SVG digital physical hybrid simulation experiment method under DFIG grid-connected scene
By connecting the doubly fed wind turbine grid-connected system with the SVG device through the RT-LAB platform and ITM algorithm, and performing segmented parallel computing, an SVG prototype was designed, which solved the problems of low simulation accuracy and poor flexibility in the existing technology and achieved efficient grid oscillation suppression.
Patent Information
- Application Number
- CN202411140187.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-20
- Publication Date
- 2026-03-03
AI Technical Summary
Existing technologies lack systematic SVG digital-physical hybrid simulation experiments in the grid-connected scenario of doubly-fed wind turbines, resulting in low simulation accuracy, poor flexibility and economy, and difficulty in effectively suppressing grid oscillations.
The RT-LAB real-time simulation platform is used to connect the digital wind power grid-connected system with the actual SVG equipment through the ITM algorithm. The simulation model is divided into multiple subsystems, and parallel computing and synchronous signal transmission are performed. The ARTEMIS module is combined to improve simulation efficiency, and an SVG prototype is designed to suppress oscillations.
It realizes power exchange between SVG and the real power grid under different operating conditions, shortens the hardware development and experimental cycle, improves the accuracy and flexibility of simulation, and effectively suppresses power grid oscillation.
Smart Images
Figure CN121596766A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of power system simulation research, and in particular to an SVG digital-physical hybrid simulation experimental method for a doubly fed induction generator (DFIG) wind turbine grid-connected scenario. Background Technology
[0002] Accelerating the development and efficient utilization of renewable clean energy (distributed energy) systems has become one of the key development strategies in my country's energy sector. The continuous increase in social electricity consumption and the ongoing investment in various renewable clean energy devices have placed higher demands on the power grid. These devices exhibit intermittency, volatility, and unpredictability during operation, making the power grid more prone to oscillations. Currently, a widely researched method for suppressing oscillations is to incorporate suppression controls into existing compensation equipment. Given the current rapid growth of renewable clean energy, while continuously increasing its installed capacity, researching oscillation suppression methods for additional equipment has become a new trend and direction for the future development and construction of smart grids. Related research aligns with the needs of national economic and social development and has significant practical implications.
[0003] Currently, the design of oscillation suppression methods based on additional devices of Static Var Generators (SVG) has been widely studied by experts at home and abroad. Among them, the establishment of a complete simulation analysis platform and the study of measures and control strategies to deal with typical operating conditions have important theoretical significance and engineering practical value for assisting the safe and stable operation of the power grid.
[0004] For current power systems, real-time simulation research on transient and steady-state conditions employs two main approaches: some involve building scaled physical experiments for real-time simulation, while others utilize software to construct simulation models and conduct research within real-time simulators. With the continuous development of real-time simulation technology, real-time simulation equipment is constantly emerging and being updated (RTDS, RT-LAB, ADPSS, etc.). Real-time digital simulators can run the constructed digital simulation models in real time. All-digital real-time simulation modeling is convenient and highly flexible, and can be used to achieve large-scale power system modeling and simulation. However, a drawback is that the components in a purely digital real-time simulation system are mathematical simulation models, which inevitably ignore certain system characteristics, reducing the accuracy of the simulation. While purely physical simulation can intuitively represent the actual state and physical meaning of the system, its scale is limited by space constraints, and different experiments require different experimental scenarios, resulting in poor experimental flexibility and economy, and increasingly restricting its application. Given the shortcomings and limitations of purely digital and purely physical simulations, hybrid simulation has attracted widespread attention from scholars. The reason why digital-physical hybrid simulation is achievable is that the digital and physical sides can establish interrelated mathematical models.
