Wind Turbine Electric Signals Simulator

Abstract

The development of green technologies in recent years in the field of wind energy conversion into electricity implies a technology transfer from the static switching field to the energy field. This paper presents a wind turbine simulator using a hardware solution following the energy conversion of a real turbine. We implemented this solution for educational and research purposes to train students in the process of electrical conversion in wind turbines. For the simulation, we chose an E82/2300 turbine, installed by ENERCON in a nearby geographical area. The turbine has the capacity to generate 2300 kW of electricity into grids. It has a direct coupling structure of the propeller to the generator. The solution is implemented on a multi-processor architecture with analog signal processing. The structure of a wind turbine is divided into three consecutive blocks, namely TUGEN, DCDC4X, and SIN3F. Each block of the simulator is designed with electronic components. The input and output signals of these blocks have similar waveforms to real signals, and their succession is interconditioned by process parameters. The innovation of the proposed solution is provided by software engineering applied to a hardware structure. The ratio between the simulated and real values is 1:60 in order to visualize the signals on a digital oscilloscope, mainly for educational purposes.

1. Introduction

The development of green technologies in recent years in the field of wind energy conversion into electricity implies a technology transfer from the static switching field to the energy field. New electrical energy conversion technologies involve optimal energy conversion solutions. These solutions are currently implemented in new wind turbines. This simulator is designed in partnership with ENERCON, a major wind turbine manufacturer in Europe.

We are witnessing continuous development in the study, development, and research of equipment and new technologies in the field of wind energy. Remarkable research has been carried out in the field of electrical energy conversion [1], in the field of optimizing the point of production of electricity from wind energy [2,3], and in the field of modeling and simulating processes in energy production facilities [4,5].

Also, this simulator was built in order to update the energy education system with new technologies used in electricity conversion equipment. The proposed system constitutes the simulation of a real wind turbine [6], similar to the ENERCON E82/2300 turbine [7].

The architecture of this simulator is modular. Each module is implemented based on the structure of the wind turbine. The hardware structure is carefully designed in order to produce identical signals to the measured signals from the wind turbine cabinets. These modules are independently implemented based on single or multi-processor structures. All these modules use ESP32-type microcontrollers [8].

The E82/2300 wind turbine is a modern turbine produced by ENERCON [7,9]. The turbine has three blades that reach a diameter of 82 m [9]. The maximum wind speed at which the turbine can operate is 25 m/s. The cut-in wind speed is about 2 m/s. The rated wind speed is 14 m/s [7].

The E82/2300 turbine has its propeller shaft directly coupled to its generator; it does not have an interposed gearbox [7,9]. The turbine is equipped with a 2300 kW synchronous generator. This is composed of two independent stator windings, each capable of generating half of the generator’s overall power. The turbine diagram is presented in Figure 1 [9].

For the wind turbine modeling, each constituent submodule was analyzed. Finally, three basic modules were developed [6]. The first module generates a three-phase signal of linearly variable amplitude and frequency, which is rectified. The second represents a voltage stabilizer having both a variable DC voltage at the input and a fixed DC voltage at the output. The last block is responsible for producing the three-phase sinusoidal voltage through a DC/AC inverter.

2. Wind Turbine Electric Simulator Design

The design of the wind turbine electric simulator is shown in Figure 2 [6,10]. Each described module is implemented as an electronic module. Signals generated by the simulator have reduced amplitude compared to electrical signals in the real environment.

Using this method, we obtain the following three modules for the simulator:

• TUGEN for turbine electrical generator signals and the three-phase controlled bridge rectifier. This module generates continuous voltage on output DC-LINK;
• DCDC4x—this module is powered by the DC-LINK bus and will output a fixed voltage to power the sinusoidal inverter;
• SIN3F—this module implements the three-phase sinusoidal inverter.

