A method for oscillation suppression of a permanent magnet direct drive wind power system based on a reactive current loop

Abstract

This application relates to the field of power system operation and control technology, and discloses a method for oscillation suppression in permanent magnet direct-drive wind power systems based on a reactive current loop. When the system is detected to be in a low-voltage ride-through state, this method extracts dynamic characteristic variables of the phase-locked loop (PLL) as input signals in real time. An additional oscillation suppression controller, configured with phase compensation, DC blocking filtering, and gain amplification stages, calculates and generates an additional reactive current command capable of suppressing oscillations. This command is then superimposed on the original reactive current setpoint and subjected to reactive current priority limiting processing before being input as the final command to the reactive power control loop of the grid-side converter. This invention accurately tunes phase compensation parameters based on a small-disturbance model and utilizes the coupling effect between reactive current and grid impedance to generate active damping. Without increasing hardware costs, it effectively suppresses system oscillations under weak grid conditions, enhancing the fault ride-through capability and operational stability of the generating units.

Description

Technical Field

[0001] This invention relates to the field of power system operation and control technology, specifically to a method for oscillation suppression in a permanent magnet direct-drive wind power system based on a reactive current loop. Background Technology

[0002] With the transformation of the global energy structure and the deepening of renewable energy strategies, the grid-connected installed capacity of permanent magnet direct-drive wind power systems continues to climb. Since wind energy-rich areas and load centers are often distributed in opposite directions, current wind power development mainly adopts a centralized construction and long-distance transmission model. This leads to a significant increase in the electrical distance between wind farms and the main grid, and consequently, an increase in the equivalent impedance of the grid connection lines, resulting in a distinctly weak grid characteristic at the wind farm's point of common coupling. Under these weak grid conditions, the grid voltage becomes more sensitive to power fluctuations, and the dynamic interaction between the converter control system of the permanent magnet direct-drive wind turbine and the AC power grid becomes increasingly complex. This easily triggers small-disturbance stability problems such as subsynchronous oscillations or low-frequency oscillations, seriously threatening the safe and stable operation of the power system.

[0003] Currently, there are two main technical approaches to oscillation suppression in permanent magnet direct-drive wind power systems: adding external equipment and optimizing the system's control. While adding flexible AC transmission devices such as static synchronous compensators (STATCOMs) or energy storage systems at the grid connection point can enhance system damping to some extent, this method faces economic challenges, including high equipment investment costs, large footprint, and difficult maintenance. Furthermore, the complex coordination and control between the external devices and the wind turbine controller often presents significant obstacles to its large-scale application in engineering projects.

[0004] In contrast, additional damping control strategies based on the wind turbine's own converter are more promising because they do not require additional main circuit hardware. However, existing control strategy optimizations often focus on active power control loops or speed control loops, emphasizing the regulation of active power output to suppress oscillations. This traditional approach has limitations in addressing stability issues under weak grid conditions. It often fails to fully analyze the strong coupling mechanism between reactive current and the phase of the point of common coupling voltage under high impedance environments in weak grids, and ignores the significant potential of the reactive current control loop in suppressing oscillation modes dominated by phase-locked loops. Especially under extreme conditions such as low-voltage ride-through, existing control strategies lack sufficient understanding of the dynamic coupling relationship between electrical subsystems, mechanical subsystems, and grid impedance, resulting in a lack of precise theoretical support for control parameter design and making it difficult to effectively utilize reactive power regulation channels to improve the overall damping characteristics of the system. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a method for oscillation suppression in permanent magnet direct-drive wind power systems based on reactive current loops. This method solves the problem of system oscillation instability caused by the dynamic coupling between the phase-locked loop and the grid impedance in permanent magnet direct-drive wind power systems under weak grid conditions and low voltage ride-through conditions.

[0006] To achieve the above objectives, this invention provides a method for suppressing oscillations in a permanent magnet direct-drive wind power system based on a reactive current loop, applicable to a control system including a grid-side converter and a phase-locked loop module. The main logic of this method lies in establishing a dynamic correlation between the reactive current and the phase of the point of common coupling voltage. By introducing a specific damping component into the reactive control loop, the positive feedback mechanism of the oscillation is disrupted.

[0007] Specifically, this method first performs operating condition monitoring and oscillation characteristic signal extraction. The system continuously monitors the three-phase voltage amplitude at the point of common coupling (PCC). Once the voltage amplitude is detected to be lower than a preset threshold, the system is determined to be in a low-voltage ride-through state. In this state, the system extracts state variables that characterize the dynamic instability of the phase-locked loop (PLL) in real time and establishes them as the oscillation characteristic input signals. This process ensures that the oscillation suppression strategy intervenes only during the most vulnerable fault recovery period of the system, avoiding impact on power generation efficiency under normal operating conditions.

[0008] Next, the method calculates and adjusts the additional damping control quantity. Its core lies in using an additional oscillation suppression controller to perform frequency domain shaping on the acquired oscillation characteristic input signal. This controller sequentially performs phase compensation processing, DC blocking filtering processing, and gain amplification processing on the signal. Through this series of signal processing steps, an additional reactive current command containing specific amplitude and phase information is generated. The physical meaning of this command is to generate a damping torque that is in opposite phase to the system's inherent oscillation mode.

[0009] Subsequently, the method performs reactive current command synthesis and limiting. The system obtains the original reactive current setpoint generated according to the low-voltage ride-through guidelines of the power grid, and algebraically superimposes it with the calculated additional reactive current command. The synthesized command after superposition integrates the steady-state support requirement for maintaining the voltage level and the dynamic damping requirement for suppressing phase oscillations. In order to protect the converter power devices, the system performs strict amplitude limiting on the synthesized command to ensure that the output current does not exceed the limit.

