A lock-free wind power converter grid connection control method, device and medium

Abstract

The application provides a wind power converter grid-connected control method based on a Kalman filter without a phase-locked loop, comprising: a frequency-locked loop for tracking a grid voltage frequency in real time and feeding back to the Kalman filter; the Kalman filter sampling a fundamental voltage of the grid based on the grid voltage frequency; and the fundamental voltage is used for phase-locked loop-free control and vector current control. The application does not use Park transformation and a phase-locked loop, reduces the calculation burden, has superior dynamic tracking performance and stability, does not use a phase-locked loop system, effectively avoids the negative problems caused by the phase-locked loop under the conditions of a weak grid and an unbalanced grid, has better performance and robustness than a traditional vector current control strategy, uses the Kalman filter to sample the grid voltage, uses the frequency-locked loop for tracking the grid voltage frequency feedback as the input of the Kalman filtering method for frequency adaptive control, improves the grid waveform quality and frequency tracking capability of the three-phase grid-connected inverter, and improves the stable operation of the grid-connected system.

Description

Technical Field

[0001] This invention relates to the fields of power systems and power electronics technology, and specifically to a grid-connected control method, equipment, and medium for wind power converters without phase-locked loops. Background Technology

[0002] When a wind power converter is connected to the grid, the traditional method for extracting the grid voltage phase angle generally uses a phase-locked loop (PLL), which is a commonly used method for grid voltage phase angle extraction. Using a PLL can synchronize the grid-connected current with the grid. However, PLLs also bring some problems. For example, when connected to a weak grid, the phase information output by the PLL PI controller will exhibit phase angle disturbances under small signal disturbances, causing the system coordinate system and control coordinate system of the grid-connected inverter to not coincide. This, in turn, deteriorates the grid-connected inverter system through the current controller.

[0003] Meanwhile, the grid-connected current affects the voltage at the PCC point through the grid impedance. As the grid impedance increases, the coupling effect between the PLL, current controller, and grid impedance gradually strengthens, adversely affecting the stable operation of the system. Therefore, PLL-based synchronization strategies may be affected by stability issues, especially under very weak grid conditions, where the problems are more pronounced.

[0004] On the other hand, due to global warming and the depletion of fossil fuels, the development of wind and solar energy has become a global consensus. The performance of wind power converters directly affects the reliability of grid connection. To achieve reliable, safe, and efficient grid connection, synchronizing the grid connection current with the voltage at the point of common coupling is one of the most important issues in connecting to the grid.

[0005] To address the aforementioned issues, two main approaches are typically employed. The first approach utilizes a phase-locked loop (PLL). One solution introduces additional feedforward or feedback terms into the current control loop, such as output impedance reshaping methods and small-signal interference compensation control. However, these solutions increase the complexity of the current control loop and are dependent on the system's operating point. Another option is to modify the PLL based on the concept of a complex phase angle vector, eliminating the frequency coupling caused by traditional PLLs. Recently, an improved parameter tuning method has been proposed to mitigate the negative impact of PLLs in weak grid conditions. While these strategies alleviate PLL instability, it remains challenging to guarantee inverter stability when connected to a high-impedance, weak grid.

[0006] Furthermore, existing literature often ignores the dynamic nonlinearity of phase-locked loops (PLLs) during design and analysis, but this can affect the stability of the controller under weak power grid disturbances. In addition, the complex trigonometric function calculations and multiple coordinate rotation transformations in PLLs increase the computational burden of the control system, thus leading to slow transient response; moreover, the design of software PLLs requires complex structures such as high-precision microcontrollers, sine wave / square wave conversion circuits, and polarity control circuits, resulting in an overall complex structure.

[0007] The second category of solutions, in order to avoid the negative impact of phase-locked loops (PLLs), avoid using PLLs to directly extract the grid voltage phase angle. One approach proposes a PLL-free direct power control (DPC) strategy based on instantaneous active and reactive power theory, without using any inner-loop current regulator. However, this method has a variable switching frequency depending on the switching state, which leads to an unexpected wideband harmonic spectrum range, making line filters difficult to design. Another approach proposes a PLL-free grid-connected strategy using current regulators, but does not consider the impact of grid imbalance on inverter control and employs a bandwidth filter with complex parameter design, affecting the system's dynamic response speed. Other methods avoid using PLLs through power synchronization and achieve effective control performance under weak grid conditions. However, this control structure is incompatible with industry-standard vector control strategies, meaning it loses the inherent overcurrent protection function of standard current control schemes. Summary of the Invention

[0008] To address the shortcomings of existing technologies, the purpose of this invention is to provide a grid-connected control method, device, and medium for wind power converters without phase-locked loops.

