PMSG-Based Wind Energy Conversion Systems Integration Into DC Microgrids With a Novel Compact Converter

DC microgrids (DC-MGs) are recognized as an efficient alternative for common AC-MGs. DC-MGs need fewer power electronic converters since most of distributed generation (DG) units and electronic loads operate with DC voltage. The integration of DGs that generate AC voltage, such as wind energy conversion systems (WECSs) into DC-MGs, however, requires an AC-DC conversion unit to perform AC-DC conversion and galvanic isolation and to control the WECS. Typically, two separate cascaded converters are used to perform such functions, which increases the cost and size of DC-MGs. In this paper, an isolated single-stage AC-DC converter is proposed that is able to perform all the required function of a WECS AC-DC conversion unit, to overcome the drawbacks of the two-converter structure. First, the operation of the proposed converter and its features as a single unit are described. The operation of the WECS with the proposed converter is then studied. A design procedure for both the steady-state and dynamic operation of the converter is developed based on a mathematical model. Experimental results obtained from a scaled-down prototype converter are also presented to confirm the feasibility of the proposed converter as a single unit. The complete WECS operation is verified by the simulation results of the aerodynamic, mechanical, and electrical system models, using MATLAB/SIMULINK.


FIGURE 1. WECS unit integration in a DC-MG (a) proposed configuration in
, [16], (b) circuit diagram of a typical two-stage AC-DC converter.
circuit diagram is shown in Fig. 1. This two-stage converter is an expensive solution because it has two power stages with 12 switches and two controllers.
The cost of the power electronic converter is a considerable portion of the overall cost of small-scale WECSs in comparison with large-scale wind farms [17]. Any cost reduction in the power electronic converters will thus significantly reduce the overall cost of WECSs.
Isolated single-stage AC-DC converters that are introduced for power supply applications as simpler, lighter, and less expensive converters also can be used in AC-DC conversion units of WECSs. These converters share the active switches between the front-end AC-DC converter and the back-end DC-DC converter stages and a single controller is used to control the converter. Some modifications, however, are required to use a single-stage AC-DC converter for WECS application because the AC-DC converter functionalities in a WECS are more complex than those in a power supply application.
An isolated single-stage AC-DC converter for WECS should perform the required control for the WECS on top of AC-DC conversion, PFC correction, and galvanic isolation. In order to investigate the possibility of using isolated AC-DC converters for a WECS, a brief review of such converters is presented as follows: • Single-stage AC-DC converters with a fixed duty cycle operation [18], [19]: In these types of converters, the input power is controlled by operating converter switches with a fixed duty cycle over the full input voltage cycle and PFC is naturally performed with the sinusoidal envelope of a train of triangle current waveform. This method is simple to implement, but the input current has high peak components that increase the peak current stress on the components and cause EMI noise, and thus can be used only for low power applications < 2kW. Moreover, it is very difficult to control the input power for MPPT and power regulation applications since a high bandwidth current sensor and voltage sensor are needed to calculate the real-time power.
• Single-stage AC-DC converters with a matrix converter topology [20]- [22]: In these types of converters, the input current is smoother than converters with a fixed duty cycle. These converters, however, require a significant number of switches and their control strategy is too complicated.
• Single-stage AC-DC converters with a diode bridge at the input stage [23], [24]: In these types of converters, the diode bridge has high conduction losses; thus, their application is limited to low power applications.
• Single-stage AC-DC converters with non-standard control methods [25]: In these types of converters, complicated control methods (i.e., variable duty cycle and frequency) are used, thus real-time power calculation and the implementation of MPPT is difficult.
• Single-stage AC-DC converters based on the standard voltage source converter (VSC) structure [26]: In these types of converters, the input current is continuous and the control system is operation is well-known for both the scholars and engineers; this converters are thus promising for use in WECS applications. It should be noted that the proposed converter in [26] was originally proposed for power supply applications. The main contributions of this paper are as follows: • Three isolation transformers are used to extend the operating power range of the converter proposed in [26].
• A new control system is proposed that enables the isolated the single-stage AC-DC converter in [26] to perform MPPT and power regulation in addition to all the control objectives in [26].
• A design procedure for the components and control system is developed by using a mathematical model of the proposed converter. 83584 VOLUME 8, 2020  • A scaled-down prototype of the proposed converter was built to confirm the feasibility of the converter as a single unit.
• A comprehensive aerodynamic, mechanical, and electrical simulation was performed in MATLAB/SIMULINK to demonstrate the converter effectiveness in a WECS.
The following parts of this paper are organized as follows.
In Section II, the proposed converter topology and operation basics are explained. In Section III, the proposed control system for the converter is explained. In Section IV, experimental results from a scaled-down converter are presented to verify converter feasibility as a single unit, and simulation results of the aerodynamic, mechanical, and electrical systems are shown to verify the effectiveness of the proposed converter in a complete WECS. Conclusions presented in Section V.