[0005] Hybrid digital-physical simulation, also known as power hardware-in-the-loop (PHIL) simulation, has seen significant development and research in numerous enterprises, institutions, and universities to date. However, current research on PHIL simulation methods, both domestically and internationally, primarily focuses on individual case studies, lacking simulation experiments involving additional devices in wind farms. Many interface devices also lack universality and systematic research. Against this backdrop, this invention researches a hybrid digital-physical simulation system for SVG (Static Var Generator). Using the ITM algorithm as the interface algorithm and an open converter system as the interface device, a wind farm grid-connected system with doubly-fed induction generators (DFIGs) is connected to the actual SVG equipment on the digital side, enabling power exchange between the SVG and the real power grid under different operating conditions. Summary of the Invention
[0006] The purpose of this invention is to provide a digital-physical hybrid simulation experimental method for SVG in a doubly-fed induction generator (DFIG) grid-connected scenario. Based on the RT-LAB real-time simulation platform, the method connects the digital-side wind power grid-connected system with the actual SVG equipment through a power interface, overcoming the time and cost of constructing various grid environments for testing and shortening the hardware development and experimentation cycle.
[0007] To achieve the above objectives, the present invention provides the following solution:
[0008] Step 1: Establish a grid-connected simulation model of DFIG and SVG based on RT-LAB.
[0009] The DFIG parallel SVG series-compensated grid-connected system is a complex power system with a large number of electrical components and various control signals. It is a multi-node, densely structured power system. For such a simulation model, solving involves a large number of matrix calculations, making it impossible for a single CPU to compute the results within a given simulation step size, resulting in low simulation efficiency. Existing simulation methods are insufficient to accurately simulate the internal system of interaction between the mechanical and electrical components of the DFIG, necessitating the design of node allocation before sending the data to an RT-LAB real-time simulator for computation.
[0010] To address this issue, the following method is adopted in the real-time simulation research based on RT-LAB:
[0011] 1) The RT-LAB real-time simulation platform consists of two parts: a host computer (human-computer interaction) and a slave computer (real-time simulation calculation). The host computer runs the Windows operating system and the RT-LAB software package to complete the simulation modeling process; the slave computer uses the OP5700 real-time simulator to compile the model and achieve real-time simulation. This functional block division rule divides the main loop into multiple parallel systems running on different CPUs, reducing the computational load and thus enabling real-time simulation.
[0012] When modeling on the RT-LAB real-time simulation platform, the requirements of main circuit partitioning and parallel processing and switching module averaging should be met. Therefore, the model needs to be divided into blocks. First, the primary system is divided, taking into account the function of each part and the number of switches. In this model, the doubly-fed induction generator (DFIG) electrical part (SM_DFIG_e) is divided into a computational subsystem. In addition, the main circuit also includes three separate computational subsystems: the DFIG mechanical part (SS_DFIG_m), the SVG subsystem (SS_SVG), and the AC system (SS_AC). A human-machine interface (SC_Console) is added as a console subsystem. The final model consists of 5 subsystems.
[0013] The transmission of signals between the split subsystems is accomplished using the OpComm module. The OpComm module is a simple filtering module that sends the input signal to a given subsystem, temporarily blocking all incoming signals and ensuring synchronous transmission of signals between the subsystems.
[0014] To prevent deadlock and serial computation and maximize parallel computation, a memory module is added between each subsystem to interrupt operations, ensuring that each subsystem completes signal transmission and reception operations within each step.
[0015] 2) After the system is divided into blocks, ARTEMIS modules are set at the output of the DFIG AC filter, the input and output of the AC line, respectively. The ARTEMIS solver uses a node method to discretize the set defined by the discrete SPS state-space equations, and can pre-calculate the matrices under different switching states, which can be directly called during simulation, thus improving the simulation efficiency of the model.
[0016] 3) The model adopts the state-space method (SSN), which can achieve accurate discretization and realize the calculation of power devices well.
[0017] After the modeling is completed, the model is converted into C language and imported into the RT-LAB lower-level machine. After it starts running, the real-time output status of the model can be observed through the analog output (AO) interface of the lower-level machine.
