A method and device for rapid measurement of complex transient frequencies in microgrids

Abstract

A rapid measurement method and device for complex transient frequencies in microgrids are disclosed, relating to the field of microgrid transient measurement. To address the shortcomings of existing microgrid methods, such as insufficient harmonic robustness, rigid computing power allocation, and unbalanced transient response, this invention suppresses harmonic amplitude aliasing using a low-pass filter and compensates for harmonics. Subsequently, the signal fluctuation intensity of the frequency measurement task is evaluated in real time using the second-order differential rate of change of voltage R, dynamically selecting the calculation mode. Data from the current point and adjacent points are selected, and Taylor expansion is performed. A dual-buffer synchronization technique is employed, allowing the frequency measurement task and control task to access different buffers, achieving conflict-free synchronization through an interrupt flag. Multi-channel redundancy verification is used to write the results into the dual buffers and notify other real-time tasks. This invention is primarily used for measuring complex transient frequencies in microgrids.

Description

Technical Field

[0001] This invention relates to the field of transient measurement of microgrids, and in particular to a method and apparatus for rapid measurement of complex transient frequencies in microgrids. Background Technology

[0002] With the rapid popularization of distributed energy resources, new power systems, represented by microgrids, have placed differentiated demands on the real-time performance and anti-interference capabilities of frequency measurement. Traditional power systems have relatively lenient requirements for frequency sampling rates. Taking 2kHz as an example (40 points per cycle, sampling period 0.5ms), the frequency changes in its steady-state operation are gradual (ROCOF < 0.5Hz / s) and the harmonic content is low (THD < 10%). Existing frequency measurement schemes, with mature filtering algorithms and dedicated hardware, can meet basic control requirements. However, in new small-to-medium power microgrids, the frequent switching of distributed power sources generates a large number of complex transient processes, exposing multiple adaptability challenges to traditional 2kHz sampling rate frequency measurement techniques, specifically as follows:

[0003] 1. Steady-state optimization algorithms are difficult to adapt to transient high-frequency changes;

[0004] Traditional frequency measurement algorithms (such as fixed-window interpolation) are designed for steady-state signals. However, during transient processes in microgrids (such as rapid charging and discharging of energy storage systems, where ROCOF can reach 5Hz / s), the 3-point interpolation method at a 2kHz sampling rate suffers from insufficient data window information, leading to a significant increase in frequency tracking error with increasing rate of change. Real-world measurements in a photovoltaic microgrid show that using traditional methods, frequency measurement delay reaches 1.5ms in islanded mode, causing PCS power adjustment lag and resulting in voltage fluctuations exceeding the national standard limit (±10%). The probability of ( ) increases by 30%.

[0005] 2. Traditional anti-aliasing designs cannot cope with new harmonic scenarios;

[0006] Traditional measurement and control devices typically set their anti-aliasing filter cutoff frequency to 500Hz, suitable for scenarios where power frequency signals are dominant. However, in new microgrids, characteristic harmonics such as 250Hz and 350Hz (amplitudes up to 15% of the fundamental frequency) generated by photovoltaic inverters and power electronic equipment are prone to aliasing distortion after sampling at 2kHz. Field tests on a wind power microgrid showed that when the harmonic distortion rate (THD) was 20%, the frequency error of the traditional 5-point interpolation method exceeded ±0.25Hz, exceeding the microgrid transient measurement standard (±0.1Hz) by 2.5 times, directly causing the failure rate of synchronous devices to rise to 18%.

[0007] 3. Low-amplitude signal processing has blind spots in scene adaptation;

[0008] Traditional frequency measurement procedures assume that the line voltage amplitude is within the normal operating range of above 50V. However, in new microgrids, under conditions such as single-phase ground faults and sudden load drops, the line voltage often drops below 10V, and the zero-crossing density increases by 2-3 times compared to steady state. This causes traditional calculation formulas to become invalid because the denominator approaches zero. A data center's backup microgrid once misjudged the frequency as 45Hz during a main grid failure due to this issue, triggering a false disconnection of the backup power supply and causing a cumulative 10 minutes of data processing interruption.