II. THE PROPOSED CONVERTER
The proposed isolated single-stage AC-DC converter is shown in Fig. 2. The converter consists of an input filter L f and C f , a six-switch three-phase inverter (S 1 −S 6 ), a capacitor in the intermediate bus (C 1 ), three dc blocking capacitors , three transformers with a turns ratio of n : 1 and leakage inductance L lk , six output rectifying diodes (D 1 − D 6 ), and an output filter (L o and C o ). The proposed converter is composed of two separate sub-converters in its topology: a six-switch three-phase AC-DC converter that will be referred to as the AC-DC sub-converter and a three-phase full-bridge (FB) DC-DC converter that will be referred to as the DC-DC sub-converter. Switches are shared between the AC-DC and DC-DC subconverters.
The key element of this design is that the two subconverters work with different frequency components of the waveform. The AC-DC sub-converter operates based on the low-frequency component of the voltage at converter terminals (A, B, and C) and the DC-DC sub-converter operates with high-frequency voltage components at the converter terminals (A, B, and C); thus, the two converters can be analyzed as two decoupled converters.
The circuit diagram of each sub-converter is shown in Fig. 3. In the following section, the operation of each subconverter is briefly explained, then the operation of the overall converter is explained.

A. AC-DC SUB-CONVERTER
The AC-DC sub-converter ( Fig. 3(a)) is a six-switch VSC. Like conventional VSCs, the AC side of the AC-DC subconverter is connected to an AC voltage source (the PMSG), but the DC side is floating. The AC-DC sub-converter operates as a current-controlled voltage source. This means that the AC-DC sub-converter generates a three-phase sinusoidal voltage with respect to the modulation signals (m abc ), as shown in Fig. 4(a). The input current is controlled so that the active power generated by the PMSG is converted to DC power and stored in the intermediate DC bus capacitor (C 1 ). Moreover, the reactive power exchange between the PMSG and the AC-DC sub-converter is set to zero to minimize the power losses in the PMSG and the converter. In order to implement such a strategy, the decoupled power control method in the rotor synchronous reference frame (dq-frame) is used [27].  As stated, the power transfer in the AC-DC sub-converter happens through the low-frequency component of the voltage and current; thus average voltage equations are used to model the operation of the AC-DC sub-converter. The KVL equations in the stator in the dq-frame can be expressed as L s di ds dt = V ds − R s i ds + ω g pL s i qs (2) where V qs , V ds , I qs , and I ds are the q-axis and d-axis components of the stator voltages and currents in the rotor dq-frame, respectively. L s is the stator inductance, R s is the stator resistance, ω g is the PMSG angular speed, ψ is the magnetic flux of the permanent magnets, and p is the number of the PMSG pole pairs. The control equation can be derived from (1) and (2) as follows: As can be seen from (3) and (4) the control equations for the AC-DC sub-converter are the same as those of a conventional VSC. The control structure for each current component is shown in Fig. 5. A PI controller is used in each loop to control the current at the converter terminals. The design details for the PI controller parameters of a VSC are explained in detail in [28].