[0018] Step 2: Before building the overall wind power grid-connected system digital-physical hybrid simulation platform, a fully digital offline simulation verification is required. Combining with Step 1, the model is divided into two systems using the Ideal Transformer Model (ITM) algorithm. The specific operation steps are as follows:
[0019] Based on the DFIG and SVG grid-connected model, the voltage source ITM algorithm is used as the partitioning interface algorithm, with the SVG output as the common partition point, to divide the grid-connected system into two parts. According to the ITM interface algorithm principle, three controlled voltage sources are connected to the SVG voltage output, while three controlled current sources are connected to the grid side. The system operates as follows: First, the three-phase voltage signal at the PCC point is collected as the control signal for the controlled voltage source, and the signal is connected to the corresponding phase's controlled voltage source, making the controlled voltage source equivalent to the grid on the SVG side. Second, the three-phase current output from the SVG side is collected as the control signal for the controlled current source, and the signal is connected to the corresponding phase's controlled current source, making the controlled current source equivalent to an SVG device on the grid side. This method enables information exchange between the SVG and the grid.
[0020] After the model is established in this step, simulation verification is required to ensure that the output voltage and current waveforms of the SVG before and after segmentation are consistent, proving that the model segmentation has a certain degree of effectiveness.
[0021] Step 3: Complete the design of the SVG prototype. The prototype adopts a two-phase three-level full-bridge structure, and the hardware mainly includes five parts: DSP core board, signal conditioning board, power circuit driver baseboard and driver board, and main power circuit topology.
[0022] In the program design, the interrupt routine is the main program. Within one interrupt cycle, it performs operations such as querying and processing serial port data, reading and calculating sampled values, controlling the algorithm, updating the PWM duty cycle, and clearing the interrupt flag.
[0023] To achieve oscillation suppression, this invention performs a dq transformation on the introduced current. Considering that the fundamental current becomes a DC quantity after the dq transformation, while the harmonic components become AC quantities, passing the current through a low-pass filter at this point will filter out the harmonic components. The transfer function of the selected Butterworth second-order low-pass filter is:
[0024]
[0025] In the formula: w c This is the cutoff frequency of the low-pass filter. The above equation is then digitized based on the bilinear transformation, resulting in the bilinear transformation formula: T represents the operation cycle.
[0026] The final conclusion is:
[0027]
[0028] In the formula: Y(k) represents the output signal and X(k) represents the input signal.
[0029] The processed current signal is subtracted from the initial current signal, and the resulting harmonic current is added to the initial current loop, which can then control the prototype to generate the corresponding oscillation suppression current.
[0030] Step 4: Connect all components of the overall digital-physical hybrid simulation platform to complete the control closed loop. The specific steps are as follows:
[0031] 1) In the simulation, the control and circuit parts of the SVG model were removed, retaining only the controlled current source connected to the system. Based on the corresponding bin file, AO and AI interfaces were set in the model and exported through the DB37 female port on the back of the lower-level machine. Considering the limited output voltage of the RT-LAB, the voltage signal and output current signal at the SVG grid-connected PCC point were conditioned and reduced to -3V to +3V, and then led out through the AO port of the RT-LAB. The AI interface was configured with three analog inputs, which were amplified proportionally and then connected to the controlled current source in the original model.
[0032] 2) The weak voltage signal output from the RT-LAB is amplified 40 times by a four-quadrant power amplifier to reach the rated voltage of the prototype before being connected to the prototype. The current signal output from the RT-LAB is connected to the prototype signal conditioning board for sampling and processing.
[0033] 3) The current introduced from the RT-LAB section is processed in the prototype program to generate an oscillation suppression current at the three-phase outlet of the prototype.
[0034] 4) A three-phase current transformer is installed at the prototype's output to collect the prototype's output current signal and convert it into a voltage signal, which is then transmitted back to the RT-LAB lower-level computer via the analog input (AI) interface. Attached Figure Description
[0035] Figure 1 This is a flowchart of an SVG digital-physical hybrid simulation experimental method for a DFIG grid-connected scenario according to the present invention.
[0036] Figure 2 To segment the phase a current waveform before and after using the ITM algorithm
[0037] Figure 3 Schematic diagram of the overall closed-loop structure
[0038] Figure 4 Physical diagram of the entire system
[0039] Figure 5 When the system induces subsynchronous oscillation, the DFIG output current waveform before and after SVG is applied is shown in the diagram.
[0040] Figure 6 When the system induces supersynchronous oscillation, the DFIG output current waveforms before and after SVG is applied are shown in the diagram.