2.1. TUGEN Simulation Module Design

This module is responsible for the simulation of the synchronous generator and controlled bridge rectifier operation. For its implementation, a sub-modular schematic was used (see Figure 3) [11]. The module uses a base board that integrates three submodules (MODSIM1-TUGEN; items 1, 2, and 3 in Figure 3), a digital signal processor DSP1 used for data communication, and an analog processing block (item 12 and 13 in Figure 3). The simulation is controlled from the potentiometer n, which represents the wind speed (item 4). The wind speed is applied in the diagram as a reference signal to the simulator. All output signals are reported in terms of this n reference.

Each submodule (MODSIM1-TUGEN), with the schematic presented in Figure 4, generates signals for one phase of the three-phase system and is initialized with reference signal given by the wind speed. To generate signals that are 120° out of phase, cascade synchronization was implemented.

The basic schematic of MODSIM1-TUGEN was implemented using an ESP32 DSP core [8] and an analog signal processing diagram.

The module MODSIM1-TUGEN has f_wind analog voltage input, representing the wind speed level in the [0…100)% domain. The f_wind signal is applied on the DSP1 processor input. The signal is converted into a digital value using the internal 12 bit analog–digital converter (ADC1 item 1 in Figure 4), and then it is translated into the [0…1) domain, thus obtaining the value for N_fwind. The signal is filtered using item 2 and then amplified to obtain the generator voltage amplitude U_a by using item 3 and frequency f_a by using item 4, in accordance with relation 1 [6,11,12,13].

$$\left\{\begin{array}{c}{U}_a=k_A\times N_{\mathit{fwind}},k_A=1\\ f_a=k_f\times N_{\mathit{fwind}},k_f=100\end{array}\right.,$$

(1)

For the next step, continuous U_DC-LINK voltage is applied to a voltage stabilizer. The voltage stabilizer raises or lowers the U_DC-LINK to a fixed value so that voltage is applied to the three-phase DC/AC inverter.

2.2. DCDC4X Simulation Module Design

This module is responsible for simulating the voltage filtering, a four-time step-up converter [13,14], and a controlled variable resistor [15,16]. This module is implemented on a separate circuit board. The schematic has a power circuit and a control circuit, as represented in Figure 5.

The electronic switching circuit is composed of two MOSFET transistors working in antiphase and two inductors for storing switching energy. The voltage produced individually by transistor circuits is summed using the D1 and D2 diodes. The power circuit elements are controlled from a control circuit. The control circuit is based on the DSP3 processor [17] that uses an ESP32 processor and some additional digital integrated circuits [8].

2.3. SIN3F Simulation Module Design

This module is responsible for generating three-phase sinusoidal voltage waveforms [18,19,20,21]. This design uses three identical submodules (MODSIM1-SIN3F)—items 1, 2, and 3 from Figure 6. Each submodule is responsible for generating the signal for a single phase [19,20,22,23].

The DSP4 block is a based on an ESP32 microcontroller [8] for communication and simulation of the main switch; see items 12 and 13. Any submodule from Figure 6 has its detailed structure shown in Figure 7 [19,20,21,22,23].

The MODSIM1-SIN3F is a DSP (DSP5)-based submodule for generating one-phase sinusoidal inverter signals. DSP5 is based on an ESP32 microcontroller [8].

3. Experimental Results of the Simulator

Following the hardware implementation and programming of the simulator CPUs, the system’s response to input variables was measured. According to the designed structure (see Figure 2), the responses of the modules must be verified. The analysis process is based on measurements taken by using four channels of a GW-INSTEK GDS-1104B digital oscilloscope.

3.1. Simulated Generator Measurements

To measure the turbine generator characteristics, TUGEN, the oscilloscope’s four channels are connected to u_a (t), u_b (t), u_c (t), and n signals—see Figure 3. The results are presented in Figure 8.

Measured values are presented in

Table 1. TUGEN three-phase generator simulated signals reported according to wind speed.