B. DC-DC SUB-CONVERTER
The DC-DC sub-converter ( Fig. 3(b)) is a three-phase PWM full-bridge (PWM-FB) DC-DC converter and has the same topology as the one proposed in [29]. The difference between the converter proposed in [29] and the DC-DC sub-converter is in their operation methods. In the former converter, the duty cycle of the converter switches are set by the output voltage controller, but in the latter converter, the duty cycle of the converter switches are set by the AC-DC sub-converter controller.
As stated, the DC-DC sub-converter transfers energy from the intermediate DC bus capacitor (C 1 ) to the output by using the high-frequency components of the voltage waveforms at point A, B, and C; thus, the DC-DC sub-converter operation can be explained based on the waveforms of the voltage over the switching cycle ( Fig. 4(b)). Fig. 4(b) shows the modulation signal, the carrier signal, associated gating signals, and the voltage across the converter terminals (V AB , V BC , V CA ) over the switching frequency, in the shaded area of Fig. 4(a). It should be noted that the gating signal of the lower switch of each leg is complementary to that of the corresponding upper switch of the same leg, with an appropriate dead-time inserted to avoid any shoot-through.
As can be seen from ( Fig. 4(b)), the terminal voltages are a square waveform and a DC component across point A-B, B-C, and C-A. A DC blocking capacitor blocks the DC component of each square wave and the AC components are applied to the appropriate transformer primary winding.
The AC component passes through the transformers, and after rectification, a low pass LC filter removes the switching frequency components from the voltage. The output LC filter is designed to prevent the inductor current dropping to zero (continuous conduction mode). This condition will reduce the current stress on the converter components and make the output current smoother.
Under these conditions, DC-DC sub-converter modes of operation can be derived. The modes of operation are explained for a set of typical modulation signals between (2π/3 < ωt < 5π/6). The equivalent circuit associated with each mode is shown in Fig. 6. It should be noted that the DC-MG is modeled with its Thevenin equivalent circuit, which includes a Thevenin voltage source (V DC−MG ) and Thevenin resistance (R DC−MG ). The blocking capacitors block the phase-to-phase low-frequency component of the voltage; thus, at t 0 , the relation between blocking capacitor Mode 1 (t 0 < t < t 1 ): At the start of this mode, switch S 5 is turned off. The converter terminal voltages are as follows: Since the primary of T B is exposed to the highest positive voltage and the primary of T C is exposed to the highest negative voltage, the secondary side current exits from T B and flows through D 3 and returns through D 2 to T C . The transformers are wye-connected at the secondary side so that the transformer primary currents are the same (I Lo /n). C b−B is charged and C b−C is discharged according to the direction of the primary currents.
Mode 2 (t 1 < t < t 2 ): At the start of this mode, switch S 3 is turned off. The converter terminal voltages are as follows: Since the primary winding of T A is exposed to the highest positive voltage and the primary winding of T B is exposed to the highest negative voltage, secondary side current exits from T A and flows through upper diodes D 1 and returns through T B , through lower diode D 6 . C b−A is charged and C b−B is discharged according to the direction of the primary currents.
Mode 3(t 2 < t < t 3 ): At the start of this mode, switch S 1 is turned off. The converter terminal voltages are equal to zero in this mode. According to blocking capacitor voltages, the primary of T C is exposed to the highest positive voltage and the primary of T B is exposed to the highest negative voltage so that the secondary side current exits from T C , flows through upper diodes D 5 and returns through T B , through lower diode D 6 . C b−C is charged and C b−B is discharged according to the direction of the primary currents.
Mode 4 (t 3 < t < t 4 ): At the start of this mode, S 1 is turned off. This mode is the same as Mode 2.
Mode 5 (t 4 < t < t 5 ): At the first of this mode, switch S 3 is turned off. This mode is the same as Mode 1.
Mode 6 (t 5 < t < t 6 ): At the first of this mode, switch S 5 is turned off. This mode is the same as Mode 3.