[0009] According to one aspect of the present invention, a phase-locked loop-free grid-connected control method for a wind power converter based on a Kalman filter is provided, characterized in that it includes:

[0010] A frequency-locked loop is used to track the grid voltage frequency in real time, and the grid voltage frequency is fed back to the Kalman filter;

[0011] The Kalman filter samples the fundamental voltage of the power grid based on the power grid voltage frequency;

[0012] Vector current control is achieved by performing phase-locked loop-free control based on the fundamental voltage.

[0013] Preferably, the step of using a frequency-locked loop to track the grid voltage frequency in real time and feeding the grid voltage frequency back to the Kalman filter includes:

[0014] The amplification gain in the frequency-locked loop is γ, and the inputs to the amplification gain γ are the phase angle error signal e and the hysteresis voltage signal qv of 90°.

[0015] The phase angle error signal e is obtained by subtracting the estimated in-phase voltage component v from the input voltage signal v′, i.e., e = v′ - v;

[0016] The frequency error signal generated after calculation is related to the frequency setpoint (ω). f The sum of these two values ​​gives the grid frequency ω.

[0017] The frequency signal ω is fed back to the Kalman filter.

[0018] Preferably, the Kalman filter samples the fundamental voltage of the power grid based on the grid voltage frequency, comprising: an observer in the Kalman filter that measures in real time online, and obtains the current state estimate of the system using the current measurement value and the observed state quantity at the previous moment through a recursive least squares method.

[0019] Preferably, the process of obtaining the state estimate is divided into a prediction stage and a correction stage; wherein,

[0020] The prediction process includes:

[0021] Assume the system state-space equation is

[0022] X(k) = AX(k-1) + W(k)

[0023] Y(k)=CX(k)+V(k)

[0024] In the formula: X(k) is the state variable at time k, A is the state transition matrix, Y(k) is the measurement variable, C is the known constant matrix, W(k) is the interference noise inside the system, and V(k) is the noise present in the actual measurement, with covariance matrices Q and R respectively.

[0025] The current system state estimate and error covariance matrix are as follows:

[0026] X(k|k-1)=AX(k-1)

[0027] P(k|k-1)=AP(k-1)A T +Q(k)

[0028] Define the Kalman filter gain K(k) to minimize the root mean square error between the actual state and the measured state, and obtain:

[0029] K(k)=P(k|k-1)C T [R(k)+CP(k|k-1)C T ] -1 ;

[0030] The correction process includes: starting from calculating the Kalman filter gain K, system state estimation and system error covariance matrix update, where I is an identity diagonal matrix:

[0031] X(k)=X(k|k-1)+K(k)[Y(k)-CX(k|k-1)]

[0032] P(k)=[IK(k)C]P(k|k-1).

[0033] Preferably, the Kalman filter is used after the correction stage to decompose the voltage signal, and observes and outputs the fundamental wave and its corresponding orthogonal signal in the voltage signal;

[0034] The observation formula in the discrete time domain

[0035]

[0036]

[0037] Where Ts is the sampling frequency, V(k) is the measurement noise matrix, X(k) is the state variable at time k, Y(k) is the measurement variable, W(k) is the internal interference noise of the system, X1 and X2 are two variables in the state variable matrix, and W1 and W2 are variables in the internal noise matrix.

[0038] Preferably, the step of performing vector current control based on the fundamental voltage without a phase-locked loop includes:

[0039] A mathematical model of vector current for a three-phase LCL grid-connected inverter is obtained based on direct power without a phase-locked loop (PLL) and instantaneous active and reactive power.

[0040] The vector current mathematical model corresponds to the PLL in the dp coordinate system, and vector current control is derived from it.