[0010] Finally, the method achieves control through current closed-loop tracking and execution. The amplitude-limited and corrected synthetic reactive current command is input to the reactive current control inner loop of the grid-side converter, and a drive pulse is generated using space vector pulse width modulation (SVM). The grid-side converter responds to this drive pulse, outputting a reactive current containing an additional damping component. When this reactive current flows through the weak grid impedance, it modulates the voltage phase at the point of common coupling (PCC), thereby generating negative feedback on the input signal of the PLL and suppressing PLL oscillations.

[0011] In a preferred embodiment, the selection of the oscillation characteristic input signal is flexible to adapt to different sampling conditions. It can utilize the difference between the synchronous angular velocity output by the phase-locked loop and the grid fundamental frequency (synchronous angular velocity increment), the difference between the phase output by the phase-locked loop and the reference phase (synchronous angle increment), or directly extract the high-frequency fluctuation component of the point of common coupling voltage. All these variables can effectively reflect the oscillation state of the system.

[0012] One of the key innovations of this invention lies in the parameter tuning mechanism for phase compensation. To ensure the effectiveness of the additional damping torque, this invention uses a small-disturbance linearization model of the system to accurately calculate the phase lag characteristics at the dominant oscillation frequency. This model covers the first path of interaction between the phase-locked loop's own integral action and the weak grid impedance, as well as the second path of voltage influence through reactive current coupling. The parameter tuning process utilizes the complex torque coefficient method to calculate the absolute value of the compensation angle required to maximize the total system damping. This absolute value of the compensation angle is designed to precisely offset the sum of the phase lag introduced by the reactive current inner-loop proportional-integral regulator and the phase lag introduced by the current-voltage coupling path. By configuring the lead time constant and lag time constant, the controller provides a precise phase lead angle at the dominant oscillation frequency, thereby achieving complete phase alignment.

[0013] Furthermore, this invention also relates to the optimized design of DC blocking filtering and gain amplification. The DC blocking filtering process employs high-pass filtering characteristics, with its cutoff frequency set much lower than the dominant oscillation frequency. This aims to completely block the DC bias without loss of the oscillation component, preventing any impact on the steady-state reactive power output accuracy of the system. The coefficients of the gain amplification process are determined using root locus analysis, selecting the optimal value that maximizes the damping ratio of the closed-loop poles of the system while maintaining stability, thus achieving a balance between damping effect and stability margin.

[0014] At the execution level, this invention employs a reactive power priority limiting strategy. When the synthesized command exceeds the converter's maximum allowable output current, the reactive current output demand is prioritized, and the maximum allowable output current is directly used as the final reactive current setpoint. Subsequently, based on the square relationship of current capacity, the remaining capacity is calculated using the Pythagorean theorem, thereby reducing the active current limit. This strategy ensures that during grid faults, regardless of the severity of oscillations, the system always prioritizes voltage support and oscillation suppression capabilities, maximizing the stability of the grid connection point.

[0015] This invention provides a method for oscillation suppression in permanent magnet direct-drive wind power systems based on a reactive current loop. It offers the following advantages:

[0016] 1. This invention taps into the potential regulation capability of the reactive current control loop of the grid-side converter and utilizes the strong coupling mechanism between the reactive current and the phase of the voltage at the point of common coupling under weak grid conditions. It achieves active stabilization without adding additional hardware devices such as STATCOM. This method transforms the impedance coupling characteristics that may lead to system instability under weak connection conditions into a control channel that applies active damping. It effectively solves the oscillation problem caused by the dynamic interaction between the phase-locked loop and the grid impedance during low voltage ride-through of permanent magnet direct-drive wind power systems, and improves the fault ride-through success rate and grid connection stability of the unit under extremely weak grid conditions.

[0017] 2. This invention designs an additional controller based on a small-disturbance linearization model of the system, which includes phase compensation, DC blocking filtering, and gain amplification. It can accurately reshape the phase for a specific dominant oscillation frequency. By quantitatively calculating and compensating for the phase lag introduced by the current loop proportional-integral regulator and the system coupling path, it ensures that the additional reactive current command can always generate positive damping torque at the oscillation frequency. This overcomes the problem of insufficient phase margin in complex operating conditions caused by the traditional empirical parameter tuning method, and achieves accurate suppression of subsynchronous or low-frequency oscillations without affecting the reactive power accuracy during steady-state operation of the system.

[0018] 3. This invention adopts a dynamic current limiting strategy that prioritizes reactive power. During faults where the converter capacity is limited, it prioritizes the output of reactive current including additional damping components and dynamically adjusts the limiting boundary of active current based on the remaining capacity. This mechanism ensures from the execution level that the oscillation suppression function will not be weakened or cut off due to converter current saturation under extreme conditions such as low voltage ride-through. This ensures that the system has sufficient control robustness and response capability when dynamic voltage support and phase stability are most needed. Attached Figure Description

[0019] Figure 1 This is a schematic diagram of the method flow of the present invention; Figure 2This is a schematic diagram of the parameter tuning process for the additional oscillation suppression controller of the present invention. Detailed Implementation

[0020] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0021] This invention provides a method for oscillation suppression in permanent magnet direct-drive wind power systems based on grid-side reactive current loops. This method is applied to permanent magnet direct-drive wind turbine generator sets that include grid-side converter control systems.

