Adaptive complex coefficient filter-based variable frequency power grid state estimation method, system, device and medium

Abstract

The adaptive complex coefficient filter-based variable frequency power grid state estimation method, system, device and medium adopt two double-input double-output adaptive complex coefficient filters to extract positive and negative sequence components of the variable frequency power grid in a cross decoupling manner; the positive sequence component extracted by the adaptive complex coefficient filter is used for a phase-locked loop; the grid frequency output by the phase-locked loop is fed back to the adaptive complex coefficient filter and the PI adaptive phase-locked loop to correct the cutoff frequency and PI parameters of the model in real time, so that the frequency change range falls within the fast dynamic response interval of the system, and the dynamic performance of the phase-locked loop under the variable frequency power grid is improved; the system and the device can perform adaptive complex coefficient filter-based variable frequency power grid state estimation according to the method; and the application has the advantages of good dynamic response performance, high efficiency and accuracy.

Description

Technical Field

[0001] This invention belongs to the field of power system synchronization technology, and particularly relates to a variable frequency power grid state estimation method, system, equipment and medium based on an adaptive complex coefficient filter. Background Technology

[0002] Compared to traditional aircraft, Multi-Electric Aircraft (MEA) offer advantages such as lighter weight, higher reliability, and lower losses, attracting increasing attention and becoming a trend in aviation technology development. With the increasing capacity of power systems and grids, research on grid technologies such as topology, control, and parameter estimation in MEA is becoming increasingly important. In MEA technology, the estimation of grid voltage parameters provides crucial information for the power system, while frequency and phase estimation are essential for power control and grid synchronization in the MEA grid. MEA grids are typical islanded microgrids with a wide frequency range of 360Hz to 800Hz, and the frequency fluctuates significantly depending on the needs of different functions.

[0003] The single synchronous reference frame phase-locked loop (SSRF-SPLL) is a classic method for power grid synchronization and is often used for parameter estimation in power frequency power grids [Zhang Xing, Zhang Chongwei. PWM Rectifier and Its Control [M], Machinery Industry Press, 2012]. Power frequency power grids have a constant frequency. In contrast, the phase-locked loop method applicable to power frequency power grids will operate far from the equilibrium point in MEA frequency converter power grids, and its performance will be significantly degraded. Estimating frequency and phase in MEA frequency converter power grids is more difficult than in power frequency power grids.

[0004] In [Chen B, Pin G, Ng WM, et al. An Adaptive Observer-Based Robust Estimator of Multi-sinusoidal Signals[J].IEEE Transactions on Automatic Control, 2018, 63(6): 1618-1631], a robust estimation method for multi-sinusoidal signals based on an adaptive observer was proposed to improve the dynamic response of frequency conversion power grid synchronization, but the computational load is large. Summary of the Invention

[0005] To overcome the shortcomings of the prior art, the present invention aims to provide a method, system, device, and medium for estimating the state of a variable frequency power grid based on an adaptive complex coefficient filter. By adaptively changing the cutoff frequency of the complex coefficient filter and the parameters of the PI controller in the phase-locked loop, the frequency variation range falls within the system's fast dynamic response range, thereby improving the dynamic performance of the phase-locked loop under variable frequency power grids. It can achieve stable, fast, and accurate estimation of grid voltage information when there are large and small frequency jumps or gradual frequency increases under variable frequency power grids, with low computational load.

[0006] To achieve the above objectives, the present invention adopts the following technical solution:

[0007] A variable frequency power grid state estimation method based on adaptive complex coefficient filters is proposed. This method employs two dual-input, dual-output adaptive complex coefficient filters in a cross-decoupling manner to extract the positive and negative sequence components of the variable frequency power grid. The positive sequence component extracted by the adaptive complex coefficient filters is used in a phase-locked loop (PLL). By feeding back the grid frequency output from the PLL to the adaptive complex coefficient filters and the PI-adaptive PLL, the cutoff frequency and PI parameters are corrected in real time to achieve an adaptive state. This ensures that the frequency variation range falls within the system's fast dynamic response range, improving the dynamic performance of the PLL under variable frequency power grid conditions and enabling rapid and accurate estimation of the voltage state information of the variable frequency power grid.