[0009] 4. The computing power bottleneck of low-cost hardware platforms has become prominent;

[0010] To control costs, new microgrids generally adopt a hardware combination of a general-purpose mid-cost processor (72MHz) and a 16-bit ADC (total cost < $15). However, traditional frequency measurement algorithms are not optimized for the fixed-point arithmetic characteristics of such chips. The single-cycle calculation time of the 4th-order Taylor expansion reaches 200μs, consuming more than 60% of the MCU's computing power. This results in a response delay of more than 50μs for real-time tasks such as fault protection and communication protocols, which cannot meet the real-time requirements of rapid microgrid control (such as power synchronization adjustment during grid-connected / off-grid switching). At the same time, the significant computing power consumption can have unpredictable impacts on control tasks.

[0011] The increasing transient complexity and decreasing hardware cost of new power systems have led to a shift in traditional frequency measurement techniques at 2kHz sampling rates from being applicable in steady state to failing in transient state. Existing solutions, limited by the inadequacy of adaptability to "fixed algorithms + general-purpose hardware," suffer from deficiencies in computing power allocation, anti-aliasing, and low-amplitude adaptability, which have become key technical bottlenecks for the stable operation of microgrids.

[0012] Therefore, there is a need for a rapid measurement method and device for complex transient frequencies in microgrids that can adaptively allocate computing power, has high anti-aliasing capability, and strong adaptability to low amplitude. Summary of the Invention

[0013] To address the shortcomings of existing microgrids, such as insufficient harmonic robustness, rigid computing power allocation, and unbalanced transient response, this invention provides a rapid measurement method and device for complex transient frequencies in microgrids that features adaptive computing power allocation, high anti-aliasing capability, and strong low-amplitude adaptability.

[0014] The present invention provides a method for rapid measurement of complex transient frequencies in microgrids, comprising the following steps:

[0015] S1, Composite anti-aliasing treatment:

[0016] Harmonic amplitude aliasing is suppressed by using a low-pass filter, and harmonics are compensated.

[0017] S2. Transient Feature Extraction:

[0018] The signal fluctuation intensity of the frequency measurement task is evaluated in real time by the second-order differential rate of change of voltage R, and the calculation mode is dynamically selected.

[0019] S3, Dynamic Hierarchical Calculation:

[0020] Based on the calculation mode, select the data of the current point and adjacent points, and perform Taylor expansion;

[0021] S4. Multi-task collaborative scheduling:

[0022] A dual-buffer synchronization technique is adopted, in which frequency measurement tasks and control tasks access different buffers respectively, and conflict-free synchronization is achieved through interrupt flags.

[0023] Further: In S1, the specific steps of the composite anti-aliasing treatment are as follows:

[0024] Hardware-level filtering: Deploy a low-pass filter between the PT sensor and the AD converter in the measurement device to suppress harmonic amplitude;

[0025] Software-level filtering: Perform cubic spline interpolation on the sampled values ​​to generate virtual sampled data;

[0026] Harmonic frequency compensation: Add a harmonic correction term to the Taylor expansion;

[0027] S5. Result Output and Verification:

[0028] The results are written to a double buffer through multi-channel redundancy verification and then notified to other real-time tasks.

[0029] Further: In S2, the real-time evaluation of signal fluctuation intensity based on the second-order differential rate of change of voltage R specifically includes: dynamically selecting the calculation mode: if the second-order differential rate of change of voltage R is less than or equal to a preset threshold, it is determined to be a smooth transient; if the second-order differential rate of change of voltage R is greater than the preset threshold, it is determined to be a severe transient.

[0030] Further: In S3, the specific steps of the dynamic hierarchical calculation are as follows:

[0031] Smooth transients: A 3-point lightweight computation mode is adopted, selecting the current point and the previous two points of data, and performing a second-order Taylor expansion;

[0032] Severe transients: A 5-point high-precision calculation mode is adopted, selecting the current point and two points before and after it, performing a 4th-order Taylor expansion and combining it with fundamental frequency pre-estimation optimization.