[0022] A permanent magnet direct-drive wind power system includes a permanent magnet synchronous generator, a generator-side converter, a DC bus, a grid-side converter, a filter circuit, and a grid-connected transformer. The stator terminals of the permanent magnet synchronous generator are connected to the AC input terminals of the generator-side converter. The DC output terminals of the generator-side converter are connected to the DC bus, on which a DC capacitor is connected in parallel. The DC bus is connected to the DC input terminals of the grid-side converter. The AC output terminals of the grid-side converter are connected to the low-voltage side of the grid-connected transformer via a filter circuit. The high-voltage side of the grid-connected transformer is connected to the grid's point of common coupling.

[0023] The grid-side converter is controlled by a grid-side controller. The grid-side controller is equipped with a signal acquisition module for real-time acquisition of the three-phase grid-connected voltage and current at the point of common coupling. The grid-side controller integrates a phase-locked loop (PLL) module. The PLL module is configured to receive the three-phase grid-connected voltage signal and, based on this signal, output the phase angle and synchronous angular velocity of a synchronous rotating coordinate system. The synchronous phase angle is used for coordinate transformation calculations in the control system, and the synchronous angular velocity is used as feedback for the system frequency.

[0024] The control logic of the grid-side controller is based on a synchronous rotating coordinate system (dq coordinate system). The grid-side controller includes a coordinate transformation module configured to use the synchronous phase angle output by the phase-locked loop to transform the collected three-phase grid-connected current and three-phase grid-connected voltage from the stationary coordinate system to the d-axis and q-axis components in the synchronous rotating coordinate system. The d-axis is oriented in the direction of the grid voltage vector.

[0025] The grid-side controller's control loop adopts a dual-closed-loop structure, including an outer voltage / power control loop and an inner current control loop. The inner current control loop is further divided into a d-axis active current control loop and a q-axis reactive current control loop. The d-axis active current control loop is configured to receive the difference between the active current setpoint and the actual d-axis current, and output a d-axis voltage command. The q-axis reactive current control loop is configured to receive the difference between the reactive current setpoint and the actual q-axis current, and output a q-axis voltage command. After being processed by the coordinate inverse transformation and space vector pulse width modulation module, the d-axis and q-axis voltage commands generate drive pulse signals to drive the power switching devices of the grid-side converter.

[0026] The power grid exhibits weak grid characteristics with certain impedance at the point of common coupling (PCC). The grid impedance is equivalent to a series circuit of a resistor and an inductor. When the grid short-circuit ratio decreases, the impact of the grid impedance on the PCC voltage increases. During system operation, when the PCC voltage amplitude is detected to be lower than the preset low-voltage ride-through threshold, the system determines that it has entered a low-voltage ride-through state.

[0027] In an embodiment of the invention, oscillation suppression logic is configured in the q-axis reactive current control loop of the grid-side converter. The grid-side controller is configured to acquire dynamic state variables output by the phase-locked loop (PLL) module during low-voltage ride-through. These dynamic state variables reflect the synchronous dynamic characteristics between the PLL and the grid voltage. Based on these dynamic state variables, the controller calculates an additional current command and superimposes it onto the input of the q-axis reactive current control loop. By adjusting the reactive current component output by the grid-side converter, the voltage phase characteristics at the point of common coupling are altered, thereby suppressing system oscillations.

[0028] See attached document Figure 1 This invention provides a method for oscillation suppression in a permanent magnet direct-drive wind power system based on a grid-side reactive current loop. This method is executed by a digital processor configured in the grid-side converter control system, and the specific process includes the following steps: Step S100: Operating Condition Monitoring and Oscillation Characteristic Signal Extraction. The grid-side controller continuously monitors the three-phase voltage amplitude at the point of common coupling (PCC). When the voltage amplitude at the PCC is detected to be lower than the preset low voltage ride-through threshold, the controller determines that the system has entered the low voltage ride-through state and activates the oscillation suppression function module.

[0029] During low-voltage ride-through, the controller extracts state variables reflecting the grid synchronization stability and PLL dynamic characteristics in real time from the phase-locked loop (PLL) module or voltage sampling module, using them as oscillation characteristic input signals. These oscillation characteristic input signals are specifically configured as one or a combination of the following physical quantities: Synchronous angular velocity increment of phase-locked loop output: Calculate the difference between the current synchronous angular velocity output by the phase-locked loop and the rated fundamental frequency angular velocity of the power grid.

[0030] The synchronization angle increment of the phase-locked loop output is obtained by calculating the difference between the current synchronization phase angle output by the phase-locked loop and the ideal reference phase angle, or by integrating the synchronization angular velocity increment.

[0031] Point of common coupling voltage fluctuation component: The high-frequency fluctuation component in the voltage amplitude of the point of common coupling is extracted. This component is obtained by band-pass filtering or high-pass filtering the real-time voltage amplitude.

[0032] Step S200: Calculation and Adjustment of Additional Damping Control Quantity. The oscillation characteristic input signal obtained in step S100 is input to the additional oscillation suppression controller. This controller is equipped with a phase compensation circuit, a DC blocking circuit, and a gain amplification circuit connected in sequence to perform frequency domain and time domain processing on the input signal. First, phase compensation processing: The input signal passes through a phase compensation stage. This stage is configured to lead or lag the phase of the input signal at a specific oscillation frequency. The goal of the correction is to compensate for the inherent phase lag of the system caused by current sampling delay, pulse width modulation delay, phase-locked loop response delay, and grid impedance coupling, ensuring that the output control signal forms a negative damping effect with the system oscillation mode.

[0033] Secondly, DC component filtering: The signal after phase compensation enters the DC blocking stage. The DC blocking stage is configured to block the DC component in the signal, allowing only the AC component reflecting oscillation characteristics to pass through. This process prevents additional control signals from causing a constant deviation in the system's steady-state reactive power output.