[0041] Figure 7 When the system induces medium-to-high frequency oscillations, the DFIG output current waveforms before and after SVG is applied are shown in the diagram. Detailed Implementation
[0042] To describe the present invention in more detail, the above-described digital-physical hybrid experimental method is verified below with reference to the accompanying drawings and specific implementation examples.
[0043] A DFIG parallel SVG grid-connected simulation model was built in MATLAB / Simulink simulation software. The model was divided into the doubly-fed induction generator (DFIG) electrical part (SM_DFIG_e), the DFIG mechanical part (SS_DFIG_m), the SVG subsystem (SS_SVG), the AC system (SS_AC), and the human-machine interface (SC_Console). OpComm and Memory modules were added to each subsystem, and the simulation step size was set to 10e-6.
[0044] Based on the established model, and according to the principle of the improved ITM interface algorithm, the SVG and the power grid are separated into two systems. At the separation point, a controlled source is used to connect them, allowing information exchange between the SVG and the power grid to be completed through the controlled source. After obtaining the simulation results, they are compared with the simulation results of an unsegmented wind farm grid-connected model. Figure 2 As shown.
[0045] Configure analog output (AO) and analog input (AI) interfaces, convert the model to C language via RT-LAB, and import it into the lower-level machine.
[0046] In the overall closed-loop system, special attention should be paid to the calculation of the conversion between digital and analog quantities during the multi-step signal transmission process. The signal conditioning board of the prototype mainly divides the acquisition and conditioning circuit into two categories: current acquisition and voltage acquisition. Current acquisition mainly uses a current Hall sensor. This sensor is a current output type, so the output of the current sensor must be connected to a sampling resistor to convert the current into voltage and send it to the ADC.
[0047] The prototype signal conditioning board has a current sensor with a turns ratio of 1000:1, meaning the primary side samples a current I of 25A, and the secondary side senses a current of 25mA. The default sampling resistor is 50 ohms. The DB25 socket on the signal conditioning board has a 2x in-phase amplifier for the analog output signal, therefore the current multiplier... Based on the adder circuit on the DSP core board, the formula for calculating AC current sampling should be:
[0048]
[0049] Simplified to:
[0050]
[0051] In the formula: I represents the sampling current, CH ADC This indicates the value of the ADC result register for the corresponding sampling channel.
[0052] Voltage acquisition uses differential mode. In practical applications, the scaling ratio is adjusted according to the amplitude range of the measured voltage to match the input range of the ADC.
[0053] The DC voltage sample, after conditioning and amplification via the DB25 socket, is by default reduced to a factor of Gdc = 1200 / 3.74 / 2 ≈ 160. Since the DC voltage does not use the addition formula of the DSP core board, the calculation formula is as follows:
[0054]
[0055] Simplified to:
[0056]
[0057] In the formula: U dc This represents the DC sampling voltage.
[0058] The AC voltage sample, after conditioning and amplification via the DB25 socket, is defaulted to a reduced value (G). ac =1200 / 4.99 / 2≈120 times; AC voltage uses an adder circuit, and the calculation formula is:
[0059]
[0060] Simplified to:
[0061]
[0062] In the formula: U ac This indicates the AC sampling voltage.
[0063] The selected current transformer has a transformation ratio of 50mV / A. Therefore, when amplifying the voltage source signal of the model's AI interface proportionally, the original transformation ratio needs to be multiplied by 200.
[0064] The overall closed-loop structure of the system is as follows Figure 3 As shown, after the connection is completed, the overall closed-loop physical system is as follows: Figure 4 As shown.
[0065] Subsynchronous oscillation, supersynchronous oscillation, and medium-to-high frequency oscillation were generated in the digital simulation model running on the lower-level machine of RT-LAB and then put into the physical side SVG device. The output current waveform of DFIG was observed, and the waveform was recorded using the OpWrite function built into RT-LAB.
[0066] The suppression effect on subsynchronous oscillations, low-frequency oscillations, and medium-to-high-frequency oscillations is as follows: Figure 5-7 As shown in the comparison, the use of SVG devices has a certain suppressive effect on oscillation, proving the effectiveness of SVG and the effectiveness of the digital-physical hybrid experimental method proposed in this invention.
