A Control Method of Converter DC Bus of Direct-Driven Wind Turbine

A Control Method of Converter DC Bus of Direct-Driven Wind Turbine

Xiaodong,Cao1,,Xuejun,Yang2,,Shijun,Liu1

1. Xinjiang Goldwind Sci & Tech Co.,Ltd., Wulumuqi 830026, China 2. Xinjiang Institute of Clean Energy Technology.， Ltd, Xinjiang 839000, China

Abstract: The direct-driven wind turbine realizes the wind turbine generator's control and grid connection control through the converter. The control of the DC bus voltage of the converter plays an important role in system control stability and device safety. Generally, the reference value of the DC bus voltage of the converter is a constant value, which is generally the maximum value under the conditions of stable operation of the converter under all operating conditions. Therefore, the bus voltage setting usually deviates from the actual minimum DC bus voltage that can be selected, which will adversely affect the efficiency of the converter, the life of the DC bus capacitor and the insulated gate bipolar transistor (IGBT). In addition, in some special grid conditions, The converter grid-connected control may have PWM exceeding the maximum value. This paper proposes the dynamic DC bus control method by analyzing the phasor relationship at the network side of the converter. Using this method, the controller can calculate the minimum bus voltage required for grid connection control in real time, so the bus voltage setting value can maintain the current minimum value and increase the system's reliability. The dynamic bus voltage control method proposed in this paper has the advantages of a clear physical relationship, simple and efficient, and accurate calculation.

Key Words: Direct-driven wind turbine, converter, floating DC bus voltage



1Introduction

The direct-drive unit adopts a full-power converter, so it has excellent grid-connection performance, strong fault voltage ride-through capability and zero voltage ride-through capability[1~3], and its reactive output replaces the reactive compensation equipment[4~6]. The full-power converter controls the DC bus voltage stability through the grid side

[7-8]. Usually, the reference value of the DC bus voltage is constant. This constant DC bus voltage setting mode has the following disadvantages:

1) When the DC voltage setting value of the converter is the minimum value that can meet the requirements of no over-modulation under normal working conditions, it is usually necessary to consider the factors such as the power supply at the network side of the converter and the maximum impedance voltage drop in case of generative reactive power. This mode gives the bus voltage sufficient margin, and the non-bus voltage setting value is designed to minimize. Under some conditions, PWM over-modulation will occur, and slight over-modulation will affect the power quality, in serious cases, the unit control will be abnormal, resulting in converter failure and shutdown.

2) When the DC bus voltage reference value is set to a constant value, it is the maximum value set according to the normal operation of the converter under various conditions, which has adverse effects on the converter efficiency, capacitor life and IGBT life.

The fan converter's DC bus voltage control strategy has been studied in the relevant literature. The paper [9] adopts the constant DC bus voltage reference value method, although comparing the absolute value of power difference with the power threshold value to output the absolute value of the difference, which is the difference by the amplitude of the grid voltage to obtain the change of the direct-axis current, thus reducing the dynamic fluctuation of the DC bus of the dual converter. Because of the constant DC bus voltage method, the utilization efficiency of the DC bus cannot be maximized under some conditions, and it can also not reduce the adverse impact on the life of the DC bus capacitor and IGBT. The paper [10] adopts the PI controller closed-loop method to realize the dynamic regulation of bus voltage. This method adds additional control links to DC bus voltage control, and needs to set PI parameters. Transient process will cause system instability, which is difficult to apply in engineering. The paper [11] proposes a method to control the switching of DC bus capacitance, which increases the DC capacitor capacity during the fault, further limits the DC bus voltage surge during the fault. Switching of DC bus capacitance reduces the power imbalance of the DC bus, and limits the DC bus overvoltage. But in this paper, the value of the DC bus voltage is not analyzed in detail.

The paper [12] analyzed the characteristics of the output instantaneous voltage of the converter by studying the circulating current of the parallel DC bus, and designed the strategy based on PR control according to the analysis results. In order to improve the dynamic response ability of the converter, the paper [13] proposes a joint controller design scheme based on the passivity theory. At the same time, it applies the passivity theory to design the controller for the tracking problem of the error system. In the paper [14], the diode rectifier direct-drive method is proposed to realize voltage stabilization and conversion, form a stable voltage source of 1100VDC, and use PWM inverter converter unit conversion to realize conversion and grid connection. These papers have studied the DC voltage control strategy, but did not research the given value of the bus voltage.