[0008] A variable frequency power grid state estimation method based on an adaptive complex coefficient filter specifically includes the following steps:

[0009] Step 1: Two adaptive complex coefficient filters are used in a cross-decoupling manner to extract the positive and negative sequence components of the frequency converter grid. When extracting the positive and negative sequence components of the frequency converter grid, the input-output transfer function of the adaptive complex coefficient filter is expressed as:

[0010]

[0011]

[0012] In the formula, Extract the positive-sequence component of the output of the adaptive complex coefficient filter. To extract the negative-order component of the output of the adaptive complex coefficient filter, v α (s), v β (s) represents the voltage component of the input adaptive complex coefficient filter, j denotes the imaginary unit, s denotes the Laplace operator, and the transfer function G p (s) is a bandpass / notch filter with a cutoff frequency ω. p Accepts the positive sequence component of the frequency converter power grid and rejects the negative sequence component; transfer function G n (s) is a bandpass / notch filter with a cutoff frequency ω. pThe negative sequence component of the frequency converter is allowed, while the positive sequence component is blocked; the passband center frequency is ω. c The sign of the passband center frequency is consistent with the extracted positive and negative sequence components, and the notch frequency is the opposite of the passband center frequency, thus realizing the extraction and separation of the positive and negative sequence components of the frequency conversion power grid.

[0013] Step 2: Apply the adaptive complex coefficient filter described in Step 1 to a phase-locked loop, and linearize the phase-locked loop using the adaptive complex coefficient filter to obtain the transfer function:

[0014]

[0015] Combining the characteristic polynomial in the transfer function with the unit step response of the third-order system in the s-domain, we obtain:

[0016]

[0017] Where: k p k i These are the proportional and integral coefficients of the PI controller in the phase-locked loop structure, respectively; μ is the coefficient that determines the actual pole location based on the natural frequency value; and ω... n ω is the undamped natural oscillation angular frequency; p The cutoff frequency ω can be determined from the transfer function of the adaptive complex coefficient filter in step 1. p The damping factor determines the transfer function; typically, the optimal damping coefficient for a second-order system is... In an adaptive enhancement-type complex coefficient filter, the cutoff frequency ω p Corresponding to The grid frequency output by the phase-locked loop in a variable frequency power grid;

[0018] Expanding the right side of equation (4) and corresponding coefficients of the left side, we can obtain:

[0019] ω p +k p =(2ξ+μ)ω n (5)

[0020]

[0021] ω p k i =μω n 3 (7)

[0022] By combining equations (5) and (6), we can obtain:

[0023]

[0024] Based on the discriminant of a quadratic function and the optimal damping ratio We can obtain:

[0025]

[0026] From the discriminant of the quadratic function in equation (8), we can see that μ is not affected by ω. p Substituting the value of ω into equations (5)-(7) above, we can obtain the cutoff frequency ω of the adaptive complex coefficient filter phase-locked loop. p The proportional coefficient k of the adaptive PI controller p and integral coefficient k i The grid frequency output by the phase-locked loop under the variable frequency power grid The relationship is as follows:

[0027]

[0028]

[0029]

[0030]

[0031] The grid frequency output by the phase-locked loop is obtained from equations (11)-(13). The relationship between the parameters of the adaptive complex coefficient filter phase-locked loop model and the parameters of the phase-locked loop, where the cutoff frequency ω of the adaptive complex coefficient filter phase-locked loop is... p It will vary with the grid frequency output by the phase-locked loop. The bandwidth of the adaptive complex coefficient filter phase-locked loop model changes accordingly; the grid frequency output by the phase-locked loop also changes. As the changes occur, the coefficient k of the PI controller in the phase-locked loop... p sum coefficient k i The value also changes adaptively, thus changing the locking range Δω of the adaptive complex coefficient filter phase-locked loop. L Value:

[0032]

[0033] Input frequency change (frequency change of mains voltage) Δω in , so that Δω in Falling within the system's fast dynamic response range ω g ±Δω L ω g For the frequency of the variable frequency power grid;

[0034] The positive-sequence voltage component extracted by the adaptive complex coefficient filter is input into the PI adaptive phase-locked loop. The input of the voltage-controlled oscillator in the phase-locked loop, plus the output of the integrator in the PI adaptive controller, provides an estimate of the grid frequency. The phase estimate is then obtained through the integrator, which is the detection of the frequency and phase of the positive-sequence component of the frequency conversion grid.

