A method for acquiring a phase-locked signal of a DSOGI-LADRC-PLL for grid-connected switching

Abstract

A method for obtaining a DSOGI-LADRC-PLL phase-locked signal for grid-connected switching of a power grid, aiming to improve the phase-locked accuracy in grid-connected switching and suppress interference, and ensure efficient and stable grid-connected operation, comprising: step one: performing Clarke transformation on three-phase voltage signals through a DSOGI module to convert the three-phase signals into two-phase orthogonal signals, and performing DSOGI filtering on the obtained orthogonal components to obtain instantaneous phase and amplitude, and performing phase error calculation and range adjustment. Step two: using high-order dynamic prediction control (HPDC) technology to model and predict the phase error. Step three: defining a gain adjustment function, and combining a fuzzy control strategy to optimize the gain of the LADRC controller in real time to adapt to different loads and disturbance conditions. Step four: using an improved LADRC controller to process the phase error and output an accurate phase-locked signal. This method is suitable for rural grid-connected switching and can provide high-precision phase synchronization and fast disturbance response to ensure stable operation of the power grid system.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of power system control and automation control, and in particular to an improved grid-connected switching DSOGI-LADRC-PLL adaptive control technology applicable to microgrid, distributed power generation, rural power grid and the like. BACKGROUND

[0002] With the rapid development of smart grid and distributed power generation, the stability and efficiency of power systems face unprecedented challenges. Grid frequency fluctuations, load changes, harmonic pollution and the nonlinear behavior of power equipment often lead to phase error, frequency instability and power quality problems of the power grid, which poses a great threat to the reliability, economy and safety of power systems. Therefore, accurate and efficient phase synchronization and frequency control have become the key to stable operation of power systems.

[0003] The phase-locked loop (PLL) is one of the commonly used synchronization controllers in power systems, which is used to synchronize the input signal with the reference signal, and then realize accurate control of the grid frequency and phase. However, as known from the articles such as "Grid Signal Synchronization Method Based on Cascaded Filter Phase-Locked Loop" and "Design of New Grid-Connected Phase-Locked Loop Under Grid Voltage Unbalance and Harmonic Distortion", the traditional PLL controller is prone to problems such as decrease in phase locking accuracy, response delay and frequency fluctuation when facing grid imbalance, frequency fluctuation and noise interference, which limits its application in modern complex power grids.

[0004] In order to solve the shortcomings of the traditional PLL method, the present application provides an improved grid-connected switching DSOGI-LADRC-PLL adaptive control method, which combines double second-order generalized integrator (DSOGI), linear active disturbance rejection control (LADRC) and HPDC technology, and comprehensively utilizes the high-order dynamic information of the error and the adaptive ability of the system to improve the accuracy, response speed and robustness of the control system. It has significant advantages in dealing with dynamic changes, frequency fluctuations and grid imbalance, and is expected to become an important development direction of phase synchronization and frequency control technology in smart grid and distributed power generation systems. SUMMARY

[0005] The purpose of this invention is to provide an improved DSOGI-LADRC-PLL adaptive control method for grid connection switching, which aims to accurately lock the grid phase and effectively suppress interference during grid connection switching to address the technical problems of grid imbalance, frequency fluctuation, and noise interference in modern power systems. This method combines dynamic gain adjustment, intelligent gain regulator module, HPDC technology, and LADRC control strategy, and has high precision, high responsiveness, and strong robustness, making it suitable for complex power system environments such as smart grids, microgrids, and distributed generation systems.

[0006] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows:

[0007] An improved method for obtaining the DSOGI-LADRC-PLL phase-locked signal during grid-connected switching includes the following steps:

[0008] Step 1: Input three-phase voltage, perform Clarke transformation on the three-phase voltage signal to convert the three-phase signal into a two-phase quadrature signal, and perform DSOGI filtering on the obtained quadrature component to obtain the instantaneous phase and amplitude, and calculate and adjust the phase error range.

[0009] Step 2: Use high-order dynamic predictive control (HPDC) technology to model and predict phase errors;

[0010] Step 3: Define the gain adjustment function and combine it with the fuzzy control strategy to optimize the gain of the LADRC controller in real time to adapt to different load and disturbance conditions;

[0011] Step 4: Use an improved LADRC controller to process phase errors and output a precise phase-locked signal.

[0012] In step 1, when inputting three-phase voltages and performing Clarke transformation on the three-phase signals, the specific steps are as follows:

[0013] Assume the three-phase voltage input signal is:

[0014]

[0015] Where: v a v b v c This is a three-phase voltage signal; V m ω represents the amplitude of the three-phase voltage; t represents time; ω represents the fundamental frequency angular velocity of the power grid.