The research in the paper [15] shows that the DC voltage of the full power converter in the permanent magnet direct-drive wind power generation system will fluctuate greatly with the change of the output power of the motor, which will adversely affect the safe operation of the converter power devices and the stable operation of the entire power generation system. In this paper, a dual PWM converter coordinated control strategy for permanent magnet direct-driven wind turbine is proposed based on the operation characteristics of wind power generation system, but it is only limited to improving the stability of DC voltage. In the paper [16], the auto-disturbance rejection controller of the converter at the motor side and the grid side of the direct-driven permanent magnet synchronous wind turbine generator is designed respectively to achieve the maximum wind energy tracking control at the motor side below the rated wind speed. However, the DC capacitor voltage constant control is still used at the grid side of the system, and the DC bus voltage setting value is not analyzed in combination with the grid connection requirements at the grid side.

This paper analyzes the voltage phasor relationship and presents a calculation method of DC bus demand value. This method calculates the minimum bus voltage setting required for system control in real time. This method not only keeps the bus voltage setting value at the current minimum required value, but also avoids introducing additional control links into the control system and increases the system control stability. The bus voltage control strategy proposed in this paper has the advantages of clear physical relationship, simple and efficient calculation, and easy to apply in engineering. In particular, when the converter's capacitive reactive power affects the grid's voltage, this method can prevent the converter from over-modulating.

1.1Principle Introduction

The network side circuit topology of the full-power converter of the direct-drive unit is shown in Figure 1. The network side is a three-phase full-bridge topology, and the output end of the converter is a LC filter. The system voltage sampling point is the grid voltage and the current feedback is inductive current.

Fig. 1: Converter grid side topology

According to the topology of the grid side of the converter shown in Figure 1. Its voltage phasor relationship is shown in Figure 2 below with taking capacitive reactive power control.

Fig. 2: voltage phasor diagram

Where, Un is line voltage of the grid voltage; is the peak value of grid phase voltage; is reactive current; is DC bus voltage; is active current; is apparent current; is the line resistance from the converter port to the junction point; is the inductance value; is the grid angular frequency; is the converter port voltage.

The apparent current is ahead of the grid voltage of angle. The apparent current is decomposed into active current and reactive current . The voltage vector generated by the active current and reactive current on the line resistance is consistent with the direction of the current itself, and their values are respectively. The voltage vector generated by active current and reactive current on the inductor is 90° ahead of the current direction, and its value are

and . According to the phase voltage vector relationship, the voltage-mode value of the filter inductor on the grid side can be obtained as shown in Formula (1).

(1)

Where, is the sum of the positive sequence voltage and negative sequence voltage.

As shown in Figure 3, is two-phase static coordinate system, and is synchronous rotating coordinate system. is positive sequence voltage, and negative sequence voltage.

Fig. 3: Positive sequence negative sequence relationship of grid voltage

Formula (2) can be obtained from the vector relationship shown in Figure 3.

(2)

According to formula (2), when , the maximum value of is:

(3)

The calculation of the positive and negative sequence components of the three-phase grid voltage in the decoupled double-synchronous rotating coordinate system can be simplified according to formula (3).

The dead time of IGBT and the narrow pulse affect the DC voltage utilization. Let the dead time of converter be ，and narrow pulse time be . Bus voltage utilization is expressed as formula (4):

(4)

The control system has detection error and delay, such as DC bus voltage sampling accuracy SNR, grid voltage accuracy SNRv and three-phase current sampling accuracy SNRI. If the DC voltage sensor accuracy SNR is and the analog circuit accuracy is , then the DC bus voltage measurement accuracy is . If the measurement value is large, the actual value is small, which causes PWM over-modulation. Therefore, the maximum error of DC bus voltage acquisition accuracy should be calculated. Similarly, taking the maximum positive deviation of the grid voltage measurement accuracy SNRv and the current measurement accuracy SNRI.

