phase and amplitude of the generator rotor current vector. From the torque formula of a permanent magnet motor, it can be seen that once the excitation flux linkage and right-angle inductance of the permanent magnet are determined, the generator torque depends on the space vector ig of the stator current, and the magnitude and phase of ig depend on id and iq. By controlling these two currents, the generator torque can be controlled. A certain speed and a certain torque correspond to a certain id and iq. By controlling these two currents, making the actual id and iq track the command values i*d and i*q, generator and speed control are achieved.
4. Control Strategy for Grid-Side Converters
Traditional grid-side converter control systems require AC voltage sensors, AC current sensors, and DC voltage sensors to detect control quantities and perform protective functions, increasing system cost and making the rectifier unit bulky. Furthermore, sensor signal loss and noise interference can degrade system performance. Therefore, researching sensor-omission control strategies is necessary. This paper, based on the traditional SVPWM method, uses a virtual flux linkage to calculate the angle, eliminating the need for AC voltage signal detection, thus reducing AC voltage sensors, lowering system cost, shrinking device size, and simplifying circuit structure. Moreover, it has a strong suppression effect on grid interference, less distortion of grid input current, and better dynamic and static control characteristics.
4.1 The concept of virtual magnet links
The concept of virtual flux is derived from virtual motor. The grid-side power supply (the dotted line part in Figure 3) can be regarded as a virtual AC "motor", and the resistance and inductance can be regarded as the stator resistance and stator leakage inductance of the virtual motor, respectively.
First, assume the three-phase power grid voltages are balanced, and neglect the incoming line reactors and line resistance r. Then, the voltage equations for the three-phase vsr in the αβ coordinate system are:
Equation (3) shows that an integrator is used when estimating the flux linkage. Problems with the initial value of the integrator can cause bias in the observed flux linkage. In this paper, two first-order low-pass filters are used to replace the integrator, eliminating the bias. The resulting virtual flux linkage observer is shown in Figure 4.
4.3 Estimation of instantaneous power
Instantaneous power is estimated using the measured line current and the estimated flux linkage. In the complex domain, instantaneous power can be calculated using the following formula:
p = re(ui * l)
q = im(uli*l)
i*l is the conjugate of il.
Instantaneous active power and reactive power can be expressed as:
p = w(ψlαilβ - ψlβilα)
p = w(ψlαilα - ψlβilβ)
4.4 SVM-DPC Control Block Diagram Introducing Virtual Magnet Link
Figure 5 shows the SVM-DPC control block diagram with virtual flux linkage introduced. Instantaneous power and flux linkage position angle can be estimated by measuring the current signal and DC-side voltage signal. The system uses the DC output voltage as the outer loop control. The difference between the output voltage and the voltage reference value is adjusted by the PI regulator to obtain the reference current value. This current value is fed into the multiplier along with the output voltage to obtain the reference value of active power. At the same time, the reactive power reference value is set to 0 to ensure unity power factor operation of the rectifier. The inner loop controls the instantaneous active and reactive power. The deviation between the calculated instantaneous power value and the reference value is output by the PI controller and transformed into the α-β coordinate system. The values ua and uβ are directly fed into the PWM modulation module to obtain the switching signal.
5. Simulation Analysis
To verify the above control method, this paper uses MATLAB/Simulink software to build simulation models of the motor side and grid side of a direct-drive wind turbine, as shown in Figures 6 and 7, respectively. The main simulation parameters are listed in the appendix, and the simulation results are shown in Figures 8 and 9.
The parameters of the permanent magnet synchronous generator are: rated speed n=750rad/min, rs=1.64ω; ld=0.01547h; lq=0.0258h; ψf=0.1848wb; pn=2; te=9.55n·m.
Figure 8 shows the waveforms of wind speed, generator speed, wind turbine output torque, generator torque, tip speed ratio, and wind energy utilization coefficient when the wind speed changes stepwise from 13 m/s to 14 m/s. As the wind speed changes, the d-axis current remains zero, while the q-axis current adjusts accordingly. The tip speed ratio remains optimal (λopt), and the wind energy utilization coefficient maintains its maximum value. Simulation results demonstrate the effectiveness of the control, showcasing the advantages of variable-speed wind power generation systems in capturing more energy and operating more stably than constant-speed systems. The feasibility of the generator-side control algorithm is verified. Maximum power point tracking (MPPT) of wind energy is achieved, and the steady-state and dynamic performance of the control strategy is validated.
Figure 9 shows the waveforms of phase a voltage and current, DC voltage, instantaneous active power, and instantaneous reactive power on the AC side of the virtual flux vector control system. Simulation results show that when the reactive power setpoint q*=0, the AC current waveform is close to sinusoidal, and the voltage and current are in phase, achieving unity power factor rectification operation. The active and reactive power exhibit good steady-state characteristics; the average value of p stabilizes at the setpoint p*, and the average value of q stabilizes at the setpoint 0, indicating good system control performance.
6. Conclusion
This paper employs a dual PWM converter as the grid-connected circuit for a direct-drive permanent magnet synchronous wind turbine, and proposes a control strategy that separates the control of the motor-side converter and the grid-side converter. Simulation results verify the correctness of the proposed control strategy: the motor side can effectively capture maximum wind energy and stabilize the DC-side voltage by tracking the optimal tip speed ratio when the wind speed is below the rated speed; the grid side can maintain the grid-side power factor and eliminates the need for an AC voltage sensor, exhibiting good dynamic and static performance. The clear division of responsibilities between the motor and grid sides makes the control method simple and effective.