C. CONVERTER DESIGN PROCEDURE
The design of several key converter components such as the blocking capacitors and the output filter was explained in detail in [26] and thus wil not be explained in this paper. In this section, only the procedure for selecting the transformer turns ratio and switch ratings will be explained.

1) TRANSFORMER TURNS RATIO
As shown in Fig. 4(b), the voltage across the transformer's primary winding is the same as that of a standard full-bridge converter so that the same design procedure as the one presented in [30] can be used to design the transformer. The main parameter needed to design the transformer is the turns-ratio and the selection of a turns-ratio value can be done according to the following procedure.
Based on an analysis of the DC-DC sub-converter, V C o is a function of the intermediate DC bus voltage (V bus ), the modulation index (M), and the turns-ratio of the transformer (n). M is governed by the AC-DC sub-converter control equations; thus, V bus and n are the key design parameters in this procedure. Since V bus is limited by the voltage rating of the converter switches, n is expressed based on the converter parameters.
The modulation index can be determined by solving (1) and (2) under steady-state operatingon conditions as follows: whereÎ a is the peak value of the PMSG phase current.
It should be noted that I ds is set to zero by the controller to perform PFC. The value of V bus is a design parameter and should be chosen so that the values of M d and M q meet the PWM linear operation condition at the converter rated power, according to Based on Fig. 4(b), a unipolar double-edge carrier signal is used to generate PWM signals. The double Fourier analysis method is used to decompose the converter terminal voltage waveforms (V AB , V BC , and V CA ) [31]. Since a single carrier signal is used for all modulation signals, the switching frequency harmonic and triplen (odd multiples of the third harmonic) sideband switching harmonics do not appear in the harmonic contents of V AB , V BC , and V CA . The terminal voltage Fourier transform for V AB can be expressed as where ω e is the angular frequency of the PMSG terminal voltage, ω c is the angular frequency of the carrier signal, θ c is the initial phase of the carrier signal, J x is a Bessel function of the first kind, U 1 = {±(3k − 2)|k ∈ N}, and U 2 = {±(3k − 1)|k ∈ N}. The harmonic contents of V BC and V CA waveforms are the same; only π/6 is replaced by −π/2 and −7π/6, respectively. VOLUME 8, 2020 Blocking capacitors block the dc and low-frequency components of (8) and high-frequency components pass through the transformer, thus, the secondary voltage of T A can be written as According to (9), the secondary voltage of the transformers can be decomposed to two sets of three-phase voltage, one set for each sideband of the harmonics of the switching frequency. These three-phase voltages are rectified and pass through an LC filter. The RMS value of the transformer secondary voltage can be calculated by adding the square root of the RMS value of each harmonic voltage as follows: The RMS voltage of the harmonics has an inverse relation with the harmonic number (x); thus, 20 harmonics are considered to calculate the value of γ (M ) to limit the calculation error to less than 5%. A value of γ (M ) has been determined for various values of M, as shown in Fig. 7, by using MATLAB software. The output converter voltage can be determined by using the three-phase rectifier equations in [32], as follows: The value of transformer's turns ratio can be selected according to The value of V bus is selected according to (5), (6), and (7) and the value of γ (M ) can be found from Fig. 7.

2) SWITCH PARAMETERS
The converter switches should block the voltage of V bus when they are turned off; as a result, the peak voltage rating of the switches is V bus . The current ratings of the switches are calculated in [26]. It should be noted that since switches in the converter are shared by the AC-DC and DC-DC sub-converter stages, the current ratings of the switches are higher than those of conventional two-stage converters. The design procedure for an example 10kW WECS that is integrated into a 380VDC microgrid is as follows: The specifications of PMSG for a typical 10kW WECS have been taken from the model used in [33]. The parameters of the PMSG are listed in Table 1. The converter can be designed as follows: • The maximum intermediate DC voltage (V bus ) has been chosen to be 800V so that 1.2kV SiC devices can be used with a 50% safety margin.
• The maximum value of M can be determined from the AC-DC converter equations ( (5) and (6)) and is equal to 0.96, which results in γ (M ) = 0.382 according to the graph in Fig. 7.