Case	CH4 Wind Speed (n) [V]	Reported Wind Speed (n) [%]	CH1 Maximum Output Voltage Amplitude [V]	CH1 Output Voltage RMS [V]	CH1 Output Signal Period [ms]	CH1 Output Signal Frequency [Hz]
(a)	1.0	33.33	5.20	3.86	53.92	18.55
	1.5	50.00	6.58	5.49	34.66	28.85
	2.0	66.67	7.37	5.12	25.44	39.31
	2.5	83.33	7.79	5.74	20.09	49.78
(b)	3.0	100.0	8.02	5.63	19.89	50.28

below.

In Table 1, the reported wind speed n [%] is determined from the measured wind speed n [V] and reported according to the maximum value 3V (see relation 2).

$$n[\%]={\displaystyle \frac{n\left[V\right]}{3\left[V\right]}}\times 100\left[\%\right]$$

(2)

Also, the maximum output voltage (U_aM) and output signal period (T_a) are measured from the oscilloscope screen using the measuring tools option. In Table 1, output voltage RMS (U_a) and output signal frequency (f_a) are calculated according to relation (3).

$$U_a[V]={\displaystyle \frac{1}{T_a}}\times \sqrt{{\int }_0^{T_a}{u}_a^2\left(t\right)\times dt}[V]={\displaystyle \frac{U_{aM}}{\sqrt{2}}}[V],f_a={\displaystyle \frac{1}{T_a}}[\mathrm{Hz}]$$

(3)

By analyzing Table 1 data, the relation between RMS voltage and frequency is

$$U_G\left[V\right]=0.758{\displaystyle \frac{\left[V\right]}{\left[V\right]}}\times n\left[V\right]+3.652\left[V\right],f_a\cong 16.87{\displaystyle \frac{\left[\mathrm{Hz}\right]}{\left[V\right]}}\cdot n[V],$$

(4)

Based on relation 2, we can observe the linear dependence between the voltage at the generator output and the frequency of its signals in relation to the wind speed. The relationship demonstrates that the simulated signals at the TUGEN output are consistent with the operating equation of the synchronous electric machine.

Using the time measurement function of the oscilloscope, we measured the time difference between CH2-CH1 (t_d21) and CH3-CH1 (t_d31), respectively, at different time speeds (n). We also measured the period of the CH1 signal (T_a). The measured phase shift between generator phases is illustrated in

Table 2. TUGEN three-phase generator simulated phase delays.

Case	CH4 Wind Speed (n) [V]	CH2-CH1 Time Delay [ms]	CH3-CH1 Time Delay [ms]	CH1 Period [ms]	CH2-CH1 Phase Delay [rad]	CH3-CH1 Phase Delay [rad]
(a)	1.0	18.00	36.00	53.92	0.333828	0.667656
	1.5	11.50	23.00	34.66	0.331795	0.663589
	2.0	8.50	17.00	25.44	0.334119	0.668239
	2.5	6.70	13.50	20.09	0.333499	0.671976
(b)	3.0	6.60	13.30	19.89	0.331825	0.668678
					120°	240°

.

Phase shift delay between phases b-a (φ_d21) and phases c-a (φ_d31) are calculated based on relation 5.

$$\left\{\begin{array}{c}{\phi }_{d21}=2\times \pi \times {\displaystyle \frac{t_{d21}}{T_a}}\\ {\phi }_{d31}=2\times \pi \times {\displaystyle \frac{t_{d31}}{T_a}}\end{array}\right.$$

(5)

As we note from Table 2, the average phase shift between phases b and a is 120°, and the average phase shift between phases c and a is 240°. We express that the output-simulated generator voltages respect a three-phase direct voltage system.

3.2. Simulated Bridge Rectifier Measurements

An important correlation is imposed by the relation between the rectifier DC voltage output (U_DC) and the wind speed (n). In order to measure this, one, two, and four oscilloscope input channels are connected with u_a (t), u_DC (t), and n according to Figure 3. In Figure 9 are some results concerning the measurements of turbine speed, and

Table 3. TUGEN rectified output voltage according to wind speed.