[0041] Preferably, the vector current mathematical model of the three-phase LCL grid-connected inverter obtained by direct power calculation without a phase-locked loop based on instantaneous active and reactive power includes: deriving the LCL filter L1 by applying Kirchhoff's voltage law, including:

[0042] Transforming the three-phase a, b, and c coordinates to the stationary coordinate system is expressed as follows:

[0043]

[0044] In the formula: i 1α i 1β These are the inverter-side currents output along the α and β axes, respectively; u cα u cβThe voltages of the filter capacitors are divided into α and β axis; u α u β These are the output voltages of the α and β axis inverters, respectively; L1 is the inverter-side inductance of the LCL filter; R1 is the parasitic resistance of inductor L1.

[0045] The filter capacitor and the AC grid share a common neutral point n. According to instantaneous power theory, the instantaneous active and reactive power of the system after the inverter-side filter inductor L1 in the stationary reference frame are expressed as follows:

[0046]

[0047]

[0048] p and q are instantaneous active power and reactive power, respectively;

[0049] Taking the derivative, the dynamic equations for instantaneous active power and reactive power are:

[0050]

[0051]

[0052] The relation is obtained based on coordinate transformation theory.

[0053] u Cα =V c cos(ωt)

[0054] u Cβ =V c sin(ωt)

[0055] get

[0056]

[0057] In the formula V c It is the amplitude of the filter capacitor voltage, relative to the V c The dynamic equation for the capacitor voltage is obtained by differentiation.

[0058]

[0059]

[0060] The dynamic equations for active power and reactive power are obtained as follows:

[0061]

[0062]

[0063] The dynamic equations for active power and reactive power satisfy the following relationship:

[0064]

[0065] Obtain the coordinate transformation matrix from the stationary coordinate system to the synchronously rotating coordinate system, where u d u q This represents the inverter output voltage in a synchronous rotating coordinate system.

[0066] Use u d u q Express the inverter output voltage u in the stationary coordinate system. α u β Use it as the modulated input signal

[0067] u α =(u Cα u d +u Cβ u q ) / V c

[0068] u β =(u Cβ u d -u Cα u q ) / V c

[0069] The dynamic equations for active and reactive power are rewritten as follows:

[0070]

[0071]

[0072] In a synchronously rotating coordinate system, since the d-axis coincides with the instantaneous voltage vector and the q-axis is orthogonal to the instantaneous voltage vector, i.e., u cd =V c ,u cq =0, therefore the instantaneous active and reactive power in the dq coordinate system can be expressed as:

[0073]

[0074]

[0075] Multiply both sides of the dynamic equations for active and reactive power by 2 / (3V) c Then, the vector current mathematical model based on DPC is obtained:

[0076]

[0077] Preferably, the vector current mathematical model based on DPC corresponds to the PLL in the dp coordinate system, and vector current control is derived. u d ,u q It is the control law of the dq axis.

[0078] According to a second aspect of the present invention, an electronic device is provided, the electronic device comprising a processor and a memory, the memory storing at least one instruction, at least one program, a code set or an instruction set, the at least one instruction, the at least one program, the code set or instruction set being loaded and executed by the processor to implement any of the methods described herein.

[0079] According to a third aspect of the invention, a computer-readable storage medium stores at least one instruction, at least one program, a code set, or an instruction set, wherein the at least one instruction, the at least one program, the code set, or the instruction set is loaded by a processor and executes any one of the methods described herein.

[0080] Compared with the prior art, the present invention has the following beneficial effects:

[0081] When using traditional phase-locked loops (PLLs) to extract the grid voltage phase angle, in weak grids or under grid imbalance conditions, the PLLs themselves exhibit instability due to their dynamic nonlinear characteristics, which affects the stability of the inverter and the grid system. Furthermore, the complex trigonometric function calculations and multiple coordinate rotation transformations within the PLLs increase the computational burden on the control system. In contrast, the PLL-based wind power converter grid-connected control method provided in this embodiment of the invention eliminates the need for Parker transformations and PLLs, thus reducing the computational burden and offering superior dynamic tracking performance and stability.