[0033] Furthermore: the three-point lightweight computing mode is as follows:

[0034] Sampling point selection: Select the current point Compared with the previous two points , The data window is 1.5ms.

[0035] The second-order Taylor expansion formula is:

[0036] ;

[0037] To predict the next point.

[0038] Furthermore: the 5-point high-precision calculation mode is as follows:

[0039] Sampling point selection: Current point and 2 points before and after. , , , The data window is 2.5ms.

[0040] Optimization of 4th-order Taylor expansion:

[0041] Introducing fundamental frequency prediction Narrowing the range of polynomial variables to .

[0042] Furthermore, in S4, the multi-task collaborative scheduling also includes dynamic allocation of computing power, assigning variable priorities to frequency measurement tasks through the processor task scheduler, specifically including:

[0043] Smooth transient state: Priority is reduced to the lowest level, freeing up computing power for the control task;

[0044] Severe transient: Priority is increased to the highest level to ensure exclusive use of computing resources and improve task switching speed.

[0045] Furthermore, the multi-channel redundancy check specifically involves: if the deviation of the three-phase frequency measurement result is greater than a preset threshold, a recalculation is triggered; if the deviation of the three-phase frequency measurement result is less than the preset threshold, the redundancy check is passed.

[0046] The measuring device disclosed in this invention for implementing the aforementioned method for rapid measurement of complex transient frequencies in a microgrid includes a transient sensing module, a dynamic calculation module, and a multi-task collaborative module.

[0047] The transient sensing module is used to evaluate the operating conditions of the frequency measurement task in real time through the second-order differential rate of change of voltage and dynamically select the calculation mode.

[0048] The dynamic calculation module is used to perform calculations on the frequency measurement task according to the selected calculation mode, and switch the hierarchical calculation mode as needed;

[0049] The multi-task collaboration module is used to synchronize the frequency measurement results with the control results through a double buffer, ensuring that the control commands are not delayed.

[0050] The beneficial effects of this invention are:

[0051] This invention addresses the specific needs of novel power systems. Without increasing hardware costs or sampling rates, it achieves a breakthrough in transient frequency measurement performance on the same hardware platform through adaptive algorithm reconstruction of the dynamic calculation module. This provides core technical support for the efficient access and reliable control of distributed energy resources, demonstrating significant engineering application value and industry versatility. Through condition-aware driven dynamic calculation and hardware-software collaborative anti-aliasing design, it achieves a leap forward in measurement accuracy, computing efficiency, transient response, and hardware adaptability on a general-purpose hardware platform with a 2kHz sampling rate. It effectively solves core technical challenges such as insufficient frequency measurement accuracy in harmonic scenarios of novel microgrids, rigid computing power allocation mechanisms, and imbalances in multiple transient response indicators, combining technological innovation with engineering practicality.

[0052] This invention achieves a generational improvement in frequency measurement performance through dynamic algorithm and hardware co-optimization without significantly increasing costs. The frequency compensation algorithm based on harmonic amplitude improves anti-aliasing capability by 40% compared to traditional methods.

[0053] This invention achieves a breakthrough in computing efficiency, measurement accuracy, and transient response on a low-cost platform with a 2kHz sampling rate for the first time through dynamic algorithm reconstruction and hardware co-optimization. It fills the industry gap in new microgrid frequency measurement technology, provides core support for efficient access and reliable control of distributed energy, and has significant engineering application value and technological innovation significance. Attached Figure Description

[0054] Figure 1 This is a schematic diagram of the frequency measurement architecture. Detailed Implementation

[0055] The following are merely preferred embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention. The embodiments described below are only for explaining the present invention and should not be construed as limiting the present invention. The scope of protection of the present invention should be determined by the scope of the claims. The embodiments of the present invention are described in detail below. In order to facilitate the description of the present invention and simplify the description, the technical terms used in the specification of the present invention should be interpreted broadly, including but not limited to conventional alternatives not mentioned in this application, and including both direct and indirect implementation methods.

[0056] Example 1

[0057] Combination Figure 1 This embodiment describes a method for rapid measurement of complex transient frequencies in a microgrid, comprising the following steps:

[0058] S1, Composite anti-aliasing treatment:

[0059] Harmonic amplitude aliasing is suppressed by using a low-pass filter, and harmonics are compensated.