[0067] This invention proposes a hybrid digital-physical simulation method for SVG in a DFIG grid-connected scenario. First, a DFIG parallel SVG grid-connected model is built based on RT-LAB, and the effectiveness of the ITM algorithm is verified in offline simulation. An SVG prototype with oscillation suppression function is designed and completed. The overall system achieves closed-loop operation through a power amplifier and voltage / current Hall effect sensors. Compared to full physical simulation, this invention overcomes the time and cost of constructing various power grid environments for testing, shortens the hardware development and experimental cycle, and better meets practical engineering needs.
[0068] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. However, the scope of protection of the present invention is not limited thereto. Any person skilled in the art can still modify or make equivalent substitutions to the specific embodiments of the present invention, and all such modifications and substitutions should be covered within the scope of protection of the claims of the present invention.
Claims
1. A method for SVG digital-physical hybrid simulation experiment in a DFIG grid-connected scenario, characterized in that, Includes the following steps: Step 1: Establish a grid-connected simulation model of DFIG and SVG based on RT-LAB; Step 2: Based on the ITM algorithm, perform offline all-digital simulation verification after model segmentation; Step 3: Complete the design of the actual SVG prototype, which includes five parts: DSP core board, signal conditioning board, power circuit driver baseboard and driver board, and main power circuit topology. Step 4: Connect all components of the overall digital physical hybrid simulation platform to complete the control closed loop.
2. The SVG digital-physical hybrid simulation experimental method in a DFIG grid-connected scenario according to claim 1, characterized in that, Step 1: To improve simulation efficiency, the simulation model is divided into blocks, namely, the doubly-fed induction generator (DFIG) electrical part (SM_DFIG_e), the doubly-fed induction generator (DFIG_m), the SVG subsystem (SS_SVG), the AC system (SS_AC), and the human-machine interface (SC_Console); ARTEMIS modules are set at the DFIG AC filter output, the AC line input and output; the state-space method (SSN) used can achieve accurate discretization and realize the calculation of power devices well.
3. The SVG digital-physical hybrid simulation experimental method in a DFIG grid-connected scenario according to claim 1, characterized in that, Step 2: Using the voltage source ITM algorithm as the segmentation interface algorithm, the grid-connected system is divided into two parts with the SVG output as the common segmentation point; according to the ITM interface algorithm principle, three controlled voltage sources are connected to the SVG voltage output, while three controlled current sources are connected to the grid side; the output voltage and current waveforms of the SVG before and after the segmentation are verified to be consistent.
4. The SVG digital-physical hybrid simulation experimental method in a DFIG grid-connected scenario according to claim 1, characterized in that, Step 3: Design and complete the actual SVG prototype; the hardware part mainly includes five parts: DSP core board, signal conditioning board, power circuit driver base board and driver small board, and main power circuit topology; in the program design, the interrupt program is the main program, which performs the following operations within one interrupt cycle: querying and processing serial port data, reading and calculating sampled values, controlling the algorithm, updating the PWM duty cycle and clearing the interrupt flag.
5. The SVG digital-physical hybrid simulation experimental method in a DFIG grid-connected scenario according to claim 1, characterized in that, Step 4: In the simulation, the control and circuit parts of the SVG model are removed, and only the controlled current source connected to the system is retained. Considering that the output voltage of RT-LAB has certain limits, the voltage signal and output current signal of the SVG grid-connected PCC point are conditioned and reduced to -3V to +3V. The weak voltage signal output by RT-LAB is amplified 40 times by a four-quadrant power amplifier to reach the rated voltage of the prototype and then connected to the prototype. The output current signal of RT-LAB is connected to the prototype signal conditioning board for sampling and processing. The current introduced from RT-LAB is processed in the prototype program to generate an oscillation suppression current at the three-phase output of the prototype. A three-phase current transformer is set at the prototype output to collect and convert the prototype output current signal into a voltage signal, and transmit it back to the RT-LAB lower computer through the analog input (AI) interface.