[0082] In a weak grid with high impedance, the phase-locked loop (PLL) and current loop controller are coupled with the grid impedance, causing the output impedance phase angle of the grid-connected inverter to decrease and making the system unstable. When the grid voltage is unbalanced in three phases, the output phase angle of the PLL participates in coordinate transformation, resulting in the grid-connected current generated by the inverter containing a large number of harmonics, which reduces the current quality. The PLL-free grid-connected control method for wind power converters based on Kalman filters provided in this embodiment of the invention can effectively avoid the negative problems caused by PLLs under weak and unbalanced grid conditions because it does not use a PLL system. Therefore, it has better performance and robustness than the traditional vector current control strategy in these situations and solves the problems caused by PLLs.

[0083] Existing phase-locked loop (PLL)-free control strategies include direct power control based on instantaneous active and reactive power theory, which does not use any inner-loop current regulator. However, the vector current in this control strategy is obtained by converting instantaneous power, and instantaneous power oscillations inject harmonics into the current, affecting the power quality of the grid. Furthermore, the variable switching frequency of the switching states leads to unexpected wideband harmonic spectrum ranges, making line filters difficult to design. Other control strategies either fail to consider the impact of grid imbalance on inverter control and employ complex bandwidth filters with intricate parameter designs, affecting the system's dynamic response speed, or their control structures are incompatible with industry-standard vector control strategies, thus losing the inherent overcurrent protection function of standard current control schemes. The PLL-free grid-connected control method for wind power converters based on Kalman filters provided in this invention not only avoids these problems but also uses a linear Kalman filter to sample the grid voltage and uses a frequency-locked loop to track the grid voltage frequency feedback as the input to the Kalman filter algorithm for frequency adaptive control. This not only improves the grid-connected waveform quality and frequency tracking capability of the three-phase grid-connected inverter but also enhances the stable operation of the grid-connected system. Attached Figure Description

[0084] Other features, objects, and advantages of the present invention will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings:

[0085] Figure 1 This is a system overall control block diagram of a three-phase LCL grid-connected inverter based on a Kalman filter without a phase-locked loop, according to an embodiment of the present invention.

[0086] Figure 2 This is a schematic diagram of the FLL structure in the phase-locked loop control of a three-phase LCL grid-connected inverter based on a Kalman filter, according to an embodiment of the present invention.

[0087] Figure 3 This is a schematic diagram illustrating the extraction of the fundamental voltage using the KFM+FLL algorithm in the phase-locked loop control of a three-phase LCL grid-connected inverter based on a Kalman filter, as described in an embodiment of the present invention.

[0088] Figure 4 To achieve the output waveform of the Matlab / Simulink simulation experiment in Implementation Plan 1 without using a Kalman filter;

[0089] Figure 5 The output waveform of the Kalman filter is shown in the Matlab / Simulink simulation experiment of the phase-locked loop-free control of a three-phase grid-connected inverter based on Kalman filtering in Implementation Scheme 1.

[0090] Figure 6The output waveform of the Matlab / Simulink simulation experiment in the phase-locked loop control of a three-phase grid-connected inverter based on Kalman filtering in Scheme 1 is shown under the condition of filter parameter mismatch. Detailed Implementation

[0091] The present invention will now be described in detail with reference to specific embodiments. These embodiments will help those skilled in the art to further understand the present invention, but do not limit the invention in any way. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention. These all fall within the scope of protection of the present invention.

[0092] This invention provides an embodiment of a phase-locked loop-free grid-connected control method for a wind power converter based on a Kalman filter, comprising:

[0093] S100, the frequency-locked loop tracks the grid voltage frequency in real time and feeds the grid voltage frequency back to the Kalman filter;

[0094] S200, the Kalman filter samples the fundamental voltage of the power grid based on the grid voltage frequency in S100;

[0095] S300 performs phase-locked loop-free control based on the fundamental voltage obtained from S200, and performs vector current control.

[0096] The above embodiments provide a phase-locked loop-free control strategy based on KMF+FLL (the control strategy adopts Kalman filtering + frequency-locked loop). The frequency-locked loop tracks the grid voltage frequency in real time and feeds back the trackable grid voltage frequency to the input of the Kalman filtering method, thereby achieving the purpose of real-time and adaptive observation of the voltage signal.