[0034] Finally, gain adjustment: the DC-filtered signal enters the gain amplification stage. This stage multiplies the signal amplitude by a preset gain coefficient and outputs an unlimited initial additional reactive current command. This gain coefficient determines the strength of the additional damping control effect.

[0035] Step S300: Reactive current command synthesis and limiting processing. The controller obtains the original reactive current setpoint of the system. This original reactive current setpoint is a reactive current support command obtained by looking up a table or calculating based on the voltage sag depth, according to the requirements of the low voltage ride-through procedure of the power grid.

[0036] The controller algebraically superimposes the initial additional reactive current command output in step S200 with the original reactive current setpoint to generate a composite reactive current command.

[0037] During the superposition process, the controller executes current amplitude limiting logic. The controller calculates the vector magnitude of the d-axis active current command and the synthesized reactive current command. If this vector magnitude exceeds the maximum allowable output current of the grid-side converter, the controller limits and truncates the synthesized reactive current command according to a preset priority strategy. The limited command serves as the final q-axis reactive current setpoint.

[0038] Step S400: Current Closed-Loop Tracking and Execution. The final q-axis reactive current setpoint is input to the q-axis current control loop of the grid-side converter. The q-axis current control loop calculates the deviation between the setpoint and the actual acquired q-axis reactive current feedback value, and generates a q-axis voltage control command through a proportional-integral regulator.

[0039] The q-axis voltage control command and the d-axis voltage control command are input together to the space vector pulse width modulation (SVPWM) module to generate switching pulses that drive the power semiconductor devices of the grid-side converter. In response to the switching pulses, the grid-side converter outputs a reactive current containing an additional damping component to the point of common coupling. This reactive current component suppresses frequency or phase oscillations of the phase-locked loop by changing the voltage phase dynamics at the point of common coupling.

[0040] In order to quantitatively tune the parameters of the additional oscillation suppression controller, it is first necessary to establish a mathematical model describing the dynamic characteristics of the system based on the weak power grid operating conditions.

[0041] This embodiment models the system in a synchronous rotating coordinate system (dq coordinate system). The steady-state operating point during low-voltage ride-through is taken as the linearized operating point. A small disturbance signal is applied to this operating point to establish a linear transfer relationship between the system's state variables. In this linearized model, there is a clear coupling relationship between the voltage disturbance at the PLL input and the angular frequency disturbance at the PLL output. This relationship is not only affected by the original current loop controller and grid impedance, but also directly by the additional oscillation suppression branch introduced in this invention.

[0042] Specifically, the q-axis voltage increment at the input of the phase-locked loop Synchronous angular velocity increment of phase-locked loop output The complex frequency domain transfer function relationship between them is constructed as follows: ; in, Represents the imaginary unit. Represents the angular frequency variable; The frequency domain expression for the q-axis voltage disturbance at the input of the phase-locked loop in a synchronous rotating coordinate system; This represents the synchronous angular velocity disturbance at the output of the phase-locked loop; This represents the transfer function of the proportional-integral (PI) regulator in the reactive current control loop of the grid-side converter. The transfer function representing the current to voltage; The transfer function represents the integral element, reflecting the integral relationship between the change in angular velocity and the change in angle. It represents the coupling transfer function of the active current component to the q-axis component of the grid-connected voltage, and is used to characterize the cross-coupling effect between voltage and current; This represents the transfer function of the additional oscillation suppression controller of the present invention.

[0043] In this mathematical model, the first term within the square brackets This describes the inherent dynamic characteristics of the system without additional control, specifically the voltage disturbance path generated by the interaction between the integral action of the phase-locked loop and the weak grid impedance. (The second term within square brackets...) The active damping path after introducing additional control is described, namely, by acquiring the angular velocity increment. via additional controller Adjusting reactive current, and thus through The closed-loop control effect affecting the grid connection point voltage. This linearized model reveals how the additional controller affects the overall stability of the system by changing the phase characteristics of the phase-locked loop input voltage, providing a theoretical basis for the accurate calculation of subsequent phase compensation parameters.

[0044] To ensure that the additional oscillation suppression controller can accurately suppress small disturbance oscillations at a specific frequency, it is necessary to quantitatively analyze the phase characteristics of the system at the oscillation frequency and calculate the required phase compensation accordingly.

[0045] This process first determines the dominant oscillation mode of the system based on the previously established small-disturbance linearization model. During the system design or adaptive adjustment phase, the transfer function of the additional controller is then... If the value is zero, perform eigenvalue analysis or frequency domain scanning on the remaining closed-loop control system. By identifying the corresponding conjugate complex roots in the pole distribution of the system's closed-loop transfer function where the real part is positive or the damping ratio is lowest, determine the dominant small-disturbance oscillation frequency of the system under low-voltage ride-through and weak power grid conditions, denoted as . .

[0046] Determine the oscillation frequency The system then further calculates the inherent phase lag generated by the original control link and the power grid coupling link at this frequency point. This phase lag mainly consists of two parts: The first part originates from the proportional-integral regulator within the reactive current loop of the grid-side converter. Based on its transfer function... Calculate its oscillation frequency The phase angle at point is denoted as .

[0047] The second part originates from the coupling path of active current to the q-axis component of the grid-connected voltage. This is derived from the active current-to-voltage coupling transfer function obtained through frequency domain analysis. Calculate its oscillation frequency The phase angle at point is denoted as This phase angle reflects the physical influence of the grid impedance characteristics and the system operating point on the voltage response phase.