[0016] When converting a three-phase signal into two-phase components in the α-β coordinate system, a coefficient σ is introduced to compensate for the voltage imbalance due to the existence of unbalanced voltages in the power grid.

[0017]

[0018] Wherein, σ is an adjustment coefficient based on the imbalance factor, and its value can be dynamically determined by detecting the three-phase voltage;

[0019] σ can be obtained from the following formula:

[0020]

[0021] Where k is a scaling factor used to adjust the intensity of unbalanced compensation; v avg For v a v b v c The average value; the max function is used to find the maximum value among them;

[0022] After Clarke transform, two-phase signals v were obtained. α With v β Two-phase signal v α With v β It is orthogonal in the α-β plane and retains the fundamental frequency component of the original three-phase signal.

[0023] In step 1, the DSOGI filtering is performed as follows:

[0024] After obtaining the α-β components, for v α and v β Filtering is performed to remove harmonic components, resulting in filtered quadrature components. and To better improve the system's response under frequency fluctuation conditions, dynamic gain adjustment is introduced. The formula is:

[0025]

[0026] Where k(t) is an adjustment coefficient that varies with time and can be adjusted according to the real-time frequency error; ξ(t) is the damping coefficient, which is adaptively adjusted according to the filtered components of the input signal; ω O (t) is the center frequency of the filter, which can be adjusted in real time according to the grid drift.

[0027] The instantaneous phase and amplitude of the orthogonal components are obtained from the filtering results, as follows:

[0028] Based on the orthogonal components after DOSGI filtering and The instantaneous phase angle and amplitude can then be obtained using the following formula:

[0029] Instantaneous phase angle θ(t):

[0030]

[0031] Amplitude V(t):

[0032]

[0033] The specific steps for calculating and adjusting the phase error range are as follows:

[0034] Phase error θ err (t) is a key quantity used by the PLL controller for calibration, which can represent the difference between the instantaneous phase angle and the desired reference phase angle;

[0035] First, an internal reference phase angle is generated by the PLL:

[0036] θ ref (t)=ω ref ·t+θ ref (0) (7)

[0037] Where, θ ref (t) represents the phase angle for reference; ω ref This is the PLL reference frequency, synchronized with the input signal frequency.

[0038] Phase error is defined as the instantaneous phase angle θ(t) and the reference phase angle θ. ref The difference (t); based on this, by adding a phase drift compensation term Δθ c (t), to compensate for phase shift caused by filtering delay or power grid frequency fluctuations:

[0039] θ err (t)=θ(t)-θ ref (t)+Δθ c (t) (8)

[0040] Where, Δθ c (t) Dynamically adjust based on frequency shift and filter delay to enhance the system's sensitivity to phase changes;

[0041] To ensure that the phase error remains within a reasonable range, such as from -π to π, an adjustment in the range is needed. Modulo operations are used to limit the phase error to the required range.

[0042] θ err (t) = mod(θ) err (t)+π,2π)-π (9)

[0043] The modulo operation can constrain the phase error within the range of (-π, π], ensuring the continuity and numerical stability of the phase error.

[0044] Step 2: Introduce High-Order Dynamic Predictive Control (HPDC) technology to achieve high-order dynamic prediction of phase error and its evolution trend; specifically:

[0045] The core of HPDC lies in error prediction and dynamic adjustment. By modeling the rate of change of phase error, the system can dynamically predict the future trend of phase error; in this invention, the phase error θ is used... err (t) can be extended to have its derivative as acceleration is Therefore, a model can be established:

[0046]

[0047] in, The phase error for the predicted next time step; This represents the phase error at the current moment; and These represent the rate of change and acceleration of the phase error, respectively; Δt is the sampling time.

[0048] This model considers both the rate and acceleration of error, allowing for a more detailed capture of the system's dynamic changes. HPDC utilizes this dynamic information to adjust the control gain at each sampling time, thus adapting to future trends.

[0049] Step 3 includes the following steps:

[0050] Step 3.1: Define the gain adjustment function:

[0051] Three dynamically predicted values ​​defined Define the gain adjustment function of the controller:

[0052]

[0053] in, and It is a gain adjustment function that is dynamically adjusted in real time based on the magnitude, rate of change, and acceleration of the prediction error;

[0054] Step 3.2: Set and optimize the gain adjustment function.

[0055] In step 3.2, the gain adjustment function is set as follows;

[0056] Fuzzy control is a highly adaptable and easy-to-implement intelligent adjustment method. This invention uses this method to adjust the parameters of the gain adjustment function based on the current error and the rate of change of error by setting a series of fuzzy rules.