(CHECK THE FIG. NUMBER)
• Using (11), the value of the effective duty cycle for the DC-DC stage D eff can be found to be 0.89.
• The transformer turns ratio n can then be found from (12) and is equal to 1.8 for this example.
• The RMS switch current can be determined from [26] and found to be 16A, which is higher than that of the switches of the AC-DC stage in a similar two-stage converter, which is 10A.

D. LOSS ANALYSIS
In this section, the converter power losses are discussed and compared with those of a conventional two-stage AC-DC converter. Since the passive components and transformers are the same in both the converters, only the power losses in the converter switches are investigated in this section. The power losses in the switches of the converter can be expressed as a summation of the switching power losses, which are essentially turn-on power losses in MOSFETs and conduction power losses [34], as follows: (13) where V on , is the voltage across the switch at turn-on time, and C oss is the output capacitance of the switch, and F sw is the switching frequency. I rms is the RMS value of the current of the switch as determined in [26], and R ds is the switch resistance. As shown in [26], with a proper design, V turn−on can be reduced to zero by using the reflected output current; thus, the converter switches work with zero voltage switching (ZVS) and switching power losses are negligible, which offsets the higher conduction losses of the proposed AC-DC converter. This feature lets the converter operate with a relatively high frequency without significant switching power losses. In conventional two-stage converters, the DC-DC converter switches are operated with ZVS and the switching frequency can be high without high power losses, but the AC-DC converter switches work without ZVS so that the switching power losses of these switches are considerable. Moreover, the input power is processed twice by a two-stage AC-DC converter. By simulating the converter as designed in the design example with C2M0025120D switches in PSIM software, the proposed converter was found to have an efficiency of 90% compared to an overall efficiency for a two-stage converter is 92%. It should be noted, however, that the proposed converter is significantly less expensive than a two-stage converter and smaller as well. If the converter's switching frequency is decreased, then its size can be increased to match that of a two-stage converter and it can have comparable efficiency at a lower cost.

III. WECS CONTROL WITH THE PROPOSED CONVERTER
In this section, the control system of a WECS implemented with the proposed converter is explained. The control system is developed based on the mathematical model of the WECS components, including the proposed isolated single-stage AC-DC converter, PMSG, and wind turbine.
The WECS unit control system is shown in Fig. 8. The control system is designed to meet the control objectives discussed in Section 1. In order to do so, an outer loop is used to perform the MPPT and power regulating control and an inner loop is used to perform PFC and actuate the outer loop command value. The operation of the outer and inner loops is explained in the following sections.