[0060] The anti-aliasing scheme combining hardware filtering and software interpolation is achieved through the integrated design of front-end low-pass filtering and ADC, and the software uses a software correction method to generate virtual sampling points using cubic spline interpolation.

[0061] It is compatible with low-cost hardware and general-purpose processors (such as STM32 / DSP) and fixed-point arithmetic optimization algorithms of 16-bit ADC. The cost of a single module is less than 800 yuan, which reduces the hardware cost by 90% compared with dedicated chip solutions, filling the gap in low-cost and high-performance frequency measurement technology for small and medium power microgrids.

[0062] Hardware-level filtering: An 8th-order Butterworth low-pass filter is deployed between the PT sensor and the AD converter of the measurement device, with a cutoff frequency of... Transfer function:

[0063] ;

[0064] Where S is the amplitude frequency domain variable;

[0065] The attenuation of 250Hz harmonics reaches 30dB, suppressing their amplitude to less than 5% of the fundamental frequency.

[0066] Software-level filtering: Cubic spline interpolation is performed on the 5 sampled values ​​to generate 10 virtual sampled data points (equivalent to a 4kHz sampling rate), and a harmonic frequency compensation term is introduced to further suppress the influence of higher harmonics. By virtually expanding the 5-point sampling to 10 points (equivalent to a 4kHz sampling rate) through cubic spline interpolation, and in conjunction with the harmonic frequency compensation term, a frequency measurement accuracy of ±0.1Hz is achieved when the harmonic distortion rate (THD) is 20%, breaking through the anti-aliasing limit of a 2kHz sampling rate.

[0067] Interpolation formula:

[0068]

[0069] in The coefficients are the piecewise polynomial coefficients, which are solved by least squares fitting.

[0070] Add 3rd and 5th harmonic correction terms to the Taylor expansion:

[0071]

[0072] coefficient , The harmonic amplitude is dynamically calculated based on the first 5 cycles.

[0073] This embodiment breaks away from the traditional model of "high-performance frequency measurement relying on dedicated chips," and is compatible with mainstream low-cost processors and ADCs. Adapted to general-purpose processors, through fixed-point arithmetic optimization and instruction pipeline design, the cost of a single frequency measurement module is less than 800 yuan (90% lower than the FFT solution); utilizing the ADC's built-in PGA (such as the ×16 gain of the ADS1115) to achieve a wide input range of 0.1V-20V, eliminating the need for additional signal conditioning circuitry and reducing hardware complexity by 30%.

[0074] Breaking through the harmonic suppression limit at a 2kHz sampling rate, this method meets the high-precision frequency measurement requirements of scenarios with a harmonic distortion rate (THD) of 20%. A 900Hz cutoff frequency LC filter is deployed at the hardware front end to attenuate the 250Hz harmonic by 30dB, weakening the source of aliasing distortion. In software, cubic spline interpolation (virtual 10 points) is performed on the 5-point sampled values, effectively increasing the sampling rate to 4kHz. Testing with a microgrid hardware-in-the-loop simulation system shows that the aliasing error decreased from 0.8% to 0.3%, and the total error at a THD of 20% is ≤ ±0.1Hz. A harmonic frequency compensation term (based on zero-crossing initial frequency estimation) is introduced into the Taylor expansion to further suppress spectral leakage caused by higher harmonics, specifically suppressing the influence of the 3rd and 5th harmonics, improving anti-aliasing capability by 40% compared to traditional solutions.

[0075] S2. Transient Feature Extraction:

[0076] The signal fluctuation intensity is evaluated in real time using the second-order differential rate of change R of the voltage; two calculation modes are intelligently switched.

[0077] Based on the operating condition judgment logic of the second-order differential rate of change of voltage R, the formula for calculating the second-order differential rate of change of voltage R is as follows:

[0078] ;

[0079] in, The voltage value at the kth sampling point. The voltage value at the (k+1)th sampling point. The voltage value at the (k-1)th sampling point. The sampling period is ΔT = 0.5 ms.