[0035] In step 2, μ is a coefficient used to determine the actual pole location based on the natural frequency value. The coefficient aims to determine the actual pole location as close to the origin as possible, i.e., μ takes the minimum possible value. To maintain the high filtering capability of the adaptive complex coefficient filter phase-locked loop.

[0036] A system for variable frequency power grid state estimation based on an adaptive complex coefficient filter includes:

[0037] Adaptive complex coefficient filter module: used to extract and separate the positive and negative sequence components of the frequency conversion power grid according to the method in step 1;

[0038] PI adaptive phase-locked loop module: used to accurately estimate the frequency and phase information of the positive sequence component of the frequency conversion power grid according to the method in step 2.

[0039] A device for a variable frequency power grid state estimation method based on an adaptive complex coefficient filter includes:

[0040] Memory, used to store computer programs;

[0041] A processor is used to implement the variable frequency power grid state estimation method based on an adaptive complex coefficient filter described in steps 1 to 2 when executing the computer program.

[0042] A computer-readable storage medium for storing a computer program that, when executed by a processor, is capable of performing variable frequency power grid state estimation based on an adaptive complex coefficient filter according to the method described in steps 1 to 2.

[0043] Compared with the prior art, the present invention has the following advantages:

[0044] 1. This invention extracts the positive and negative sequence components of the frequency converter grid by using two adaptive complex coefficient filters, and uses a PI adaptive phase-locked loop to quickly and stably estimate the voltage information of the frequency converter grid accurately. By feeding the output grid frequency back to the adaptive complex coefficient filters and the PI adaptive phase-locked loop, the model parameters are corrected in real time to achieve an adaptive state, which is efficient and accurate.

[0045] 2. This invention adaptively changes the cutoff frequency of the complex coefficient filter and the parameters of the PI controller, so that the frequency change range falls within the system's fast dynamic response range, thereby improving the dynamic performance of the phase-locked loop under the frequency conversion power grid. During large frequency jumps, small frequency jumps, and gradual frequency increases, this invention can quickly and stably estimate the grid voltage information accurately.

[0046] Compared to single synchronous reference frame phase-locked loop (SSRF-SPLL) and complex coefficient filter phase-locked loop (ECCF-PLL), this invention has high dynamic response performance, can quickly track frequency information and has small error; compared to the multi-sinusoidal signal robust estimation method based on adaptive observer, this invention greatly reduces the amount of computation and improves the working efficiency of MEA frequency conversion power grid. Attached Figure Description

[0047] Figure 1 This is a structural diagram of a variable frequency power grid state estimation system based on an adaptive complex coefficient filter in an embodiment of the present invention.

[0048] Figure 2 This is an experimental result diagram of the frequency estimation waveform when the power grid frequency undergoes a small frequency jump of 450-460Hz according to the present invention.

[0049] Figure 3 This is a diagram showing the experimental results of phase error estimation when the power grid frequency experiences a small frequency jump of 450-460Hz.

[0050] Figure 4 This is an experimental result diagram of the frequency estimation waveform when the power grid frequency undergoes a large frequency jump of 450-750Hz.

[0051] Figure 5 This is a diagram showing the experimental results of phase error estimation when the power grid frequency experiences a large frequency jump of 450-750Hz.

[0052] Figure 6 This is an experimental result diagram of the frequency estimation waveform of the power grid frequency during the gradual increase (450-750Hz) of the present invention.

[0053] Figure 7 This is a diagram showing the experimental results of phase error estimation for the power grid frequency during the gradual increase (450-750Hz) of this invention. Detailed Implementation

[0054] The present invention will now be described in further detail with reference to the accompanying drawings.