[0097] The Kalman filter method is a linear Kalman filter that tracks the grid voltage to observe its fundamental component, used for sampling the grid voltage. This method uses the concept of state space to describe the system, where each update of the state variables depends only on the previous estimate and the new input data, without requiring all past data. This improves computational speed, avoids affecting the dynamic response speed of the system, and makes the estimates more effective and reliable.

[0098] In a preferred embodiment of the present invention, S100 is implemented. The amplification gain in the frequency-locked loop is γ, and the input to the amplification gain γ is the phase angle error signal e and the signal qv with a 90° hysteresis voltage. The phase angle error signal e is obtained by subtracting the estimated in-phase voltage component v from the input voltage signal v′, i.e., e = v′ - v. The frequency error signal generated after the calculation is compared with the frequency setpoint (ω). f The sums are used to obtain the grid frequency ω; the frequency signal ω is fed back to the Kalman filter.

[0099] In a preferred embodiment of the present invention, step S200 is implemented. Specifically, the observer of the Kalman filter, which performs real-time online measurement, obtains the current state estimate of the system using the current measurement value and the observed state quantity from the previous moment through a recursive least-squares method. When the Kalman filter is used for state estimation, the estimation process includes two steps: a prediction stage and a correction stage.

[0100] In another preferred embodiment of the present invention, the prediction and correction steps described above are further explained.

[0101] First, assume the system state-space equations are as follows:

[0102] X(k) = AX(k-1) + W(k)

[0103] Y(k)=CX(k)+V(k)

[0104] In the formula: X(k) is the state variable at time k, A is the state transition matrix, Y(k) is the measurement variable, C is the known constant matrix, W(k) is the interference noise inside the system, and V(k) is the noise present in the actual measurement, with covariance matrices Q and R respectively.

[0105] Prediction Phase: In the prediction part of the estimation process, the current system state estimate and error covariance matrix are as follows:

[0106] X(k|k-1)=AX(k-1)

[0107] P(k|k-1)=AP(k-1)A T +Q(k)

[0108] To predict the current system state, we need to define the Kalman filter gain K(k) to minimize the mean square error between the actual and measured states. This is obtained by:

[0109] K(k)=P(k|k-1)C T [R(k)+CP(k|k-1)C T ] -1

[0110] The correction stage of the Kalman filter described above uses a state estimation correction method to provide a more accurate estimate for the next time step. This process begins with calculating the Kalman filter gain K. The system state estimation and system error covariance matrix update are achieved through the following formula, where I is an identity diagonal matrix:

[0111] X(k)=X(k|k-1)+K(k)[Y(k)-CX(k|k-1)]

[0112] P(k)=[IK(k)C]P(k|k-1)

[0113] The Kalman filter is used here to decompose the voltage signal, outputting the fundamental wave and its corresponding orthogonal signals. For simplicity, the observation formula in the discrete-time domain is expressed as follows:

[0114]

[0115]

[0116] In a preferred embodiment of the present invention, an S300 Kalman filter and a frequency-locked loop are implemented in conjunction with direct power theory to derive vector current control (VCC), i.e., phase-locked loop-free control is achieved. This is a direct power control (DPC) strategy based on instantaneous active and reactive power theory. Based on DPC, a vector current mathematical model of a three-phase LCL grid-connected inverter is obtained. This mathematical model is the same current mathematical model as that of a traditional phase-locked loop (PLL) in the dq coordinate system, thereby deriving vector current control (VCC).

[0117] Specifically, based on the vector mathematical model of the DPC three-phase LCL grid-connected inverter, the mathematical model of L1 in the LCL filter is derived using Kirchhoff's voltage law. Transforming the three-phase a, b, and c coordinates to the stationary coordinate system, it can be expressed as:

[0118]

[0119] In the formula: i 1α i 1β These are the inverter-side currents output along the α and β axes, respectively; u cα u cβ The voltages of the filter capacitors are divided into α and β axis capacitors; u α u β These are the output voltages of the α-axis and β-axis inverters, respectively.