In addition, the voltage drop of IGBT, control delay, active and reactive current fluctuation, grid voltage fluctuation and filter cut-off frequency can affect the bus voltage setting, and it can be solved by voltage compensation value obtained by experiment.

Through the above analysis, the following bus voltage demand calculation expression as (5):

(5)

The DC voltage compensation value includes the influence of the voltage drop of IGBT, control delay, active and reactive current fluctuation, grid voltage fluctuation and filter cut-off frequency.

To determine the value of , run the system and the algorithm after other parameters are set, and select the compensation value that makes PWM less than the modulation value.

2Experimental Platform

The experimental platform is shown in Figure 4. The designed algorithm is tested and verified on the RTDS hardware-in-the-loop simulation platform of Goldwind PCS09I converter with rated power is 3000kW.

Fig. 4: RTDS simulation platform

The system model is built in RSCAD as shown in Figure 5. Based on this model, hardware-in-the-loop simulation is performed. The grid side voltage model is designed as a constant voltage source to avoid grid impedance's influence on the experiment.

Fig. 5: rscad hardware in the loop simulation model

3Experimental Result

The algorithm is tested based on the hardware-in-the-loop simulation platform. The grid-side control of the converter adopts the active and reactive power decoupling control mode, and adjust the value under normal working conditions to make the PWM wave just pass the modulation. Then change the grid voltage, capacitive reactive power and inductive reactive power, to test whether SVPWM wave over-modulation and verify whether the DC voltage setting is reasonable.

3.11.0Pu and Reactive Power

Without changing the algorithm parameters, set the grid voltage to 1.0 Pu, make the converter output active power to 3000 kW, adjust the capacitive reactive value, and obtain the DC voltage setting value calculated by formula (5). Under the same condition, set the constant DC bus voltage value. By observing the SVPWM waveform, determine that the DC voltage required for the grid side control is exactly the corresponding value when the SVPWM is over-modulated. Then, the difference between the set value of DC bus voltage calculated by the algorithm and the demand value of DC bus obtained by the experiment can be obtained. The test results are shown in Table 1. The DC bus voltage value calculated by the algorithm is the DC bus value required by the actual system control under the voltage of 1.0 Pu grid side.

Table 1: 1.0 Pu grid side voltage under different capacitive reactive power

Fig. 8: Set value of bus voltage under different capacitive reactive power of 1.0pu grid voltage

Fig. 9 SVPWM value of 1.0pu grid voltage under different capacitive reactive power

3.20.95Pu and Reactive Power

In the same way, set the grid voltage to 0.95 Pu, make the converter output active power to 3000 kW, adjust the capacitive reactive value, and obtain the DC voltage setting value calculated by the formula (5). The test results are shown in Table 2. The DC bus voltage value calculated by the algorithm is the DC bus value required by the actual system control under the voltage of 1.0 Pu grid side.

Table 2: Test of 0.95 Pu grid side voltage under different capacitive reactive power

Fig. 6: set value of bus voltage under different capacitive reactive power of 0.95pu grid voltage

Fig. 7: SVPWM value of 0.95pu grid voltage under different capacitive reactive power

3.31.05Pu and Reactive Power

Set the grid voltage to 1.05 Pu, make the converter output active power to 3000 kW, adjust the capacitive reactive value, and obtain the DC voltage setting value calculated by formula (5). The test results are shown in Table 2. The DC bus voltage value calculated by the algorithm is the DC bus value required by the actual system control under the voltage of 1.0 Pu grid side.

Table 1: 1.05 Pu grid side voltage under different capacitive reactive power

Fig. 8 set value of bus voltage under different capacitive reactive power of 1.05pu grid voltage

Fig. 11 SVPWM value of 1.05pu grid voltage under different capacitive reactive power

3.4Prototype test

The prototype test is GW150-3.0. 0 kVar reactive power, 300 kVar capacitive reactive power and 500 kVar capacitive reactive power are tested respectively. Also, whether the set value of bus voltage is correct or not is judged by whether the SVPWM of the unit has over-modulation, as shown in Figures 11, 12, and 14. It can be seen that the set value of bus voltage calculated according to the algorithm meets the unit's operational requirements under normal operation and power generation.