[0055] A variable frequency power grid state estimation method based on an adaptive complex coefficient filter specifically includes the following steps:

[0056] Step 1: Two adaptive complex coefficient filters are used in a cross-decoupling manner to extract the positive and negative sequence components of the frequency converter grid. When extracting the positive and negative sequence components of the frequency converter grid, the input-output transfer function of the adaptive complex coefficient filter is expressed as:

[0057]

[0058]

[0059] In the formula, Extract the positive-sequence component of the output of the adaptive complex coefficient filter. To extract the negative-order component of the output of the adaptive complex coefficient filter, v α (s), v β (s) represents the voltage component of the input adaptive complex coefficient filter, j denotes the imaginary unit, s denotes the Laplace operator, and the transfer function G p (s) is a bandpass / notch filter with a cutoff frequency ω. p Accepts the positive sequence component of the frequency converter power grid and rejects the negative sequence component; transfer function G n (s) is a bandpass / notch filter with a cutoff frequency ω. p The negative sequence component of the frequency converter is allowed, while the positive sequence component is blocked; the passband center frequency is ω. c The sign of the passband center frequency is consistent with the extracted positive and negative sequence components, and the notch frequency is the opposite of the passband center frequency, thus realizing the extraction and separation of the positive and negative sequence components of the frequency conversion power grid.

[0060] See Figure 1 In actual operation, the passband center frequency is ω c Equivalent to the grid frequency output by the phase-locked loop

[0061] Step 2: Apply the adaptive complex coefficient filter described in Step 1 to a phase-locked loop (PLL), and linearize the PLL using the adaptive complex coefficient filter to obtain the transfer function:

[0062]

[0063] Combining the characteristic polynomial in the transfer function with the unit step response of the third-order system in the s-domain, we obtain:

[0064]

[0065] k p k i These are the proportional gain and integral gain of PI, respectively; μ is the coefficient used to determine the actual pole location based on the natural frequency; and ω... n ω is the undamped natural oscillation angular frequency; pThe cutoff frequency ω can be determined from the transfer function of the adaptive complex coefficient filter in step 1. p The damping factor determines the transfer function; typically, the optimal damping coefficient for a second-order system is... In an adaptive enhancement-type complex coefficient filter, the cutoff frequency ω p Corresponding to This refers to the grid frequency output by the phase-locked loop in a variable frequency power grid.

[0066] Expanding the right side of equation (4) and corresponding coefficients with the left side, we can obtain...

[0067] ω p +k p =(2ξ+μ)ω n (5)

[0068]

[0069]

[0070] By combining equations (5) and (6), we can obtain

[0071]

[0072] According to the discriminant of a quadratic function (due to the proportional gain k...), p It can only have real values ​​(as opposed to complex numbers), therefore the discriminant of the quadratic equation (8) should be greater than or equal to 0, combined with the optimal damping ratio achievable

[0073]

[0074] From the discriminant of the quadratic function in (8), we know that μ is not affected by ω. p To maintain the high filtering capability of the adaptive complex coefficient filter phase-locked loop, the actual pole location should be as close to the origin as possible, i.e., μ should be taken to its minimum possible value. Substituting into equations (5)-(7) above, the cutoff frequency ω of the adaptive complex coefficient filter can be obtained. p The proportional coefficient k in an adaptive PI controller p and integral coefficient k i The grid frequency output by the phase-locked loop under the variable frequency power grid Relationship:

[0075]

[0076]

[0077]

[0078]

[0079] The grid frequency output by the phase-locked loop is obtained from equations (11)-(13). The relationship between the parameters of the adaptive complex coefficient filter phase-locked loop model and the parameters of the phase-locked loop, where the cutoff frequency ω of the adaptive complex coefficient filter phase-locked loop is... p It will vary with the grid frequency output by the phase-locked loop. The bandwidth of the adaptive complex coefficient filter phase-locked loop model changes accordingly; the grid frequency output by the phase-locked loop also changes. As the changes occur, the proportional coefficient k in the PI controller of the phase-locked loop also changes. p and integral coefficient k i The value also changes adaptively, thus changing the locking range Δω of the adaptive complex coefficient filter phase-locked loop. L Value:

[0080]

[0081] Input frequency change (frequency change of mains voltage) Δω in , so that Δω in Falling within the system's fast dynamic response range ω g ±Δω L ω g To improve the dynamic performance of the phase-locked loop (PLL) under the frequency conversion grid, the PLL based on the adaptive complex coefficient filter can also quickly and stably estimate the grid voltage state information under the frequency conversion grid.