[0048] In order for the damping torque generated by the additional controller to effectively suppress oscillations, the phase of the output signal of the additional control loop needs to match the inherent phase characteristics of the system to form a positive damping effect. Ideal damping control requires that the additional control signal and the speed deviation signal caused by oscillations maintain a specific phase relationship (usually by using phase compensation to make the phase of the total loop close to 0 degrees or 180 degrees, depending on the feedback polarity design, to generate a pure damping component).

[0049] Based on the principle of the complex torque coefficient method, this embodiment calculates the absolute value of the compensation angle required by the phase compensation stage in the additional oscillation suppression controller using the following formula.

[0050] ; The above equation defines a precise phase compensation strategy: the phase compensation provided by the additional controller should offset the phase angle introduced by the current loop Pl regulator. and the phase angle introduced by the system voltage coupling path. sum.

[0051] The above calculations yielded... The subsequent determination of the specific time constant (i.e., the lead time constant) for the phase compensation stage. and lag time constant This is the direct basis for [the calculation]. This calculation process ensures that the controller does not simply increase the gain, but rather targets a specific oscillation frequency. Precise phase alignment was performed to maximize the suppression of the target oscillation mode without affecting the performance of other frequency bands of the system.

[0052] The controller is configured in the computing unit of the grid-side converter control system and is used to convert the input oscillation characteristic signal into an additional reactive current command with specific phase and amplitude characteristics.

[0053] The transfer function of the additional oscillation suppression controller is denoted as: To achieve precise damping injection for specific frequency band oscillation modes, while isolating DC steady-state components and ensuring the stability of the control system, this controller adopts a series cascade structure. Its mathematical model consists of a phase compensation element. Isolation and straightness links and gain stage Together they constitute.

[0054] The complete transfer function expression for the additional oscillation suppression controller is as follows: ; in, Represents the Laplace operator. The time constant of the DC blocking element The specific technical details of each step are described below: Phase compensation stage : This stage is located at the front end of the signal processing chain (or may swap positions with other stages depending on the digital implementation order), and is configured to perform dynamic phase correction on the input signal. Its transfer function is designed as a lead-lag correction structure. ; in, This is the lead time constant of the phase compensation stage. Defined as the lag time constant of the phase compensation stage.

[0055] The values ​​of these two time constants are not chosen arbitrarily, but are strictly based on the phase compensation angle. Confirmed. (By setting...) and The ratio and value of these factors make this component dominate the oscillation frequency. The phase angle offset generated at that point is exactly equal to the required compensation angle. This process ensures that the signal from the controller, when superimposed on the reactive current loop, generates a pure damping torque opposite to the original oscillation mode of the system, rather than introducing new reactive power fluctuations or exacerbating instability. In the specific implementation of the digital control system, this stage can be discretized into difference equations using a bilinear transformation.

[0056] Isolation link : To prevent the DC component or low-frequency drift in the input signal from causing a persistent deviation in the system's steady-state reactive power output, a DC blocking element (or high-pass filter / Nashout element) is connected in series with the controller. Its transfer function expression is: ; in, This is the time constant of the DC blocking element.

[0057] The time constant The value of determines the low-frequency cutoff frequency of the controller. In this embodiment, It is configured to have a value between 3 and 10 seconds (e.g., 10 seconds). The selection of this parameter is based on the following: firstly, it needs to be large enough to ensure that the oscillation frequency... The amplitude gain of this stage is close to 1 and the phase shift is close to 0, so as not to interfere with the function of the phase compensation stage. On the other hand, it is necessary to effectively filter out the non-oscillating DC bias caused by measurement error or slow system variation, so as to ensure that the additional control only applies to the dynamic oscillation process.

[0058] Gain stage : This step is configured to linearly amplify or reduce the amplitude of the signal after phase correction and DC blocking, and its expression is represented by a scaling factor: ; in, The gain coefficient of the additional controller. is the gain constant.

[0059] This coefficient directly determines the strength of the additional damping control. The controller will pass through... and Multiply the processed intermediate variable The final additional reactive current control quantity is obtained. .

[0060] Through the synergistic effect of the above three stages, the controller can process the input signal reflecting the dynamic characteristics of the phase-locked loop (such as...). This is transformed into a reactive current correction command that can effectively suppress system oscillations during low-voltage ride-through under weak power grid conditions.

[0061] See attached document Figure 2 To ensure that the controller can perform its intended damping function in the actual physical system, this embodiment provides a rigorous set of parameter calculation and tuning procedures. These procedures, based on the aforementioned system transfer function model and the calculated phase compensation requirements, quantitatively design the three key parameter sets of the controller.

[0062] Parameter tuning specifically includes the following steps: Step 1: Time constant of DC blocking circuit Configuration Isolation link Its main function is to isolate the DC component, and its time constant This determines the cutoff frequency of the high-pass filter. Parameter configuration follows the principle of frequency band separation: the cutoff frequency must be much lower than the system's dominant oscillation frequency. This is to avoid the DC blocking element having a significant impact on the phase and amplitude at the oscillation frequency.

[0063] In this embodiment, the small disturbance oscillation angular frequency of the system during low voltage ride-through is first determined. ,in Subsequently, setting The value of is chosen such that it satisfies the following frequency domain constraints: ; Based on this constraint, this embodiment will The configuration is set to a fixed value, with a range of 3 to 10 seconds. In a specific preferred embodiment, the following is selected: Under these parameters, the amplitude gain of the DC blocking element at the oscillation frequency is approximately 1, and the phase shift is approximately 0, thus ensuring the independence of subsequent phase compensation design.

[0064] Step 2: Time constant of phase compensation stage and The calculation of this step aims to determine the lead time constant. and lag time constant This makes the phase compensation stage At the oscillation angular frequency The phase angle generated at that point is exactly equal to the absolute value of the angle to be compensated. .