[0057] Step 3.2.1: In this invention, the input variable error is... Error change rate With changing acceleration The variables are fuzzified into three variables: x1, x2, and x3. The output variable k is then... pi With k di The variables are fuzzyened into multiple language variables, such as small (S), medium (M), and large (L). Other variables are similar and will not be explained here.

[0058] Step 3.2.2: Mathematical Representation of Membership Functions

[0059] This invention uses triangular membership functions, and the membership degree can be expressed as:

[0060]

[0061] Where a, b, c, d, e, and f are all endpoints of the intervals defining linguistic variables, and the shape and position of the membership function are determined by these endpoint parameters.

[0062] Step 3.2.3: Construction of the fuzzy rule table;

[0063] The constructed fuzzy rule table is as follows:

[0064] Input variable error With error change rate Combine them, and then output variable k. pi With k di It is determined by both of the fuzzy variables;

[0065] When both inputs are x1 or x3, the output is L; when at least one input is x2, both will output M; when the two inputs are different and neither is x2, the output is S; where L refers to the larger output fuzzy set; M refers to the medium output fuzzy set; and S refers to the smaller output fuzzy set.

[0066] Fuzzy rules are the core of a fuzzy controller; they link the input fuzzy set with the output fuzzy set.

[0067] The fuzzy rule table is constructed as follows:

[0068] Input variable error With error change rate Combine them, and then output variable k. pi With k di It is determined by both of the fuzzy variables.

[0069] When both inputs are x1 or x3, the output is L; when at least one input is x2, both will output M; when the two inputs are different and neither is x2, the output is S; where L refers to the larger output fuzzy set; M refers to the medium output fuzzy set; and S refers to the smaller output fuzzy set.

[0070] Example of the meaning of the above rules:

[0071] Rule 1: If For x1 and If x1, then k pi Let L, k di Let L be the value.

[0072] Rule 2: If For x2 and If x2, then k pi For M, k di For M.

[0073] These rules are expressed in IF-THEN form and are set through experimentation.

[0074] Step 3.2.4: Fuzzy Reasoning

[0075] During fuzzy inference, the fuzzy controller links the fuzzy sets of input and output according to fuzzy rules. The inference computation employs the Mamdani fuzzy inference method.

[0076] Assuming input error With error change rate The corresponding membership functions are as follows:

[0077] Find the corresponding rule based on the fuzzy rule table, for example: if It is x1 and If it is x1, then k p (t) is L.

[0078] Then, the Min-Max method is used as the fuzzy inference method, selecting the minimum membership degree of the input variable as the membership degree of the output fuzzy set; at this time, the membership degree is: μ M (k p (t))=min(0.7,0.8)=0.7.

[0079] Wherein, min(0.7,0.8) takes the minimum value between 0.7 and 0.8, which is 0.7;

[0080] Step 3.2.5: Deblurring

[0081] Defuzzification transforms the fuzzy output obtained from fuzzy inference into precise numerical values ​​for practical gain adjustment. The defuzzification method used in this invention is the centroid method, and the formula is as follows:

[0082]

[0083] Where: n is the number of output fuzzy sets; μ i It is the membership degree of the output fuzzy set i; It is the specific output value corresponding to the fuzzy variable i.

[0084] Assuming the output is The fuzzy sets small (S), medium (M), and large (L) correspond to respectively but:

[0085]

[0086] Similarly, assuming the output The fuzzy sets small (S), medium (M), and large (L) correspond to respectively but:

[0087]

[0088] The final accurate output k can be obtained from the above formula. p (t) and k d (t), and feed it back to the control system.

[0089] Step 4: The controller outputs a precise phase-locked signal to ensure that the system maintains excellent stability and accurate phase tracking performance when the power grid frequency fluctuates or is subjected to external disturbances;

[0090] The goal of the LADRC control law is to use phase error feedback to generate control inputs so that the system phase error approaches zero.

[0091] The LADRC control law can be expressed as:

[0092]

[0093] Where, k p (t) represents the proportional gain, used to adjust the feedback of phase error; k d (t) is the differential gain, used to suppress rapid changes in phase error; It is the disturbance estimate predicted by HPDC, used to adjust the system output.

[0094] The present invention also includes a method for optimizing the gain of an LADRC controller. By defining a gain adjustment function and optimizing the parameters of the LADRC controller in real time, the method can adapt to different load and disturbance conditions. This method can solve the technical problems of existing technologies such as "Research on Control Methods Based on Fuzzy PID Controllers" which are difficult to deal with dynamic disturbances, system nonlinearity and accurate phase error tracking in complex environments.

[0095] Specifically, the following steps are included:

[0096] Step S1: Define the gain adjustment function of the controller;

[0097] Step S2: Set and optimize the gain adjustment function;

[0098] In step S1, the proportional gain k is defined. p (t) and differential gain k d (t).