[0082] The positive-sequence voltage component extracted by the adaptive complex coefficient filter is input into the PI adaptive phase-locked loop. The input of the voltage-controlled oscillator in the phase-locked loop, plus the output of the integrator in the PI adaptive controller, provides an estimate of the grid frequency. The phase estimate is then obtained through the integrator, thus realizing the detection of the frequency and phase of the positive-sequence component of the frequency conversion grid.

[0083] Example:

[0084] A variable frequency power grid state estimation method based on an adaptive complex coefficient filter specifically includes the following steps:

[0085] See Figure 1 Three-phase frequency converter grid voltage V abc The Clark transformation converts the three-phase stationary coordinate system abc into a two-phase stationary coordinate system for voltage components v. α and v β The converted voltage component v α The input signal to the first adaptive complex coefficient filter is subtracted from the positive-sequence voltage component output after filtering by the adaptive complex coefficient filter in the α-axis coordinate system. Subtract the negative sequence voltage component The positive sequence voltage component can be extracted through cross-decoupling. The converted voltage component v β The input signal to the first adaptive complex coefficient filter is subtracted from the positive-sequence voltage component output after filtering by the adaptive complex coefficient filter in the β-axis coordinate system. Subtract the negative sequence voltage component Afterwards, the positive sequence voltage component can be extracted through cross-decoupling.

[0086] The converted voltage component v α The input signal to the second adaptive complex coefficient filter is subtracted from the negative sequence voltage component output after filtering by the adaptive complex coefficient filter in the α-axis coordinate system. Subtract the positive sequence voltage component Subsequently, the negative sequence voltage component can be extracted through cross-decoupling. The converted voltage component v β The input signal to the second adaptive complex coefficient filter is subtracted from the negative sequence voltage component output after filtering by the adaptive complex coefficient filter in the β-axis coordinate system. Subtract the positive sequence voltage component Subsequently, the negative sequence voltage component can be extracted through cross-decoupling.

[0087] The positive sequence voltage component at the output of the adaptive complex coefficient filter and The two-phase stationary coordinate system αβ is transformed into a two-phase rotating coordinate system dq using the Park transformation, and the transformed two-phase rotating coordinate system v is output. dq At the input terminal of the phase-locked loop (PLL), the PLL control module provides the inverter grid frequency ω in a rotating coordinate system by adding the input of the voltage-controlled oscillator to the output of the integrator in the PI controller. g The estimated value (i.e., the grid frequency output by the phase-locked loop under the variable frequency power grid) Then, the phase estimate is obtained through an integrator. Phase estimate Feedback is sent to the Park conversion module for Park conversion, and the grid frequency output by the phase-locked loop is also fed back. ω fed back to the adaptive complex coefficient filter p The coefficient k in the neutralization PI adaptive controller p k i This allows for frequency adaptation.

[0088] From the transfer function of the adaptive complex coefficient filter, we know that the cutoff frequency ω pThe damping factor determines the transfer function; typically, the optimal damping coefficient for a second-order system is... In an adaptive complex coefficient filter, the cutoff frequency This represents the grid frequency output by the phase-locked loop (PLL) in a variable frequency power grid; the grid frequency output by the PLL... Substituting the values, we can obtain the parameter values ​​of the model.

[0089]

[0090]

[0091]

[0092] The model parameter ω is known p k p k i The value varies with the grid frequency output by the phase-locked loop. The changes are corrected in real time to achieve an adaptive state, corresponding to Figure 1 The coefficients in the model are:

[0093] k1=k2=0.707 (18)

[0094] k3 = 0.1215 (19)

[0095] The positive and negative sequence components of the frequency converter grid are extracted by using two adaptive complex coefficient filters. The extracted positive sequence voltage component is converted from a two-phase stationary coordinate system αβ to a two-phase rotating coordinate system dq. The converted grid voltage is input into an adaptive phase-locked loop. A PI adaptive phase-locked loop is used to detect the frequency and phase of the positive sequence component of the frequency converter grid (usually, the positive sequence voltage component is detected, and the negative sequence component is not needed).