[0065] The parameter calculation is based on the following two simultaneous equations: Center frequency constraint: The center frequency of the phase compensation stage is set to the oscillation angular frequency. To obtain the maximum phase lead angle at that frequency, the constraint relationship is expressed as: ; Phase angle constraint: at frequency At that point, the phase lead angle determined by the time constant should satisfy the target compensation value: ; Alternatively, an equivalent lead-lag ratio can be used in engineering. Perform the calculation: ; ; The processor, based on the above formula, utilizes known... and Numerical solution yields a unique and And update it in the control algorithm.

[0066] Step 3: Gain Coefficient Adjustment After confirming , and After that, the shape of the controller was determined, and only the amplitude gain remained. To be determined. This step uses root locus analysis to determine the optimal gain.

[0067] Constructing the open-loop transfer function: Substitute the designed phase compensation and DC blocking components into the open-loop transfer function model of the system.

[0068] Plotting the root locus: using the gain coefficient Plot the root locus of the closed-loop system as a variable (gradually increasing from 0 to positive infinity).

[0069] Determine the optimal value: Observe the movement trajectory of the conjugate complex poles representing the dominant oscillation mode in the complex plane.

[0070] along with As the damping ratio increases, the dominant pole gradually moves towards the left half of the complex plane (the direction of the negative real part), indicating that the system damping ratio increases.

[0071] when If the value is increased beyond a certain critical value, the pole may turn toward the imaginary axis or other high-frequency poles may become unstable.

[0072] The selection criteria are: The dominant pole with the largest damping ratio on the root locus (i.e., the pole with the smallest angle to the negative real axis) and all poles located in the stable region of the left half-plane. The value is used as the final controller gain.

[0073] Through the above three steps, all parameters of the additional oscillation suppression controller have been solidified. This parameter combination ensures that the controller can provide accurate phase compensation and appropriate damping amplitude under the target operating conditions, thereby effectively suppressing the dominant oscillation mode of the phase-locked loop in weak power grids.

[0074] This section explains how the grid-side controller converts the calculated control parameters into actual current commands and coordinates them with the system's existing control strategy. Within each digital control cycle of the grid-side converter, the controller executes the following logical steps to generate the final reactive current setpoint: The real-time calculation controller for the additional damping control quantity first reads the phase-locked loop synchronization angular velocity increment at the current sampling moment. Using the transfer function structure and setting parameters, the controller calculates the current additional reactive current command through discretized difference equations. .

[0075] The calculation process follows the following frequency domain control law formula: ; in, The transfer function of the additional oscillation suppression controller, This indicates that the expression is in the frequency domain and is used for dynamic analysis and design of control systems.

[0076] In its practical application in the time domain, this formula is represented as a weighted linear combination of the current input value, historical input values, current output value, and historical output values. The calculated... It is a dynamic AC signal containing amplitude and phase information, and its function is to generate a damping torque that is out of phase with the grid oscillation mode.

[0077] Obtaining the base reactive current setpoint: Simultaneously, the controller operates the Low Voltage Ride-Through (LVRT) support module in parallel. This module generates a base reactive current setpoint based on the real-time detected voltage sag depth at the point of common coupling, according to grid connection guidelines or a preset voltage droop characteristic curve. This basic given value mainly includes a DC component or a slowly varying component used to support the recovery of the grid connection point voltage.

[0078] Algebraic superposition of signals: The controller is equipped with an addition unit that calculates the dynamic additional reactive current command. Compared with the base reactive current setpoint Perform algebraic summation to synthesize the unlimited reactive current total command. .

[0079] The superposition logic satisfies the following formula: ; Through this superposition method, the system provides reactive voltage support as specified in the standard, while simultaneously controlling the reactive current. A small dynamic modulation component is introduced. This component does not change the steady-state reactive power output level of the system, but only suppresses the oscillation of the phase-locked loop during the dynamic process.

[0080] Priority limiting and final instruction output: To protect the power semiconductor devices of the grid-side converter, the controller imposes strict amplitude limits on the superimposed current command. Considering that maintaining voltage stability during low-voltage ride-through is the system's primary task, this embodiment employs a reactive power-priority limiting strategy.

[0081] The specific limiting logic is as follows: First, the controller reads the maximum allowable output current of the grid-side converter. .

[0082] Subsequently, the controller determines the total reactive current command that is not limited. Does the absolute value exceed .

[0083] like If so, the reactive current is cut off, and the final reactive current setpoint is... Values or (Depending on the original instruction symbol), the additional damping function is limited by the hardware capacity at this time; like Then the final reactive current setpoint Direct value .

[0084] After determining the final reactive current setpoint Then, the controller calculates the active current limit based on the remaining current capacity.

[0085] ; The original active current setpoint of the system must be limited to Within the range.

[0086] This limiting logic ensures that during grid faults, regardless of changes in the output of the additional damping controller, the grid-side converter always prioritizes the tracking capability of the reactive current loop (including foundation support and damping components), only reducing active current when capacity is insufficient, thereby maximizing the effectiveness of oscillation suppression. The final determined... The reactive current is fed into the inner loop PI controller to drive the converter.

[0087] This section describes alternative embodiments and extensions of the oscillation suppression method of the present invention in terms of input signal selection, controller implementation, and applicable system objects. Although the main embodiment described above preferably uses the phase-locked loop synchronization angular velocity increment as input, based on the control principle disclosed in this invention, the technical solution can be modified and extended as follows (not obvious).