[0120] Considering that the filter capacitor and the AC grid share a common neutral point n, according to instantaneous power theory, the instantaneous active and reactive power of the system downstream of the inverter-side filter inductor L1 in the stationary reference frame can be expressed as:

[0121]

[0122]

[0123] p and q are the instantaneous active power and reactive power, respectively. To obtain the dynamic equation for power, we differentiate the above equation. The dynamic equations for instantaneous active power and reactive power are as follows:

[0124]

[0125]

[0126] According to coordinate transformation theory, the following relationship can be obtained:

[0127] u Cα =V c cos(ωt)

[0128] u Cβ =V c sin(ωt)

[0129] Where ω is the angle and t is the time;

[0130] Then we have:

[0131]

[0132] In the formula V c This is the amplitude of the filter capacitor voltage. Differentiating the above equation yields the following dynamic equation for the capacitor voltage:

[0133]

[0134]

[0135] Combining the above equations, the dynamic equations for active power and reactive power can be obtained as follows:

[0136]

[0137]

[0138] It can be noted that the above formulas satisfy the following relationship:

[0139]

[0140] The above formula constructs a coordinate transformation matrix from a stationary coordinate system to a synchronously rotating coordinate system, where u d u q This represents the inverter output voltage in a synchronous rotating coordinate system. It can be expressed as u. d u q Express the inverter output voltage u in the stationary coordinate system. α u β When used as a modulated input signal, it can be specifically represented as:

[0141] u α =(u Cα u d +u Cβ u q ) / V c

[0142] u β =(u Cβ u d -u Cα u q ) / V c

[0143] Combining the above formulas, the dynamic equations for active power and reactive power can be rewritten as follows:

[0144]

[0145]

[0146] In a synchronously rotating coordinate system, since the d-axis always coincides with the instantaneous voltage vector and the q-axis is orthogonal to the instantaneous voltage vector, i.e., u... cd =V c ,u cq =0, therefore the instantaneous active and reactive power in the dq coordinate system can be expressed as:

[0147]

[0148]

[0149] Multiply both sides of the above dynamic equations for active and reactive power by 2 / (3V) c Then, the vector current mathematical model based on DPC is obtained as follows:

[0150]

[0151]

[0152] Therefore, the control strategy without a phase-locked loop based on the Kalman filter method and frequency-locked loop in this embodiment starts from the direct power control model, combines the advantages of direct power control (DPC) and vector current control (VCC), and integrates the functions of the frequency-locked loop and Kalman filter method. Due to the Park transform and phase-locked loop (PLL), compared with the traditional vector current control, it not only achieves the same function as the phase-locked loop, but also reduces the computational load of the control method, avoids the slow dynamic performance of the phase-locked loop, and improves its dynamic performance.

[0153] In other embodiments of the invention, the frequency-locked loop can also be used to calculate the real-time voltage frequency of the power grid, providing frequency feedback for the Kalman filtering method, or other devices or method logic that can obtain real-time, adaptive observation voltage signals.

[0154] In other embodiments of the invention, the three-phase LCL grid-connected inverter may also be other converters, power electronic devices or topologies, or may be used in other scenarios involving obtaining voltage frequency.

[0155] In other embodiments of the invention, direct power control (DPC) derives a vector current mathematical model, or other methods may be used to obtain the vector current mathematical model.

[0156] In other embodiments of the invention, the Kalman filtering method can also be calculated using optimized methods or other similar methods.

[0157] In other embodiments of the invention, the overall method of Kalman filtering + frequency-locked loop can also be used for its optimization.

[0158] The following simulation examples will further illustrate the application of the above structure and method.

[0159] Figure 1 This is a block diagram of the overall system control of a three-phase LCL grid-connected inverter based on a Kalman filter without a phase-locked loop in this embodiment. Figure 2 This is a block diagram of the FLL structure in this invention. γ is the amplification gain, and its inputs are the phase angle error signal e and the signal qv with a 90° hysteresis voltage. The phase angle error signal e can be obtained by subtracting the estimated in-phase voltage component v from the input voltage signal v, i.e., e = v′ - v. The frequency error signal generated after the calculation is compared with the frequency setpoint (ω). f The summation yields the grid frequency ω. This frequency signal ω is further fed back to the Kalman filter to achieve real-time, adaptive observation of the voltage signal. Finally, the extraction process of the fundamental voltage using the Kalman filtering method is as described in this invention. Figure 3 As shown.