[0096] To verify the effectiveness of the phase-locked loop (PLL) proposed in this invention in variable frequency power grid applications, the performance of the variable frequency power grid state estimation method based on adaptive complex coefficient filter (AECCF-PLL) of this invention, a typical PLL (SSRF-SPLL), and a complex coefficient filter PLL (ECCF-PLL) were compared experimentally under different power grid conditions. The coefficient k of the PI controller in the complex coefficient filter PLL and the typical PLL was also compared. p =222.1, k i =11974, the cutoff frequency ω in a complex coefficient filter p =222.1. Table 1 provides a summary of the dynamic performance of the variable frequency power grid state estimation method based on the adaptive complex coefficient filter and two other phase-locked loops under different conditions.

[0097] Table 1 Performance comparison of different methods

[0098]

[0099] Figure 2 and Figure 3 The experimental waveforms show frequency estimation and phase error when a small frequency jump (450-460Hz) occurs in a variable frequency power grid. The method of this invention (AECCF-PLL) can accurately estimate the power grid frequency within 9ms. It can be seen that the state estimation method of this invention has a fast dynamic response and its phase error converges to zero. In addition, the two phase-locked loops both contain steady-state error, and their stability is also better than the other two phase-locked loops.

[0100] Figure 4 and Figure 5 The experimental waveforms show frequency estimation and phase error when a large frequency jump (450-750Hz) occurs in a variable frequency power grid. The method of this invention (AECCF-PLL) can accurately estimate the power grid frequency within 8ms and the phase error approaches zero after a brief oscillation. The other two phase loops show continuous oscillation. It can be seen that the state estimation method of this invention has a shorter settling time and smoother dynamic behavior compared with the other two phase-locked loops.

[0101] Figure 6 and Figure 7 The experimental waveforms show frequency estimation and phase error during a gradual frequency ramp-up (450-750Hz) in a variable frequency power grid. The method of this invention (AECCF-PLL) achieves near-zero frequency tracking error and zero steady-state error. The complex coefficient filter phase-locked loop (ECCF-PLL) has a significant delay in tracking the frequency and a continuous steady-state error. The single synchronous reference coordinate system phase-locked loop (SSRF-SPLL) continues to oscillate and fluctuate. It can be seen that the state estimation method of this invention has the advantage of particularly good dynamic performance response compared with the other two phase-locked loops, and can quickly track frequency and phase information.