[0080] In this embodiment, the second-order differential rate of change of voltage R is used to determine the drastic voltage change at the current sampling time. The distribution of R values ​​is shown in Table 1 after simulating different transient scenarios of microgrid operation using MATLAB Simulink hardware-in-the-loop simulation.

[0081] Table 1: Distribution of R-values;

[0082]

[0083] Un is the system rated voltage.

[0084] Threshold filtering logic:

[0085] The midpoint between the maximum R value of the smooth transient (0.25V / ms2) and the minimum R value of the severe transient (0.35V / ms2), i.e., 0.3V / ms2, is selected to ensure reliable differentiation between the two types of operating conditions.

[0086] For power consumption scenarios with drastic load changes or broad requirements for transient characteristics, the R value can be changed according to actual needs.

[0087] The real-time evaluation of signal fluctuation intensity based on the second-order differential rate of change of voltage R specifically includes: dynamically selecting the calculation mode: if the second-order differential rate of change of voltage R is less than or equal to a preset threshold, it is determined to be a smooth transient; if the second-order differential rate of change of voltage R is greater than the preset threshold, it is determined to be a severe transient.

[0088] When the preset threshold is 0.3, the calculation mode is dynamically selected based on the second-order differential rate of change of voltage R: when When, it is determined to be a smooth transient state; when At that time, it was determined to be a violent transient;

[0089] Switching threshold:

[0090] During smooth transients (such as gradual load changes): enable 3-point sampling + 2nd order Taylor series expansion; reduce computing power utilization to 25%; ensure that the processor runs 3 real-time tasks simultaneously (such as control tasks, fault detection, and communication).

[0091] During severe transients (such as grid-connected / off-grid switching): 5-point sampling + 4th-order Taylor expansion is enabled; the Taylor expansion is optimized by combining fundamental frequency pre-estimation with hardware multiplication acceleration (such as DSP MAC instructions) to control the error at harmonic distortion rate (THD) of 20% within ±0.1Hz, and the computing power utilization is stabilized within 50%. This ensures that frequency measurement tasks, control tasks, fault detection, and other real-time tasks run in parallel, with multi-task timing deviation <50μs, meeting the stringent requirements of microgrid controllers for computing power allocation.

[0092] Intelligent allocation of computing power driven by transient intensity;

[0093] A transient adaptive computing power allocation mechanism was constructed to resolve the contradiction between computing power waste and computing power overload caused by fixed algorithms, and to realize intelligent scheduling of processor resources.

[0094] S3, Dynamic Hierarchical Calculation:

[0095] To address the issues of lag in frequency response at zero crossings and poor dynamic adaptability of Taylor expansion, a wide ROCOF range of 0.1-5 Hz / s is covered. Data from the current point and adjacent points are selected, and Taylor expansion is performed.

[0096] When the transient state is smooth ( When using a 3-point lightweight computing mode, the response time is <1.5ms, which is 6 times faster than the zero-crossing scheme (10ms), achieving a 40% improvement in computing efficiency. By selecting the current point and the previous two points (1.5ms data window) and performing a second-order Taylor expansion, the computing power consumption is reduced to 25%.

[0097] Sampling point selection: Current point Compared with the previous two points , The data window is 1.5ms.

[0098] The second-order Taylor expansion formula is:

[0099] ;

[0100] To predict the next point, it is obtained through linear extrapolation;

[0101] Computing power optimization: Absolute value operations and higher-order multiplications are omitted, the number of calculation instructions per cycle is less than 200, and the computing power usage is ≤25%.

[0102] When violent transient ( When the frequency changes rapidly, the tracking error is controlled within ±0.1Hz, in conjunction with the harmonic frequency compensation term, to meet the control margin requirements of grid-connected and off-grid switching (200ms cycle) and achieve stable frequency measurement under a wide range of transient conditions.

[0103] Select the current point and two points before and after it (2.5ms data window), perform a 4th-order Taylor expansion and combine it with fundamental frequency prediction optimization, and keep the computing power usage within 50%.