[0088] First, regarding the diversified implementation of the oscillation characteristic input signal. In an alternative embodiment of the invention, the input signal of the additional oscillation suppression controller is not limited to the phase-locked loop synchronization angular velocity increment, but is configured as any state variable capable of characterizing the system's grid-connected point synchronization stability or voltage dynamics. Specifically, the input signal can be configured as the synchronization angle increment output by the phase-locked loop. When the synchronization angle increment is used as the input, the transfer function order of the system control path changes because the angle is the time integral of the angular velocity. Under this configuration, the internal structural parameters of the additional oscillation suppression controller need to be adjusted accordingly, specifically by adding a differential action or adjusting the lead time constant in the phase compensation stage to compensate for the 90-degree phase lag introduced by the integral stage, thereby ensuring that the final output reactive current command can still provide positive damping torque at the oscillation frequency. In addition, the input signal can also be configured as the voltage amplitude deviation at the point of common coupling. Under weak grid conditions, there is a strong coupling relationship between the grid-connected point voltage amplitude and the phase angle, and voltage fluctuations directly reflect the power oscillation state of the system. The controller collects the voltage amplitude deviation, extracts the oscillation component through bandpass filtering, and inputs it to the auxiliary controller. By adjusting the reactive current, the controller smooths out voltage fluctuations, thereby indirectly suppressing the phase oscillation of the phase-locked loop.

[0089] Secondly, regarding the generalized implementation of the controller's mathematical model. The transfer function described in the above embodiments is expressed in the form of a Laplace transform in the continuous domain. In actual digital control systems, this transfer function is discretized into difference equations for computation. This invention does not limit the specific discretization algorithm; the analog transfer function can be mapped to the discrete Z-domain using the bilinear transform method (Tustin transform), backward difference method, or zero-pole matching method. Regarding the controller's order, although the main embodiment demonstrates a second-order phase compensation structure, this invention covers linear controller structures of any order. As long as the controller satisfies the amplitude gain and phase compensation conditions at the dominant oscillation frequency determined through frequency response analysis, it falls within the scope of this invention. For example, the phase compensation stage can be composed of multiple first-order lead-lag stages connected in series, or a second-order generalized integrator (SOGI) structure can be used to achieve signal extraction and phase shifting at a specific frequency.

[0090] Finally, regarding the system-level expansion of applicable objects, the oscillation suppression method based on the grid-side reactive current loop of this invention is physically based on the phase-locked loop coupling mechanism of voltage source converters under weak grid connections. Therefore, the application of this method is not limited to the grid-side converters of permanent magnet direct-drive wind power generation systems, but is also applicable to the grid-side converters of doubly-fed induction generator (DFIG) wind turbines and the grid-connected inverters of photovoltaic power generation systems. In DFIG wind turbines, the grid-side converter is also responsible for DC bus voltage control and reactive power support, and its control architecture is highly similar to that of permanent magnet direct-drive systems. The additional damping control strategy of this invention can be directly transplanted to the grid-side controller of the DFIG. For photovoltaic inverters, when they are in a weak grid or at a low short-circuit ratio connection point, they also face the risk of phase-locked loop instability. The control logic provided by this invention, which modulates the voltage phase at the grid connection point through reactive current, is also applicable; it is only necessary to replace the control object from the grid-side converter of the wind turbine to the power stage of the photovoltaic inverter.

Claims

1. A method for oscillation suppression in a permanent magnet direct-drive wind power system based on a reactive current loop, characterized in that, Includes the following steps: Step S100, Operating Condition Monitoring and Oscillation Feature Signal Extraction: Continuously monitor the three-phase voltage amplitude at the point of common coupling. When the three-phase voltage amplitude is detected to be lower than the preset low voltage ride-through threshold, the system is determined to have entered the low voltage ride-through state. In the low voltage ride-through state, state variables reflecting the dynamic characteristics of the phase-locked loop are extracted in real time, and the state variables are used as oscillation feature input signals. Step S200, Calculation and adjustment of additional damping control quantity: In response to the oscillation characteristic input signal extracted in step S100, phase compensation processing, DC blocking filtering processing and gain amplification processing are sequentially performed on the oscillation characteristic input signal to generate an additional reactive current command that can suppress oscillation; Step S300, Synthesis and Limiting of Reactive Current Command: Obtain the original reactive current given value generated based on the low voltage ride-through procedure of the power grid, calculate the synthesized reactive current command by performing an algebraic superposition operation on the additional reactive current command generated in step S200 and the original reactive current given value, and perform amplitude limiting processing on the synthesized reactive current command. Step S400, Current Closed-Loop Tracking and Execution: The synthesized reactive current command processed in step S300 is used as the final q-axis reactive current setpoint and input to the q-axis reactive current control loop of the grid-side converter. A drive signal is generated through space vector pulse width modulation to drive the grid-side converter to output reactive current containing additional damping components, thereby suppressing the oscillation of the phase-locked loop.

2. The oscillation suppression method for a permanent magnet direct-drive wind power system based on a reactive current loop according to claim 1, characterized in that, In step S100, the real-time extraction of state variables reflecting the dynamic characteristics of the phase-locked loop as oscillation characteristic input signals specifically includes any one of the following methods: Method 1: Obtain the synchronous angular velocity output by the phase-locked loop module, calculate the synchronous angular velocity increment by performing a difference operation between the synchronous angular velocity and the rated base frequency angular velocity of the power grid, and use the synchronous angular velocity increment as the oscillation characteristic input signal; Method 2: Obtain the synchronization phase angle output by the phase-locked loop module, calculate the synchronization angle increment by performing a difference operation between the synchronization phase angle and the reference phase angle, and use the synchronization angle increment as the oscillation characteristic input signal; Method 3: Collect the three-phase voltage of the common connection point, calculate the high-frequency fluctuation component by performing amplitude synthesis and filtering operations on the three-phase voltage, and use the high-frequency fluctuation component as the oscillation characteristic input signal.