[0099] In step S1, the phase error at the current moment is first defined. Rate of change of phase error and the acceleration due to changes in phase error Then define the proportional gain k respectively. p (t) and differential gain k d (t) is:

[0100]

[0101] in, and It is a gain adjustment function that is dynamically adjusted in real time based on the magnitude, rate of change, and acceleration of the predicted phase error.

[0102] Step S2 includes the following steps:

[0103] S2-1: Input variable error Error change rate With changing acceleration The variables are fuzzed into three variables: x1, x2, and x3; the output variable k is then fuzzified. pi With k di Fuzzyization into multiple linguistic variables;

[0104] S2-2: Obtain the membership function, then the membership degree can be expressed as:

[0105]

[0106] Where a, b, c, d, e, and f are all endpoints of the intervals defining linguistic variables, and the shape and position of the membership function are determined by these endpoint parameters;

[0107] S2-3: Construct a fuzzy rule table; the fuzzy rule table is specifically constructed as follows:

[0108] Input variable error With error change rate Combine them, and then output the variable. and It is determined by both of the fuzzy variables;

[0109] When both inputs are x1 or x3, the output is L; when at least one input is x2, both will output M; when the two inputs are different and neither is x2, the output is S; where L refers to the larger output fuzzy set; M refers to the medium output fuzzy set; and S refers to the smaller output fuzzy set.

[0110] S2-4: Perform fuzzy inference on the error, error rate of change, and error acceleration in the control system;

[0111] S2-5: Defuzzify the fuzzy output values;

[0112] Defuzzification transforms the fuzzy output obtained from fuzzy inference into precise numerical values ​​for practical gain adjustment. The defuzzification method used in this invention is the centroid method, and the formula is as follows:

[0113]

[0114] Where: n is the number of output fuzzy sets; μ i It is the membership degree of the output fuzzy set i; It is the specific output value corresponding to the fuzzy variable i;

[0115] Assuming the output is The fuzzy sets small (S), medium (M), and large (L) correspond to respectively but:

[0116]

[0117] Similarly, assuming the output The fuzzy sets small (S), medium (M), and large (L) correspond to respectively but:

[0118]

[0119] The final accurate output k can be obtained from the above formula. p (t) and k d (t), and feed it back to the control system.

[0120] This method combines Linear Active Disturbance Suppression Control (LADRC) and fuzzy gain optimization to address the limitations of existing control methods in accurately tracking phase errors under dynamic disturbances, system nonlinearity, and complex environments. Specifically, it utilizes the phase error feedback characteristic of the LADRC control law to achieve precise suppression of phase errors, while dynamically adjusting the proportional and derivative gains through fuzzy control enables the system to adapt in real-time to changes in error, error rate of change, and acceleration. This approach significantly improves the controller's adaptability, disturbance rejection capability, and robustness, exhibiting excellent stability and tracking accuracy, especially in nonlinear systems or scenarios with frequent disturbances.

[0121] This method significantly improves the phase tracking performance and response speed of the system in dynamic environments, ensuring excellent system stability under grid frequency fluctuations or external disturbances. Furthermore, by fuzzifying the input variables and performing defuzzification based on the centroid method, this method reduces the reliance on precise system modeling and enhances its adaptability to complex disturbances and uncertain environments. Compared with traditional methods, this method demonstrates higher control performance and application value in accurate phase tracking, dynamic gain optimization, and handling complex nonlinear disturbances.

[0122] Compared with the prior art, the present invention has the following technical effects:

[0123] 1) This invention presents an improved DSOGI-LADRC-PLL adaptive control method for grid connection switching, aiming to improve the accuracy, response speed, and robustness of phase synchronization in power systems. This method is widely applicable to smart grids, microgrids, distributed generation systems, and other fields, enhancing the synchronization accuracy and robustness of the system and ensuring the stable operation of the power system.

[0124] 2) An improved Clarke transform method is proposed. By introducing coefficients to compensate for voltage imbalance and combining it with DSOGI filtering, the instantaneous phase and amplitude of the two-phase quadrature components are extracted, and the phase error is dynamically adjusted, which significantly improves the phase synchronization accuracy under voltage imbalance.

[0125] 3) High-order dynamic predictive control (HPDC) technology is introduced to dynamically model and predict phase error, making full use of the error's rate of change and acceleration information, thereby optimizing the system's response speed and stability under grid frequency fluctuation conditions.

[0126] 4) Define a gain adjustment function and combine it with a fuzzy control strategy to optimize the gain parameters of the LADRC controller in real time for different load and disturbance conditions, thereby enhancing the system's adaptability and anti-disturbance capability.