Claims

1. A variable frequency power grid state estimation method based on an adaptive complex coefficient filter, characterized in that, Two dual-input, dual-output adaptive complex coefficient filters are used in a cross-decoupling manner to extract the positive and negative sequence components of the frequency conversion grid. The positive sequence component extracted by the adaptive complex coefficient filters is used in a phase-locked loop (PLL). By feeding the grid frequency output from the PLL back to the adaptive complex coefficient filters and the PI-adaptive PLL, the cutoff frequency and PI controller parameters are corrected in real time to achieve an adaptive state. This ensures that the frequency variation range falls within the system's fast dynamic response range, improving the dynamic performance of the PLL under the frequency conversion grid and enabling rapid and accurate estimation of the voltage state information of the frequency conversion grid. The specific steps are as follows: Step 1: Two adaptive complex coefficient filters are used in a cross-decoupling manner to extract the positive and negative sequence components of the frequency converter grid. When extracting the positive and negative sequence components of the frequency converter grid, the input-output transfer function of the adaptive complex coefficient filter is expressed as: （1） （2） In the formula, , Extract the positive-sequence component of the output of the adaptive complex coefficient filter. , Extract the negative-order component of the output of the adaptive complex coefficient filter. , The voltage component is the input to the adaptive complex coefficient filter. Represents the imaginary unit. Represents the Laplace operator and its transfer function. For a bandpass / notch filter, its cutoff frequency is... Accepts positive sequence components of the frequency converter power grid and rejects negative sequence components; transfer function For a bandpass / notch filter, its cutoff frequency is... The negative sequence component of the frequency converter is allowed, while the positive sequence component is blocked; the passband center frequency is The sign of the passband center frequency is consistent with the extracted positive and negative sequence components, and the notch frequency is the opposite of the passband center frequency, thus realizing the extraction and separation of the positive and negative sequence components of the frequency conversion power grid. Step 2: Apply the adaptive complex coefficient filter described in Step 1 to a phase-locked loop, and linearize the phase-locked loop using the adaptive complex coefficient filter to obtain the transfer function: （3） Combining the characteristic polynomial in the transfer function with the unit step response of the third-order system in the s-domain, we obtain: （4） in: , These are the proportional and integral coefficients of the PI controller in the phase-locked loop structure. These are coefficients used to determine the actual pole locations based on the natural frequency values. The frequency of undamped natural oscillation; The cutoff frequency can be determined from the transfer function of the adaptive complex coefficient filter in step 1. The damping factor of the transfer function is determined; the optimal damping coefficient for a second-order system is... In an adaptive complex coefficient filter, the cutoff frequency is... Corresponding to , The grid frequency output by the phase-locked loop in a variable frequency power grid; Expanding the right side of equation (4) and corresponding coefficients of the left side, we can obtain: （5） （6） （7）。 2. The variable frequency power grid state estimation method based on an adaptive complex coefficient filter according to claim 1, characterized in that, By combining equations (5) and (6), we can obtain: （8） Based on the discriminant of a quadratic function and the optimal damping ratio We can obtain: （9） From the discriminant of the quadratic function in equation (8), we can see that... Unaffected Substituting the value of into equations (5)-(7) above, we can obtain the cutoff frequency of the adaptive complex coefficient filter phase-locked loop. Adaptive PI controller proportional coefficient and integral coefficient The grid frequency output by the phase-locked loop under the variable frequency power grid The relationship is as follows: （10） （11） （12） （13） The grid frequency output by the phase-locked loop under the variable frequency power grid is obtained from equations (11)-(13). The relationship between the parameters of the adaptive complex coefficient filter phase-locked loop model and the parameters of the phase-locked loop, where the cutoff frequency of the adaptive complex coefficient filter phase-locked loop is... It will vary with the grid frequency output by the phase-locked loop. The bandwidth of the adaptive complex coefficient filter phase-locked loop model changes accordingly; the grid frequency output by the phase-locked loop also changes. As the changes occur, the proportional gain of the PI controller in the phase-locked loop also changes. and integral coefficient The value also changes adaptively, thus changing the locking range of the adaptive complex coefficient filter phase-locked loop. Value: （14） Input frequency variation (frequency variation of mains voltage) ,make Falling within the system's fast dynamic response range , For the frequency of the variable frequency power grid; The positive-sequence voltage component extracted by the adaptive complex coefficient filter is input into the PI adaptive phase-locked loop. The input of the voltage-controlled oscillator in the phase-locked loop, plus the output of the integrator in the PI adaptive controller, provides an estimate of the grid frequency. The phase estimate is then obtained through the integrator, which is the detection of the frequency and phase of the positive-sequence component of the frequency conversion grid.

3. The variable frequency power grid state estimation method based on an adaptive complex coefficient filter according to claim 2, characterized in that, In step 2 It is a coefficient that determines the actual pole location based on the natural frequency value. The coefficient aims to locate the actual pole as close to the origin as possible based on the natural frequency value. Take the minimum possible value This is to maintain the high filtering capability of the adaptive complex coefficient filter phase-locked loop.

4. A system based on the variable frequency power grid state estimation method based on an adaptive complex coefficient filter as described in any one of claims 1 to 3, characterized in that, include: Adaptive complex coefficient filter module: used to extract and separate the positive and negative sequence components of the frequency conversion power grid according to the method in step 1; PI adaptive phase-locked loop module: used to accurately estimate the frequency and phase information of the positive sequence component of the frequency conversion power grid according to the method in step 2.

5. A device based on the variable frequency power grid state estimation method based on an adaptive complex coefficient filter as described in any one of claims 1 to 3, characterized in that, include: Memory, used to store computer programs; A processor, configured to implement the variable frequency power grid state estimation method based on an adaptive complex coefficient filter as described in any one of claims 1 to 3 when executing the computer program.

6. A computer-readable storage medium for storing a computer program, characterized in that, When the computer program is executed by a processor, it is capable of performing variable frequency power grid state estimation based on an adaptive complex coefficient filter according to any one of claims 1 to 3.

Images