Fig. 12 bus voltage value under zero capacitive reactive power of the unit

Fig. 13 SVPWM value under zero capacitive reactive power

Fig. 14 bus voltage value under 300kVar capacitive reactive power of the unit

Fig. 15 SVPWM value under 300kVar capacitive reactive power

Fig. 16 bus voltage value under 500kVar capacitive reactive power of the unit

Fig. 15 SVPWM value under 500kVar capacitive reactive power

4Conclusion

This paper proposes a dynamic DC Voltage control strategy for fan converter. By analyzing the principle of bus voltage composition through physical phasor calculation, obtains a specific calculation method to achieve the minimum voltage set value required by the system control. Compared with the constant bus voltage mode, it not only meets the requirements of converter grid-connected control, but also reduces the bus support capacitance and IGBT voltage stress, and improves the service life and converter DC bus efficiency. The proposed DC bus voltage control method has the advantages of a clear physical relationship, simple, efficient, accurate calculation, and is easy to apply in engineering. In particular, the method proposed in this paper can effectively solve the problem of over-modulation when the converter generates capacitive reactive power.

References

[1]Liu, B. , et al. Wind Power Status and Development Trends. Journal of Northeast Dianli University (2016).

[2]BinrongWen, et al. Power fluctuation and power loss of wind turbines due to wind shear and tower shadow.

Frontiers of Mechanical Engineering (2017).

[3]Bao, X. , et al. Review and Prospect of Low-Speed High-Torque Permanent Magnet Machines. Transactions of China Electrotechnical Society (2019).

[4]Sio, Jpmk , and R. Reginatto . Reactive power control of DFIG-based wind turbines for voltage support during faults. Simpósio Brasileiro de Sistemas Elétricos – SBSE 2020 2020.

[5]Li, Shuhui , et al. Characteristic study of vector-controlled direct driven permanent magnet synchronous generator in wind power generation. Electric Power Components and Systems 37.10(2009):1162-1179.

[6]Xiao, L. , et al. Harmonic Suppression for the Motor-side Converter of the Directly-driven Wind Turbine With PM Synchronous Generator. Proceedings of the CSEE (2011).

[7]Dong, Shuhui , H. Li , and Y. Wang . Low voltage ride through capability enhancement of PMSG-based wind turbine. International Conference on Sustainable Power Generation and Supply (SUPERGEN 2012) IET, 2013.

[8]Liu, Zhongyi , C. Liu , and G. Li . Power coordinated control of wind turbines with permanent magnet synchronous generator for low voltage ride through. 2014 IEEE Power & Energy Society General Meeting IEEE, 2014.

[9]Du, X. , and L. Zhu . Research on Control Method of DC Bus Voltage Based on Power Feedforward. Electric Power ence & Engineering 380-384(2013):3065-3068.

[10]Ma, Y. , et al. DC Side Bus Voltage Control of Wind Power Grid-Connected Inverter Based on Second-Order Linear Active Disturbance Rejection Control. Energies 12(2019).

[11]Chen, H. , et al. A coordinated control strategy for low voltage ride-though of direct-drive permanent magnet wind power generating units. Power System Technology 37.5(2013):1464-1470.

[12]Li, Q. , M. Chen , and F. Xiao . Modeing OF CIRCULATING CURRENT IN PARALLEL WIND POWER GENERATION SYSTEMS. Acta Energiae Solaris Sinica (2017).

[13]Wang, P. , et al. Passivity-based combined controller design for wind power applications. Diangong Jishu Xuebao/Transactions of China Electrotechnical Society 29.11(2014):201-209.

[14]Hansen, A. D. , and G. Michalke . Modelling and control of variable-speed multi-pole permanent magnet synchronous generator wind turbine. Wind Energy 11.5(2010):537-554.

[15]Cao, H. , et al. Control Strategy of Permanent Magnet Direct-Driven Wind Generation System Based on Hybrid Clamp Dual PWN Rectifier. Transactions of China Electrotechnical Society (2016).

[16]Wan, Jianru , et al. Direct Torque Controlled Permanent Magnet Synchronous Motor Based on Active Disturbance Rejection Controller. Journal of Computational and Theoretical Nanoscience 4.6(2011):2087-2091.