[0127] 5) An improved LADRC controller is used to output a phase-locked signal. By introducing a phase drift compensation term, the effects of filtering delay and frequency fluctuation are effectively overcome, ensuring high-precision synchronization and stable operation of the power system in complex environments.

[0128] 6) This invention proposes a method for extracting orthogonal components of three-phase voltage signals using DSOGI filtering and Clarke transform, as well as for calculating and adjusting phase errors. This invention improves the traditional Clarke transform by introducing a coefficient σ to compensate for voltage imbalance; furthermore, to better enhance the system's response under frequency fluctuation conditions, dynamic gain adjustment is introduced. Additionally, a phase drift compensation term Δθ is added to the original phase error calculation. c (t) compensates for phase shift caused by filtering delay or power grid frequency fluctuations. Attached Figure Description

[0129] The present invention will be further described below with reference to the accompanying drawings and embodiments:

[0130] Figure 1 This is a flowchart of the present invention. Detailed Implementation

[0131] A DSOGI-LADRC-PLL adaptive control method for grid connection switching includes the following steps:

[0132] Step 1: Input three-phase voltage, perform Clarke transformation on the three-phase voltage signal through the DSOGI module to convert the three-phase signal into a two-phase quadrature signal, and perform DSOGI filtering on the obtained quadrature component to obtain the instantaneous phase and amplitude, and calculate and adjust the phase error range.

[0133] Step 1.1: Clarke Transform of the Three-Phase Input Signal Model

[0134] Assume the three-phase voltage input signal is:

[0135]

[0136] Where: v a v b v c This is a three-phase voltage signal; V m ω represents the amplitude of the three-phase voltage; ω represents the fundamental frequency angular velocity of the power grid.

[0137] When converting a three-phase signal into two-phase components in the α-β coordinate system, due to the problem of unbalanced voltage in the power grid, this invention introduces a coefficient σ to compensate for the voltage imbalance:

[0138]

[0139] Wherein, σ is an adjustment coefficient based on the imbalance factor, and its value can be dynamically determined by detecting the three-phase voltage.

[0140] σ can be obtained from the following formula:

[0141]

[0142] Where k is a scaling factor used to adjust the intensity of unbalanced compensation; v avg For v a v b v c The average value; the max function is used to find the maximum value among them.

[0143] After Clarke transform, two-phase signals v were obtained. α With v β They are orthogonal in the α-β plane and retain the fundamental frequency components of the original three-phase signals.

[0144] Step 1.2: DSOGI Filtering

[0145] After obtaining the α-β components, for v α and v β Filtering is performed to remove harmonic components. In standard DSOGI, the transfer function typically uses a fixed gain parameter k and a center frequency ω. O Filter characteristics are defined to decompose the orthogonal components of the signal. In this invention, dynamic gain adjustment is introduced to better improve the system's response under frequency fluctuation conditions. The formula is then:

[0146]

[0147] Where k(t) is an adjustment coefficient that varies with time and can be adjusted according to the real-time frequency error; ξ(t) is the damping coefficient, which is adaptively adjusted according to the filtered components of the input signal; ω O (t) is the center frequency of the filter, which can be adjusted in real time according to the grid drift.

[0148] Step 1.3: Calculation of instantaneous phase angle and amplitude of orthogonal components

[0149] Based on the orthogonal components after DOSGI filtering and The instantaneous phase angle and amplitude can then be obtained using the following formula:

[0150] Instantaneous phase angle θ(t):

[0151]

[0152] Amplitude V(t):

[0153]

[0154] Step 1.4: Calculation and Range Adjustment of Phase Error

[0155] Phase error θ err (t) is a key quantity used by the PLL controller for calibration, which can represent the difference between the instantaneous phase angle and the desired reference phase angle.

[0156] First, an internal reference phase angle is generated by the PLL:

[0157] θ ref (t)=ω ref ·t+θ ref (0) (7)

[0158] Where, θ ref (t) represents the phase angle for reference; ω ref This is the PLL reference frequency, synchronized with the input signal frequency.

[0159] Phase error is defined as the instantaneous phase angle θ(t) and the reference phase angle θ. ref The difference (t). Based on this, the present invention adds a phase drift compensation term Δθ. c (t), to compensate for phase shift caused by filtering delay or power grid frequency fluctuations:

[0160] θ err (t)=θ(t)-θ ref (t)+Δθ c (t) (8)

[0161] Where, Δθ c (t) Dynamically adjust based on frequency offset and filter delay to enhance the system’s sensitivity to phase changes.

[0162] To ensure that the phase error remains within a reasonable range (e.g., -π to π), adjustments to the range are necessary. Modulo operations can generally be used to limit the phase error to the desired range.