A. OUTER CONTROL LOOP
The outer loop consists of the PMSG speed reference generator and speed controller blocks that are shown in Fig. 8.  The objective of the outer control loop is to control the WECS unit in two different modes. First, when the input wind power is smaller than the wind turbine and the PMSG rated power, the outer loop controller performs MPPT control; this mode is referred to as Region 2 of the WT characteristics. Second, when the input wind power is larger than the wind turbine and the PMSG rated power, the outer loop controller limits the wind power absorbed by the wind turbine and the rated PMSG power (power regulating mode); this mode is referred to as Region 3 of the WT characteristics.
The control for each region can be implemented by using different methods [35]. MPPT in Region 2 can be implemented using various algorithms such as tip speed ratio (TSR) algorithm [36], [37], optimal torque algorithm [38], power signal feedback [39], perturb and observe algorithm [40], fuzzy-based [41], and neural network (NN)-based algorithms [42]. In this paper, the TSR algorithm is used to implement MPPT in Region 2. For Region 3 of operation, various methods such as pitch control [43], stall control [44], yaw [33], and furl [45] can be used. In the present study, stall control is used as the literature shows its advantages over other methods such as yaw and furl for small-scale wind turbines; however, any other control method that has been proposed in the literature can be implemented in Region 2 and Region 3 using the proposed converter. The control methods used in this paper for Region 2 and Region 3 will be briefly explained in this section.
The absorbed power absorbed by the WT can be expressed as: (14) where ρ is the air density, A is the WT swept area (π r 2 , r is the WT radius), C p is the power coefficient, and V wind is the wind speed. The value of C p is the critical factor in controlling the power absorbed by the WT.
When the wind speed is in Region 2, the value C p should be selected as the largest value that allows MPPT to be performed and in Region 3, the value of C p should be selected so that the power absorbed by the WT remains constant.
Given this context, the value of C p is a function of blade pitch angle β, and the ratio of WT angular speed to wind speed is called tip speed ratio. The relation between WT angular speed and wind speed (ω * g − V wind ) is determined from the design characteristic of WT, such as shown in Fig. 9(a).
In Region 3, the proposed converter forces the blades to operate in the stall region to limit the power absorbed. Stall is an aerodynamic phenomenon and it can happen if the angle of attack (the angle between the chord line and the relative wind speed) exceeds a critical value. When the blades operate in the stall condition, airflow is separated from the upper surface of the blades so that the lift forces on the blades are reduced and power captured is limited. As a result, stall control can be achieved if the blades are appropriately designed for stall regulation and the rotor speed is controlled based on the second part of the (ω * g − V wind ) curve at wind speeds higher than the rated wind speed,.
A look-up table (LUT) is used in the control system to determine the reference angular speed value for the turbine and generator (ω * g ) with respect to the wind speed (V wind ). Fig. 9(b) shows a graph of C p and P WT vs wind speed for a WT operating with the stated control method. It should be noted that a low pass filter (LPF) is used to remove small instantaneous fluctuations of wind speed and on the angular speed reference signal.
The angular reference speed of WT (ω * g ) is compared with the measured speed from the PMSG shaft. The error signal is fed to a PI controller (speed controller) to generate one of the reference values for the inner loop. The following state equations for the mechanical system can be used to design the speed controller: where T e is the mechanical torque of the PMSG, T f is the shaft static friction torque, F is the viscous friction coefficient, θ g is the PMSG shaft position, and J is the mechanical system momentum of inertia. The control variable for the speed control loop is T e . A limiter block is used to limit T e to the rated value and ensure the power absorbed by the WT does not exceed the converter ratings.

B. INNER CONTROL LOOP
The inner loop has PI controllers in the synchronous rotor reference frame (dq-frame) to shape the PMSG current based on (1) and (2), as shown in Fig. 8. The objective of this loop is to control the electrical torque of the PMSG (T e ) at a specific value that is specified by the outer loop (T * e ) while simultaneously performing power factor correction (PFC). According to these conditions, the setpoint values for the q and d axis currents can be calculated as follows: • Since the q-axis is aligned with the phase 'A' voltage induced by the PMSG, the relationship between I * qs and T * e can be expressed as • Since the current should be in phase with the voltage, the d-axis component of the current should be set to zero (I * ds = 0). As stated in Section II.A, a PI controller is used to implement the inner loop current controllers. The modulation signals in dq-frame can be determined by using (3) and (4), as follows:

IV. EXPERIMENTAL AND SIMULATION RESULTS
In this section, the feasibility of the proposed converter is confirmed by a scaled-down prototype and tested as a single unit. The performance of the complete WECS is confirmed by simulations of the aerodynamic, mechanical, and electrical systems in MATLAB/SIMULINK.