[0160] Based on the above embodiments, the system was simulated and verified using MATLAB / Simulink software. The simulation parameters are shown in Table 1. The system was also verified on the grid-connected side of a low-power wind power converter simulator using an LCL filter.

[0161] Table 1

[0162]

[0163] Implementation Plan 1:

[0164] A simulation model was built in Matlab / Simulink to verify the effectiveness of the theoretical analysis and control methods. This was done to compare and verify the effectiveness of using the Kalman filter method to sample voltage. Figure 4 and Figure 5 The simulation waveforms before and after using the Kalman filter method are shown respectively.

[0165] from Figure 4 As can be seen from this, this strategy can still maintain the grid-connected current i without using a phase-locked loop. g While the voltage is in phase and frequency with the PCC voltage, power oscillation issues cause the grid-connected current quality to fail to meet standards, resulting in a grid-connected current THD of 6.28%. Figure 5 It can be seen that after using the Kalman filter method, the power oscillations of the output active and reactive power are significantly reduced, the grid-connected current quality is significantly improved, and the current THD is reduced to 1.89%. This demonstrates that the measures adopted in this paper can greatly mitigate the broadband harmonics of the current caused by power oscillations, and also enable the inverter system to output stable active and reactive power.

[0166] To test the robustness of this control strategy under parameter uncertainty, it was determined that filter parameters may change due to aging in practical applications, where the inverter-side inductor plays a crucial role in the LCL filter's filtering function. Therefore, Figure 6 Simulation results were presented when the actual value of filter L1 did not match the reference value. When the actual value of L1 was 200% of the reference value, the grid-connected current THD was 1.43%. When it changed to 60% of the reference value, the grid-connected current THD was 2.72%, still meeting the grid-connected current quality requirements. The results show that the proposed control scheme does not lose steady-state performance when system parameters change, and has good robustness.

Claims

1. A phase-locked loop-free grid-connected control method for a wind power converter based on a Kalman filter, characterized in that, include: A frequency-locked loop is used to track the grid voltage frequency in real time, and the grid voltage frequency is fed back to the Kalman filter; The Kalman filter samples the fundamental voltage of the power grid based on the power grid voltage frequency; Vector current control is achieved by performing phase-locked loop-free control based on the fundamental voltage. in, The step of performing vector current control based on the fundamental voltage without a phase-locked loop includes: The vector current mathematical model of a three-phase LCL grid-connected inverter is obtained based on the direct power analysis without a phase-locked loop (PLL) for instantaneous active and reactive power. Specifically: ； Where L1 is the inverter-side inductance of the LCL filter, R1 is the parasitic resistance of inductor L1, ω is the grid frequency, and t is time. The vector current mathematical model corresponds to the PLL in the dp coordinate system, and the vector current control is derived as follows: u d ,u q These represent the inverter output voltages in a synchronous rotating coordinate system; The above control method does not include Parker transformation and phase-locked loop.

2. The method for grid-connected control of a wind power converter based on a Kalman filter without a phase-locked loop, as described in claim 1, is characterized in that... The method of using a frequency-locked loop to track the grid voltage frequency in real time and feeding the grid voltage frequency back to the Kalman filter includes: The frequency-locked loop has an amplification gain of γ, and the inputs to the amplification gain γ are the phase angle error signal e and the hysteresis voltage signal of 90°. qv; The phase angle error signal e is obtained from the input voltage signal. v′ Subtract the estimated in-phase voltage component from v To obtain, that is e = v′ v； The frequency error signal generated after calculation is different from the frequency setpoint ( ω f The power grid frequency is obtained by adding them together. ω ; The grid frequency ω is fed back to the Kalman filter.

3. The method for grid-connected control of a wind power converter based on a Kalman filter without a phase-locked loop, as described in claim 2, is characterized in that... The Kalman filter samples the fundamental voltage of the power grid based on the grid voltage frequency, and includes: an observer that measures in real time online in the Kalman filter, which uses the least squares method in a recursive form to obtain the current state estimate of the system using the current measurement value and the observed state quantity at the previous moment.