[0163] θ err (t) = mod(θ) err (t)+π,2π)-π (9)

[0164] The modulo operation can constrain the phase error within the range of (-π, π], ensuring the continuity and numerical stability of the phase error.

[0165] Step 2: Use high-order dynamic predictive control (HPDC) technology to model and predict phase error;

[0166] Step 2.1: Error Dynamic Modeling and Prediction

[0167] The core of HPDC lies in error prediction and dynamic adjustment. By modeling the rate of change of phase error, the system can dynamically predict the future trend of phase error. In this invention, the known phase error θ... err (t) can be extended to have its derivative as acceleration is Therefore, a model can be established:

[0168]

[0169] in, The phase error for the predicted next time step; This represents the phase error at the current moment; and These represent the rate of change and acceleration of the error, respectively; Δt is the sampling time.

[0170] This model considers both the rate and acceleration of error, allowing for a more detailed capture of the system's dynamic changes. HPDC utilizes this dynamic information to adjust the control gain at each sampling time, thus adapting to future trends.

[0171] Three dynamically predicted values ​​defined The gain adjustment function of the controller can be obtained:

[0172]

[0173] Among them, f(·) and g(·) are gain adjustment functions that are dynamically adjusted in real time according to the magnitude, rate of change and acceleration of the predicted phase error.

[0174] Step 3: Define the gain adjustment function and combine it with the fuzzy control strategy to optimize the gain of the LADRC controller in real time to adapt to different load and disturbance conditions;

[0175] Step 3.1: Gain Adjustment Function

[0176] The intelligent gain regulator module is used to monitor the dynamic state of the system in real time and adaptively adjust the corresponding parameters of the gain regulation function of the LADRC controller, thereby optimizing system performance.

[0177] Define gain adjustment functions f(·) and g(·), and use fuzzy control to adjust the gain k. p With gain k d Adjustments were made, and the formulas are shown in equations (11) and (12).

[0178] The intelligent gain regulator uses fuzzy rules (such as increasing k when the error increases) to adjust the gain. pWhen the rate of change of error increases, k increases. d Generate gain adjustment strategy.

[0179] Step 3.2: Set the gain adjustment function

[0180] Fuzzy control is a highly adaptable and easily implemented intelligent adjustment method. This invention uses this method to adjust the parameters of the gain adjustment function based on the current error and the rate of change of the error by setting a series of fuzzy rules.

[0181] Step 3.2.1: In this invention, we will input the variable error. Error change rate With changing acceleration The variables are blurred into three variables: x1, x2, and x3. The output variable is k. p The variables are fuzzyened into multiple language variables, such as small (S), medium (M), and large (L). Other variables are similar and will not be explained here.

[0182] Step 3.2.2: Mathematical Representation of Membership Functions

[0183] This invention uses triangular membership functions, and the membership degree can be expressed as:

[0184]

[0185] Where a, b, c, d, e, and f are all endpoints of the intervals defining linguistic variables, and the shape and position of the membership function are determined by these endpoint parameters.

[0186] Step 3.2.3: Construction of the fuzzy rule table

[0187] Fuzzy rules are the core of a fuzzy controller; they relate the input fuzzy set to the output fuzzy set. The following is an example of constructing a fuzzy rule table:

[0188] Table 1: Fuzzy Rule Table

[0189]

[0190] Example of the meaning of the above rules:

[0191] Rule 1: If It is x1 and If it is x1, then k p It is L.

[0192] Rule 2: If It is x2 and If it is x2, then k p It's M.

[0193] These rules are expressed in IF-THEN form and are set through experimentation.

[0194] Step 3.2.4: Fuzzy Reasoning

[0195] During fuzzy inference, the fuzzy controller links the fuzzy sets of input and output according to fuzzy rules. The inference computation employs the Mamdani fuzzy inference method.

[0196] Assume the input error is Error change rate The corresponding membership functions

[0197] Find the corresponding rule based on the fuzzy rule table, for example: if It is x1 and If it is x1, then k p It is L.

[0198] Then, the Min-Max method is used as the fuzzy inference method, selecting the minimum membership degree of the input variable as the membership degree of the output fuzzy set. At this point, the membership degree is: μ M (k p =min(0.7,0.8) = 0.7

[0199] Step 3.2.5: Deblurring

[0200] Defuzzification transforms the fuzzy output obtained from fuzzy inference into precise numerical values ​​for practical gain adjustment. The defuzzification method used in this invention is the centroid method, and the formula is as follows:

[0201]

[0202] Where: n is the number of output fuzzy sets; μ i It outputs the membership degree of fuzzy set i; k pi It is the specific output value corresponding to the fuzzy variable i.