A. EXPERIMENTAL RESULTS
A scaled-down proof-of-concept prototype of the proposed converter was built to confirm the converter's feasibility. The following specifications were used to build the prototype: line to line RMS voltage, V LL = 120 V, line voltage angular frequency, ω e = 377 rad/sec, DC microgrid voltage V DC−MG = 48 V, converter rated power P conv = 1kW, and switching frequency F sw = 50 kHz. Switches S 1 -S 6 were implemented with IPx60R190P6 MOSFETs and DPG10I400PM devices were used as rectifying diodes (D 1 -D 6 ). The input inductors (L f ) were 1.2 mH, the input capacitors (C f ) were 1.2 µF, the blocking capacitors were C b = 680 nF, and the transformer turns ratio was 5:1. The values of L o and C o were 2 mH, and 1 mF, respectively. The converter was implemented with the same controller that was discussed in Section 3 and with double edge carrier conventional SPWM, using the slave DSP of a dSPACE 1103 (TM320F240). It was shown in Section 3 that the transient performance of the converter is the same as that of a typical three-phase, six-switch, non-isolated bidirectional AC-DC converter; thus, only steady-state results are presented here. Fig. 10 shows typical converter waveforms. Fig. 10(a) shows the three-phase input currents. It can be seen that the inner current controllers keep the PMSG current sinusoidal. Fig. 10(b) shows a typical phase (phase A) voltage and current. It can be seen that input voltage and current are in phase since I d is set to zero, as shown in Fig. 8. Fig. 10(c) shows the output current in the line frequency scale. It can be seen that it is a smooth DC waveform.

B. SIMULATION RESULTS
In this section, the performance of the overall system, including the WT, PMSG, proposed converter, and DC microgrid model is studied. Two studies are presented in this section. First, the dynamic performance of the proposed converter (Fig. 2) is compared to that of a conventional two-stage converter ( Fig. 1(b)), from the point-of-view of the wind turbine control system. Second, the proposed converter is integrated into a complete WECS, and its performance under different wind conditions is investigated. VOLUME 8, 2020

1) COMPARISON WITH THE CONVENTIONAL TWO-STAGE CONVERTER
In this study, both the proposed converter and conventional two-stage converter are designed with the specifications shown in Table 1. An input step command power (4 kW-10kW) is applied to both converters and the power absorbed is measured. The controllers are designed according to the procedure presented in [27]. The transfer function of the active power current component (I qs ), with respect to its reference value, can be expressed as The value of τ i is chosen to be 50 µsec. The coefficients of the PI current controllers in Fig. 5 are chosen to be k p = 24, and k i = 100. Simulation results for both converters are shown in Fig. 11. Fig. 11(a) shows the step response of the proposed converter and Fig. 11(b) shows the step response of the conventional two-stage converter. As can be seen, the dynamic response of both converters is the same, which confirms the discussion in Section II.1. This study demonstrates that any procedure that can be used to control a two-stage converter can be applied to the proposed converter. Step response for AC input power (a) the proposed converter (Fig. 2), (b) the conventional two-stage converter ( Fig. 1(b)).