[0104] 5-point high-precision calculation mode (violent transient):

[0105] Sampling point selection: Current point and 2 points before and after. , , , The data window is 2.5ms.

[0106] Optimization of 4th-order Taylor expansion:

[0107] Introducing fundamental frequency prediction (Based on results from the previous period), the range of polynomial variables is narrowed to... This reduces the number of iterations and lowers the computation time by 30% compared to traditional methods.

[0108] S4. Multi-task collaborative scheduling:

[0109] The dual-buffer (A / B) synchronization technology is adopted, and the frequency measurement task and the control task access different buffers respectively, and the conflict-free synchronization is achieved through the interrupt flag.

[0110] A "frequency measurement result buffer A / B" is set up, and the frequency measurement task and the control task access different buffers respectively. Synchronization is achieved through the interrupt flag (FLAG_FREQ_READY) to eliminate critical section contention.

[0111] Dynamically adjust task priority:

[0112] During smooth transients, reduce the priority of frequency measurement tasks to free up computing power; during severe transients, increase the priority of frequency measurement and control tasks to ensure exclusive use of computing resources.

[0113] Assign variable priorities to frequency measurement tasks using a processor task scheduler (such as FreeRTOS):

[0114] Smooth transient state: Priority is reduced to the lowest level (releasing computing power to the control task);

[0115] Severe transient: Priority is increased to the highest level (ensuring exclusive use of computing resources), and task switching latency is <10μs.

[0116] The dual-buffer architecture and dynamic priority scheduling strategy is a conflict-free synchronization method for frequency measurement results using dual buffers; the task priority dynamic adjustment technology based on transient intensity controls the timing deviation of multiple tasks within 50μs.

[0117] S5. Result Output and Verification:

[0118] After ensuring the reliability of the results through multi-channel redundancy verification (recalculation is triggered when the deviation of the three-phase frequency measurement result is >0.5Hz), the results are written into the buffer and other real-time tasks are notified.

[0119] Example 2

[0120] This embodiment, in conjunction with Example 1, describes the specific implementation steps of a rapid measurement method for complex transient frequencies in a microgrid.

[0121] Signal acquisition phase:

[0122] The ADC samples the three-phase line voltages at a rate of 2kHz and stores them in a circular buffer (depth of 10 points).

[0123] The transient intensity detection module is triggered every 3 sampling points (1.5ms).

[0124] Operating condition assessment stage:

[0125] Calculate the second-order differential rate of change R of the voltage at the current 3 points, and determine whether to use the 3-point / 5-point calculation mode;

[0126] If it is a 5-point mode, add the first 2 points of historical data to the calculation window.

[0127] Frequency calculation stage:

[0128] 3-point mode: Performs a 2nd-order Taylor expansion, taking <80μs;

[0129] 5-point mode: Performs 4th order Taylor expansion + harmonic compensation, time <150μs;

[0130] Both modes use cubic spline interpolation to correct for aliasing at the sampling points.

[0131] Result output stage:

[0132] Multi-channel redundancy check (recalculation triggered when the deviation of the three-phase frequency measurement result is >0.5Hz);

[0133] The results are written to a double buffer and controlled by interrupt notifications for tasks such as fault detection.

[0134] Comparison of key technical parameters in MATLAB simulation:

[0135]

[0136] Example 2

[0137] Combination Figure 1 This embodiment describes a measuring device for implementing a rapid measurement method for complex transient frequencies in a microgrid, comprising a transient sensing module, a dynamic calculation module, and a multi-task collaborative module.

[0138] The transient sensing module is used to evaluate the operating conditions of the frequency measurement task in real time through the second-order differential rate of change of voltage and dynamically select the calculation mode.

[0139] The dynamic calculation module is used to perform calculations on the frequency measurement task according to the selected calculation mode, and switch the hierarchical calculation mode as needed;

[0140] The multi-task collaboration module is used to synchronize the frequency measurement results with the control results through a double buffer, ensuring that the control commands are not delayed.