[0203] Assuming the output k p The fuzzy sets small (S), medium (M), and large (L) correspond to respectively but:

[0204]

[0205] Similarly, assuming the output The fuzzy sets small (S), medium (M), and large (L) correspond to respectively but:

[0206]

[0207] The final accurate output k can be obtained from the above formula. p (t) and k d (t), and feed it back to the control system.

[0208] Step 4: The controller outputs a precise phase-locked signal to ensure that the system maintains excellent stability and accurate phase tracking performance when the grid frequency fluctuates or is subjected to external disturbances.

[0209] Step 4.1: The goal of the LADRC control law is to use phase error feedback to generate control inputs to ensure that the system phase error approaches zero.

[0210] The LADRC control law can be expressed as:

[0211]

[0212] Where, k p (t) represents the proportional gain, used to adjust the feedback of phase error; k d (t) is the differential gain, used to suppress rapid changes in phase error; It is the disturbance estimate predicted by HPDC, used to adjust the system output.

[0213] The present invention also includes a method for optimizing the gain of an LADRC controller. By defining a gain adjustment function and optimizing the parameters of the LADRC controller in real time, the method can adapt to different load and disturbance conditions. This method can solve the technical problems of existing technologies such as "Research on Control Methods Based on Fuzzy PID Controllers" which are difficult to deal with dynamic disturbances, system nonlinearity and accurate phase error tracking in complex environments.

[0214] Specifically, the following steps are included:

[0215] Step S1: Define the gain adjustment function of the controller;

[0216] Step S2: Set and optimize the gain adjustment function;

[0217] In step S1, the proportional gain k is defined. p (t) and differential gain k d (t).

[0218] In step S1, the phase error at the current moment is first defined. Rate of change of phase error and the acceleration due to changes in phase error Then define the proportional gain k respectively. p (t) and differential gain k d (t) is:

[0219]

[0220] in, and It is a gain adjustment function that is dynamically adjusted in real time based on the magnitude, rate of change, and acceleration of the predicted phase error.

[0221] Step S2 includes the following steps:

[0222] S2-1: Input variable error Error change rate With changing acceleration The variables are fuzzed into three variables: x1, x2, and x3; the output variables are then... and Fuzzyization into multiple linguistic variables;

[0223] S2-2: Obtain the membership function, then the membership degree can be expressed as:

[0224]

[0225] Where a, b, c, d, e, and f are all endpoints of the intervals defining linguistic variables, and the shape and position of the membership function are determined by these endpoint parameters;

[0226] S2-3: Construct a fuzzy rule table; the fuzzy rule table is specifically constructed as follows:

[0227] Input variable error With error change rate Combine them, and then output the variable. and It is determined by both of the fuzzy variables;

[0228] When both inputs are x1 or x3, the output is L; when at least one input is x2, both will output M; when the two inputs are different and neither is x2, the output is S; where L refers to the larger output fuzzy set; M refers to the medium output fuzzy set; and S refers to the smaller output fuzzy set.

[0229] S2-4: Perform fuzzy inference on the error, error rate of change, and error acceleration in the control system;

[0230] S2-5: Defuzzify the fuzzy output values;

[0231] Defuzzification transforms the fuzzy output obtained from fuzzy inference into precise numerical values ​​for practical gain adjustment. The defuzzification method used in this invention is the centroid method, and the formula is as follows:

[0232]

[0233] Where: n is the number of output fuzzy sets; μ i It is the membership degree of the output fuzzy set i; It is the specific output value corresponding to the fuzzy variable i;

[0234] Assuming the output is The fuzzy sets small (S), medium (M), and large (L) correspond to respectively but:

[0235]

[0236] Similarly, assuming the output The fuzzy sets small (S), medium (M), and large (L) correspond to respectively but:

[0237]

[0238] The final accurate output k can be obtained from the above formula. p (t) and k d (t), and feed it back to the control system.

[0239] This method combines Linear Active Disturbance Suppression Control (LADRC) and fuzzy gain optimization to address the limitations of existing control methods in accurately tracking phase errors under dynamic disturbances, system nonlinearity, and complex environments. Specifically, it utilizes the phase error feedback characteristic of the LADRC control law to achieve precise suppression of phase errors, while dynamically adjusting the proportional and derivative gains through fuzzy control enables the system to adapt in real-time to changes in error, error rate of change, and acceleration. This approach significantly improves the controller's adaptability, disturbance rejection capability, and robustness, exhibiting excellent stability and tracking accuracy, especially in nonlinear systems or scenarios with frequent disturbances.