2) OVERALL WECS UNIT PERFORMANCE WITH THE PROPOSED CONVERTER
In this study, the performance of the overall WECS operation with the proposed converter is examined. This study is performed by simulating the aerodynamic, mechanical, and electrical systems with a unified model. A typical WECS system, such as the system shown in Fig. 1(b), is modeled in the MATLAB/Simulink.
The key parameters of the electrical parts of the WECS were determined in the design example in Section 2. The switching frequency of the converter (F sw ) is chosen to be 25 kHz. A three-bladed stall-regulated WT that was designed in [46] by the authors to deliver 10kW rated power is used for this study. The WT is directly connected to the PMSG. The 10kW PMSG is modeled with the same parameters as [33]. The mechanical and electrical system parameters are given in Table 1. The electrical system is modeled with the power system toolbox of MATLAB/Simulink library components and other tools are used to model the mechanical part of the WECS system to assure the mechanical system modeling accuracy. The wind profile is generated by TurbSim [47] and is applied to a WT with ω * r − V wind characteristics that is shown in Fig. 9(a). Aerodynamic forces are calculated by AeroDyn [48] and mechanical forces are calculated by FAST (Fatigue, Aerodynamics, Structures, and Turbulence) software [49].
The aim of the study is to show that the overall WECS system can operate in different modes and that the proposed converter, with isolation, can follow the control system commands to achieve both the MPPT and the power regulation control schemes (shown in Fig. 9) in Regions 2 and 3, respectively. In order to do this, a typical turbulent wind profile with a turbulence intensity of class ''b'' and an average wind speed of 11 m/s has been generated by TurbSim software based on the IEC 61400-1 standard [50]. The wind profile is applied to the WT and simulation results are shown in Fig. 12. Fig 12(a) shows the wind profile that is generated by TurbSim software. The wind profile is selected to cover both Regions 2 and 3 of WT operation. The wind profile is applied to the WECS system as an input. Fig 12(b) shows the angular speed of the WT and PMSG. The reference speed is generated based on the characteristic of the WT that is shown in Fig. 9. It can be seen that when the wind speed is in region 3, the WT is reduced to limit the absorbed power by the WT. When the wind speed is in region 2, the WT speed varies proportionally with wind speed to perform MPPT.  Fig. 12(c) shows the output power of the PMSG. It can be seen that when the wind speed is in Region 2, the absorbed power by WT is less than the rated current of the PSMG and the WECS works in MPPT mode, however, when the wind speed increases to Region 3, the power absorbed by the WT is limited to the rated power of the WT and the PMSG (10 kW), and the WECS works in power regulating mode. Fig. 12(d) shows the electrical torque variations caused by the power absorbed by the WT. This value is used to calculate the current reference value for the q-axis component. Fig. 12(e) shows the reference signal I * qs that is calculated from T e by using (16) and the measured I qs value of the q-axis component of the stator current. It can be seen that measured I qs value follows the reference value that is set by the outer loop in both MPPT and power regulating regions. This means the proposed converter can perform MPPT and power regulating functions in the same way as a conventional VSC. Fig. 12(f) shows the reference I * ds and measured I_ ds values of the d-axis component of the stator current. It can be seen that I ds is always zero; thus, the converter performs power factor correction (PFC). Fig. 12(g) shows the harmonic content of the output current. It can be seen that the output current has a harmonic that is almost 3.3% of its DC component at a frequency that is 6 times the frequency of. It also has a harmonic that is about 1.4% of its DC component at a frequency that is 12 times the frequency of, but this is negligible.
It can be seen from the simulation results that the WECS with the proposed converter can perform all control and optimal operation requirements of a typical WECS.

V. CONCLUSION
This paper proposed a new configuration for interfacing wind energy conversion system (WECS) units to a DC microgrid (DC-MG) that uses a novel AC-DC converter. The new converter is a single-stage AC-DC converter designed for WECS units and its features are as follows: • The converter provides AC-DC conversion and galvanic isolation with only six active switches.
• The converter performance and control system is very similar to the conventional six switch AC-DC converter; thus, understanding its operation basics and its implementation with standard power electronic blocks is easy.
• The converter can follow MPPT controller command signals in the same way as any two-stage conventional converter; thus, the converter is compatible with most MPPT methods that have been proposed in the literature.
• The converter uses a three-phase six-switch DC-DC converter; thus, the component stress is less than that in single-phase four-switch DC-DC converters.
In this paper, the new configuration of the WECS unit with the new converter is explained. The basic steady-state operation of the new converter is discussed and a dynamic model for the new converter is developed. The feasibility of the converter was confirmed with experimental results from a scaled-down prototype and the overall performance of the WECS unit was studied with simulations that were based on a comprehensive model of the mechanical and electrical parts of the WECS system. VOLUME 8, 2020 The results for both the converter and the overall WECS confirmed that the proposed converter can be used in small-scale WECSs to reduce the cost and size of the small-scale WECSs that help power a DC microgrid.