Claims

1. A method for rapid measurement of complex transient frequencies in a microgrid, characterized in that, Includes the following steps: S1, Composite anti-aliasing treatment: Harmonic amplitude aliasing is suppressed by using a low-pass filter, and harmonics are compensated. S2. Transient Feature Extraction: The signal fluctuation intensity of the frequency measurement task is evaluated in real time by the second-order differential rate of change of voltage R, and the calculation mode is dynamically selected. In S2, the real-time evaluation of signal fluctuation intensity based on the second-order differential rate of change of voltage R specifically includes: dynamically selecting the calculation mode: if the second-order differential rate of change of voltage R is less than or equal to a preset threshold, it is determined to be a smooth transient; if the second-order differential rate of change of voltage R is greater than the preset threshold, it is determined to be a severe transient. The formula for calculating the second-order differential rate of change R of voltage: ； in, The voltage value at the kth sampling point. The voltage value at the (k+1)th sampling point. The voltage value at the (k-1)th sampling point. The sampling period; S3, Dynamic Hierarchical Calculation: Based on the calculation mode, select the data of the current point and adjacent points, and perform Taylor expansion; In S3, the specific steps of the dynamic hierarchical calculation are as follows: Smooth transients: A 3-point lightweight computation mode is adopted, selecting the current point and the previous two points of data, and performing a second-order Taylor expansion; Severe transients: A 5-point high-precision calculation mode is adopted, selecting the current point and two points before and after it, performing a 4th-order Taylor expansion and combining it with fundamental frequency pre-estimation optimization; The three-point lightweight computing mode is as follows: Sampling point selection: Select the current point Compared with the previous point 1 And the next point The data window is 1.5ms. The formula for calculating the current frequency is: ； The 5-point high-precision calculation mode is as follows: Sampling point selection: Current point and 2 points before and after. , , , The data window is 2.5ms; optimized with 4th-order Taylor expansion: Introducing fundamental frequency prediction Narrowing the range of polynomial variables to x represents the difference between the actual frequency f and the estimated fundamental frequency. The ratio of S4. Multi-task collaborative scheduling: A dual-buffer synchronization technique is adopted, in which frequency measurement tasks and control tasks access different buffers respectively, and conflict-free synchronization is achieved through interrupt flags.

2. The method for rapid measurement of complex transient frequencies in a microgrid according to claim 1, characterized in that, In S1, the specific steps of the composite anti-aliasing treatment are as follows: Hardware-level filtering: Deploy a low-pass filter between the PT sensor and the AD converter in the measurement device to suppress harmonic amplitude; Software-level filtering: Perform cubic spline interpolation on the sampled values ​​to generate virtual sampled data; Harmonic frequency compensation: Add a harmonic correction term to the Taylor expansion; S5. Result Output and Verification: The results are written to a double buffer through multi-channel redundancy verification and then notified to other real-time tasks.

3. The method for rapid measurement of complex transient frequencies in a microgrid according to claim 1, characterized in that, In S4, the multi-task collaborative scheduling also includes dynamic allocation of computing power, which assigns variable priorities to frequency measurement tasks through the processor task scheduler, specifically including: Smooth transient state: Priority is reduced to the lowest level, freeing up computing power for the control task; Severe transient: Priority is increased to the highest level to ensure exclusive use of computing resources and improve task switching speed.

4. The method for rapid measurement of complex transient frequencies in a microgrid according to claim 2, characterized in that, The multi-channel redundancy check specifically involves: if the deviation of the three-phase frequency measurement result is greater than a preset threshold, a recalculation is triggered; if the deviation of the three-phase frequency measurement result is less than the preset threshold, the redundancy check is passed.

5. A measuring device for implementing a rapid measurement method for complex transient frequencies in a microgrid as described in any one of claims 1-4, characterized in that, It includes a transient sensing module, a dynamic computing module, and a multi-task collaboration module; The transient sensing module is used to evaluate the operating conditions of the frequency measurement task in real time through the second-order differential rate of change of voltage and dynamically select the calculation mode. The dynamic calculation module is used to perform calculations on the frequency measurement task according to the selected calculation mode, and switch the hierarchical calculation mode as needed; The multi-task collaboration module is used to synchronize the frequency measurement results with the control results through a double buffer, ensuring that the control commands are not delayed.

Images