4. The method for grid-connected control of a wind power converter based on a Kalman filter according to claim 3, characterized in that, The process of obtaining state estimates is divided into a prediction stage and a correction stage; among them, The prediction process includes: Assume the system state-space equation is ； In the formula: X Y(k) is the state variable at time k, A is the state transition matrix, Y(k) is the measurement variable, C is the known constant matrix, W(k) is the interference noise inside the system, and V(k) is the noise present in the actual measurement. The covariance matrix of W(k) is Q, and the covariance matrix of V(k) is R. The current system state estimate and error covariance matrix are as follows: ； Define Kalman filter gain K ( k ), used to minimize the root mean square error between the actual state and the measured state, is obtained as follows: ； The correction process includes: starting from calculating the Kalman filter gain K, system state estimation and system error covariance matrix update, where I is an identity diagonal matrix: 。 5. The method for grid-connected control of a wind power converter based on a Kalman filter according to claim 4, characterized in that, The Kalman filter is used after the correction stage to decompose the voltage signal and observe and output the fundamental wave and its corresponding orthogonal signal in the voltage signal. Observation formula in discrete time domain ； ； Where Ts is the sampling frequency, V(k) is the measurement noise matrix, X(k) is the state variable at time k, Y(k) is the measurement variable, W(k) is the internal interference noise of the system, X1 and X2 are two variables in the state variable matrix, and W1 and W2 are two variables in the internal noise matrix.

6. The method for grid-connected control of a wind power converter based on a Kalman filter without a phase-locked loop, as described in claim 1, is characterized in that... The vector current mathematical model for a three-phase LCL grid-connected inverter, derived from the direct power calculation without a phase-locked loop based on instantaneous active and reactive power, includes: deriving the LCL filter L1 using Kirchhoff's voltage law, including: Three phases a, b, c The coordinate transformation to the stationary coordinate system is expressed as follows: ； In the formula: i 1α , i 1β These are the inverter-side currents output along the α and β axes, respectively. The voltages of the filter capacitors are divided into α-axis and β-axis; u α , u β These are the output voltages of the α and β axis inverters, respectively; L1 is the inverter-side inductance of the LCL filter; R1 is the parasitic resistance of inductor L1. The filter capacitor and the AC power grid share a common neutral point. n Based on instantaneous power theory, the inverter-side filter inductance in the stationary reference frame is obtained. L The instantaneous active and reactive power of the system after step 1 are expressed as follows: ； p and q These are instantaneous active power and reactive power, respectively. Taking the derivative, the dynamic equations for instantaneous active power and reactive power are: ； The relation is obtained based on coordinate transformation theory. ； get ； In the formula V c It is the amplitude of the filter capacitor voltage, for the aforementioned V c The dynamic equation for the capacitor voltage is obtained by differentiation. ； The dynamic equations for active power and reactive power are obtained as follows: ； The dynamic equations for active power and reactive power satisfy the following relationship: ； Obtain the coordinate transformation matrix from the stationary coordinate system to the synchronously rotating coordinate system, where u d , u q This represents the inverter output voltage in a synchronous rotating coordinate system. use u d , u q Indicate the inverter output voltage in the stationary coordinate system. u α , u β Use it as the modulated input signal ； The dynamic equations for active and reactive power are rewritten as follows: ； In a synchronously rotating coordinate system, since the d-axis coincides with the instantaneous voltage vector and the q-axis is orthogonal to the instantaneous voltage vector, that is... u Cd = V c ,u Cq = 0, therefore the instantaneous active and reactive power in the dq coordinate system can be expressed as: ； Multiply both sides of the dynamic equations for active and reactive power by 2 / (3) V c Then, the vector current mathematical model based on DPC is obtained: 。 7. An electronic device, characterized in that, The electronic device includes a processor and a memory, wherein the memory stores at least one instruction, at least one program, a code set, or an instruction set, and the at least one instruction, the at least one program, the code set, or the instruction set is loaded and executed by the processor to implement the method according to any one of claims 1-6.

8. A computer-readable storage medium, characterized in that, The storage medium stores at least one instruction, at least one program, code set, or instruction set, wherein the at least one instruction, the at least one program, the code set, or the instruction set is loaded by a processor and the processor executes the method described in any one of claims 1-6.

Images