Case	CH4 Wind Speed (n) [V]	CH2 Maximum Output Voltage Amplitude [V]	CH2 Rectified Output Voltage [V]	CH2 Output Signal Period [ms]	CH2 Output Signal Frequency [Hz]
(a)	1.00	4.32	4.14	10.63	94.04
	1.50	5.04	4.85	5.78	173.16
	2.00	5.80	5.62	4.23	236.31
	2.50	6.60	6.31	3.33	299.85
(b)	3.00	6.56	6.24	3.32	301.20

presents these measured values.

By using the option to measure the maximum value and the DC value of a signal, we used the CH2 channel in relation to the wind speed value (n), as well as the CH4 channel; see the values presented in Table 3 in the columns “Maximum Output Voltage Amplitude” and “Rectified Output Voltage”. Also, using the oscilloscope’s option to measure the time difference, we measured the period of the CH2 signal and noted it in Table 3 in the column “Output Signal Period” (T_rd).

The “Output Signal Frequency” (f_rd) column is obtained by applying relation 6 to the signal period (T_rd) in Table 3 above.

$$f_{rd}[\mathrm{Hz}]={\displaystyle \frac{1}{T_{rd}[s]}}$$

(6)

By analyzing rectified output voltage (U_DC) data from Table 3, a linear relation between the rectified voltage of the TUGEN simulated bridge rectifier and the wind speed can be obtained.

$$U_{\mathit{DC}}[V]=1.132{\displaystyle \frac{[V]}{[V]}}\times n[V]+3.168[V]$$

(7)

The bridge rectifier operation for one phase is presented in Figure 10. For this, the oscilloscope inputs are connected to the output generator voltage u_a (t), the current waveform passes through T1, the T1 gate pulses, and the wind level is equivalent to the speed of the generator, n.

3.3. Simulated Voltage Stabilizer Measurements

To measure DCDC4X module functionality, one, two, and four inputs from the oscilloscope are connected to U_DC-LINK signal, U_C-DC (see Figure 5), and wind speed signal (n). The capacitor effect can be observed by measuring U_C-DC voltage (see Figure 11 below).

In this case, we measured with the oscilloscope the DC voltage related to the wind speed at the terminals of the UC-DC capacitor bank. The measured data are presented in

Table 4. DCDC4X-filtered DC-LINK relation reported according to wind speed.

Case	CH4/Wind Speed (n) [V]	CH2/Filtered DC-LINK (UC-DC) [V]
(a)	1.00	4.34
	1.50	4.86
	2.00	5.41
	2.50	6.30
(b)	3.00	6.30

.

The relation between the filtered DC-LINK, U_DC-LINK, and wind speed is presented in the table above (Table 4).

The filtered DC-LINK voltage—U_C-DC—powers the DC-DC step-up convertor; see Figure 5. The step-up converter is composed by two units (L1-Q1-D1 and L2-Q2-D2), with each unit being activated alternately [8]. To analyze the DC-DC operation and measure the Q1 and Q2 gate signals, the oscilloscope’s first input is connected to the Q1 gate control signal (G1) and the third oscilloscope channel input is connected to the Q2 gate control signal (G2); see Figure 5. The wind speed is measured on oscilloscope channel 4. Measured signals are displayed in Figure 12a.

The applied pulses to the Q1 and Q2 transistors determine a switching process in DC-DC circuits. To monitor the switching effect, the Q1 gate control (G1) is connected to oscilloscope input 1, and the drain voltage of Q1, called u_z1 (t), is monitored on oscilloscope input 2. The same goes for transistor Q2—the gate signal is connected to oscilloscope input 3, and the drain signal is connected to input 4. The measured results are presented in Figure 12b.