[0240] This method significantly improves the phase tracking performance and response speed of the system in dynamic environments, ensuring excellent system stability under grid frequency fluctuations or external disturbances. Furthermore, by fuzzifying the input variables and performing defuzzification based on the centroid method, this method reduces the reliance on precise system modeling and enhances its adaptability to complex disturbances and uncertain environments. Compared with traditional methods, this method demonstrates higher control performance and application value in accurate phase tracking, dynamic gain optimization, and handling complex nonlinear disturbances.

Claims

1. A method for obtaining the DSOGI-LADRC-PLL phase-locked signal during grid connection switching, characterized in that, Includes the following steps: Step 1: Input three-phase voltage, perform Clarke transformation on the three-phase voltage signal to convert the three-phase signal into a two-phase quadrature signal, and perform DSOGI filtering on the obtained quadrature component to obtain the instantaneous phase and amplitude, and calculate and adjust the phase error range. Step 2: Use high-order dynamic predictive control (HPDC) technology to model and predict phase errors; Step 3: Define the gain adjustment function and combine it with the fuzzy control strategy to optimize the gain of the LADRC controller in real time to adapt to different load and disturbance conditions; Step 4: Use an improved LADRC controller to process phase errors and output a precise phase-locked signal; In step 1, when inputting three-phase voltages and performing Clarke transformation on the three-phase signals, the specific steps are as follows: Assume the three-phase voltage input signal is: （1）； in: , , It is a three-phase voltage signal; This refers to the amplitude of the three-phase voltage; t For time; The fundamental frequency angular velocity of the power grid; Convert the three-phase signal to When dealing with two-phase components in a coordinate system, a coefficient is introduced due to the unbalanced voltage problem in the power grid. To compensate for the imbalance of various voltages: （2）； in, The adjustment coefficient is based on the imbalance factor and its value can be dynamically determined by detecting the three-phase voltage. It can be obtained from the following formula : (3); in, This is a scaling factor used to adjust the intensity of unbalanced compensation. for , , The average value; the max function is used to find the maximum value among them; After Clarke transform, two-phase signals were obtained. and Two-phase signal and exist They are orthogonal in the plane and retain the fundamental frequency components of the original three-phase signals.

2. The method according to claim 1, characterized in that, In step 1, the DSOGI filtering is performed as follows: In obtaining After portioning, for and Filtering is performed to remove harmonic components, resulting in filtered quadrature components. and To better improve the system's response under frequency fluctuation conditions, dynamic gain adjustment is introduced. The formula is: （4）； in, The adjustment factor is a time-varying factor that can be adjusted based on real-time frequency error; This is the damping coefficient, which will be adaptively adjusted according to the filtered components of the input signal; This is the center frequency of the filter, which can be adjusted in real time according to grid drift.

3. The method according to claim 2, characterized in that, The instantaneous phase and amplitude of the orthogonal components are obtained from the filtering results, as follows: Based on the orthogonal components after DOSGI filtering and The instantaneous phase angle and amplitude can then be obtained using the following formula: Instantaneous phase angle : （5）； Amplitude : （6）。 4. The method according to any one of claims 1 to 3, characterized in that, The specific steps for calculating and adjusting the phase error range are as follows: Phase error It is a key quantity used by the PLL controller for calibration, and it can represent the difference between the instantaneous phase angle and the desired reference phase angle; First, an internal reference phase angle is generated by the PLL: （7）； in, Used as the reference phase angle; This is the PLL reference frequency, synchronized with the input signal frequency. Phase error is defined as the instantaneous phase angle. and reference phase angle The difference; based on this, by adding a phase drift compensation term. To compensate for phase shifts caused by filtering delays or power grid frequency fluctuations: （8）； in, The system is dynamically adjusted based on frequency shift and filter delay to enhance its sensitivity to phase changes. To ensure that the phase error remains within a reasonable range, adjustments to the range are needed. Modulo operations are used to limit the phase error to the required range. （9）； The modulo operation can constrain the phase error within... Within the range, ensure the continuity of phase error and the stability of the value.

5. The method according to claim 1, characterized in that, In step 2, high-order dynamic predictive control (HPDC) technology is introduced to achieve high-order dynamic prediction of phase error and its evolution trend; specifically: Through phase error Its rate of change can be extended to acceleration is Therefore, a model can be established: （10）； in, The phase error for the predicted next time step; This represents the phase error at the current moment; and These are the rate of change and acceleration of the phase error, respectively. Sampling time.

6. The method according to claim 1, characterized in that, In step 4, the controller outputs a precise phase-locked signal to ensure that the system maintains excellent stability and accurate phase tracking performance when the grid frequency fluctuates or is subjected to external disturbances. Specifically, it is expressed as follows: （20）； in, This is the proportional gain, used to adjust the feedback for phase error; It is the differential gain, used to suppress rapid changes in phase error; It is the disturbance estimate predicted by HPDC, used to adjust the system output.

Images