3. The oscillation suppression method for a permanent magnet direct-drive wind power system based on a reactive current loop according to claim 1, characterized in that, In step S200, the specific steps of sequentially performing phase compensation processing, DC blocking filtering processing, and gain amplification processing on the oscillation characteristic input signal include: Phase compensation processing is performed: using a transfer function with lead and lag characteristics, the phase of the oscillating characteristic input signal at the oscillation frequency is corrected by lead or lag to compensate for the inherent phase lag of the system. Perform DC blocking filtering: Using a transfer function with high-pass filtering characteristics, the DC component in the signal is blocked, allowing only the AC component reflecting oscillation characteristics to pass through; Perform gain amplification processing: Using a proportional gain coefficient, the amplitude of the signal after the above processing is linearly amplified or reduced, and the additional reactive current command is output.

4. The oscillation suppression method for a permanent magnet direct-drive wind power system based on a reactive current loop according to claim 3, characterized in that, The parameters used for performing the phase compensation process are predetermined based on the system's small-disturbance linearization model. The determination process includes: Based on the small disturbance linearization model, the dominant small disturbance oscillation frequency of the system under low voltage ride-through and weak power grid conditions is identified. The first phase lag angle introduced by the proportional-integral regulator is obtained by substituting the dominant small disturbance oscillation frequency into the transfer function model of the reactive current control loop of the grid-side converter. The second phase lag angle is obtained by substituting the dominant small disturbance oscillation frequency into the coupling path model of the active current on the q-axis component of the voltage at the point of common coupling. The required absolute value of the compensation angle is calculated using the complex torque coefficient method, such that the value of the absolute value of the compensation angle is the sum of the first phase lag angle and the second phase lag angle, and the parameters for phase compensation processing are set accordingly.

5. The oscillation suppression method for a permanent magnet direct-drive wind power system based on a reactive current loop according to claim 4, characterized in that, The small-disturbance linearization model is a mathematical model used to describe the complex frequency domain transfer relationship between the q-axis voltage increment at the input of the phase-locked loop and the synchronous angular velocity increment at the output of the phase-locked loop; the complex frequency domain transfer relationship includes two parallel signal transmission paths: First path: Describes the inherent voltage disturbance propagation relationship generated by the interaction between the integral action of the phase-locked loop itself and the impedance of the weak power grid; The second path describes the active damping transmission relationship that regulates reactive current through additional damping control, and then affects the voltage at the point of common coupling through the coupling effect of current and voltage.

6. The oscillation suppression method for a permanent magnet direct-drive wind power system based on a reactive current loop according to claim 4, characterized in that, The parameters of the phase compensation process include a lead time constant and a lag time constant, and the values ​​of the lead time constant and the lag time constant are selected to satisfy the following two conditions: Condition 1: The value of the center frequency determined by the lead time constant and the lag time constant is the dominant small disturbance oscillation frequency; Condition 2: The phase lead angle determined by the lead time constant and the lag time constant corresponds precisely to the absolute value of the compensation angle at the dominant small disturbance oscillation frequency.

7. The oscillation suppression method for a permanent magnet direct-drive wind power system based on a reactive current loop according to claim 3, characterized in that, The parameter settings for performing DC blocking filtering and gain amplification processes follow the following principles: The time constant of the DC blocking filter is set so that the cutoff frequency of the processing stage is much lower than the dominant small disturbance oscillation frequency, ensuring that no amplitude attenuation and phase shift occur at the oscillation frequency. The proportional gain coefficient of the gain amplification process is determined by the root locus analysis method. It is selected such that the damping ratio of the closed-loop pole corresponding to the dominant small disturbance oscillation frequency is maximized in the complex plane, and all system poles are located in the stable region.

8. The oscillation suppression method for a permanent magnet direct-drive wind power system based on a reactive current loop according to claim 1, characterized in that, In step S300, the synthesis and limiting of the reactive current command specifically includes: The unlimited total command is obtained by performing an addition operation between the additional reactive current command and the original reactive current setpoint. Determine whether the absolute value of the unlimited total command exceeds the maximum allowable output current of the grid-side converter; If the judgment result is exceeded, the maximum allowable output current is used as the final q-axis reactive current setpoint, and the active current limit is reduced according to the capacity occupied by the setpoint to ensure the dynamic response of the reactive control loop.

9. The oscillation suppression method for a permanent magnet direct-drive wind power system based on a reactive current loop according to claim 8, characterized in that, The limiting value for reducing active current based on the given value specifically includes: The maximum allowable output current of the grid-side converter is calculated by performing a power operation on the maximum allowable output current to obtain the square value of the current. The square value of the reactive current is calculated by performing a power operation on the final q-axis reactive current setpoint. The square difference is calculated by subtracting the square value of the maximum current from the square value of the reactive current. The amplitude limit of the d-axis active current of the grid-side converter is obtained by performing an arithmetic square root operation on the squared difference.

10. The oscillation suppression method for a permanent magnet direct-drive wind power system based on a reactive current loop according to claim 3, characterized in that, The process of generating the additional reactive current command in step S200 is mathematically represented by multiplying the oscillation characteristic input signal by three transfer function terms in sequence. The three transfer function terms are: Phase compensation term: It consists of the ratio of a differential term containing a lead time constant to an inertial term containing a lag time constant; DC blocking term: It consists of the ratio of the differential term containing the DC blocking time constant to the inertial term containing the DC blocking time constant; Gain term: It is a proportionality coefficient in constant form.

Images