To obtain the DC-DC step-up converter, the characteristic oscilloscope input channels are connected to the Q1 gate signal, u_G1 (t), output u_z (t) voltage—named DZ-LINK—Q1 drain signal, and the wind speed n. The control of the output voltage u_z (t) is monitored by the DSP3 processor—see Figure 5. The results are shown in Figure 13a. Overvoltage control is achieved by changing the converter output impedance [15,16]. Operation is monitored by connecting one, two, and four oscilloscope inputs to the Q3 gate control (G3), output voltage u_z (t), and wind speed level. The results are displayed in Figure 13b.

The gate control pulses have a 6 kHz switching frequency, and the PWM duty cycle can vary between 10% and 50% depending on the output voltage value. Analyzing the oscilloscope data, the DCDC4X input and output characteristic curves via turbine speed are presented in Figure 14a, and the output characteristic is shown in Figure 14b.

In Figure 14, we can observe the constant output values of the U_dc and U_z voltages in response to the DCDC4X block at values above 83.3% of the maximum wind speed value. Also, the U_z voltage remains constant when the U_c-dc voltage exceeds a value of 4.86 V.

3.4. Simulated Three-Phase Sinusoidal Inverter Measurements

To measure the functionality of the SIN3F inverter, the oscilloscope’s first probe is connected to u_GT1 (t), the second probe to u_GT4 (t), the third probe to the first phase output, u_R2 (t), and the fourth channel is connected to u_R1 (t); see Figure 6. The oscilloscope-measured data are shown in Figure 15.

According to Figure 15, each power transistor is controlled during half a period, producing positive or negative pulses [18,19]. Pulses have the same period (chopping frequency is 6 kHz), but the pulse width is modulated with a sinusoidal signal. The pulse level is imposed by the input voltage (U_Z).

By connecting oscilloscope channel 3 to u_R2 (t), channel 1 to u_R1 (t), channel 2 to u_R (t), and channel 4 to input U_z, we obtain one-phase signal processing from the SIN3F output to the load. Measured data is presented in Figure 16a.

In order to observe the three-phase functionality, the oscilloscope probes are connected to u_R (t), u_S (t), and u_T (t). The signals are shown in Figure 16b.

We remark that the generated signals have the same amplitude, same frequency (50 Hz), and a phase shift of 120° [18,19,21]. The output inverter values do not depend on wind speed, and the inverter input voltage is fixed by the DCDC4X module.

4. Conclusions

The described simulator reproduces all the electrical signals from a wind turbine at lower amplitude but retains the same shape. Therefore, it can be used both for educational and research purposes by simulating some operating limit situations. The simulator is structured on three main component blocks. including the generator and the controlled rectifier, the voltage conditioning block, and the inverter and network switching.

The system architecture is modular. The modules are electrically chained. Each module simulates a part of a wind turbine. The three-phase modules (TUGEN and SIN3F) are designed using three submodules (MODSIM1-TUGEN, respective MODSIM1-SIN3F) that operate in parallel. The submodules are based on ESP32 microcontrollers that are synchronized between them using a chained architecture. The ESP32 processors are used as DSP units, each one programmed according to the described signal processing algorithm presented.

The construction design of the simulator is achieved using common electronic components such as transistors, diodes, resistors, integrated circuits, and microcontrollers. The integrated circuits are used as comparators and simple mathematical functions such as addition, subtraction, constant multiplication. The generation and conditioning of sinusoidal or continuous waveforms, respectively the communication of signals between functional blocks is performed by ESP32 microcontrollers.

The innovative solution of this work consists of the implementation mechanism of the simulator and the signal processing mode. The solution approached by the authors consists of the creation of simulated equipment with a real interface at a reduced scale of the electrical part of a wind turbine. The simulator generates signals similar to the electrical signals generated by the equivalent blocks in the real turbine. The solution presented in the study generates signals only for operation under normal or limiting conditions.