A new energy station rotational inertia breathing simulation test method and system

Abstract

The application relates to the technical field of new energy power system testing, and discloses a new energy station rotational inertia breathing simulation testing method and system. The method introduces an adaptive closed-loop signal adjustment mechanism based on response feedback on the basis of periodic frequency change rate signal injection and millisecond-level data acquisition, adjusts the amplitude and period of a test signal in real time through a proportional-integral strategy and a sufficiency criterion, and sets a safety boundary constraint; meanwhile, a multi-dimensional response quality evaluation framework including waveform fidelity scoring, response symmetry index and time delay consistency index is constructed, and abnormal mode recognition and degradation classification diagnosis are realized by combining weighted fusion scoring and a sliding window statistical process control. Adaptive signal adjustment and multi-dimensional quality evaluation form a closed-loop cooperation, and the adaptability of the test to different types of stations and the accuracy of inertia response performance evaluation are improved.

Description

Technical Field

[0001] This invention relates to the field of testing technology for new energy power systems, and in particular to a method and system for simulating the breathing motion of rotational inertia of new energy power stations. Background Technology

[0002] With the increasing penetration of wind power, photovoltaics, and new energy storage devices in power systems, the rotational inertia response capability of renewable energy power plants is gradually becoming an important factor affecting grid frequency stability. Since renewable energy units generally use power electronic converters for grid connection, they do not possess a natural rotational inertia comparable to traditional synchronous generators. Therefore, virtual inertia control technology is typically needed to simulate inertia response behavior. Accurate assessment of rotational inertia response capability has become a crucial aspect of grid connection acceptance and ancillary service evaluation for renewable energy power plants.

[0003] Currently, the testing and evaluation of the rotational inertia response capability of renewable energy power plants mainly relies on passive observation of the plant's response process after a real frequency disturbance occurs in the power grid. However, due to the randomness of the timing and amplitude of real frequency disturbances in the power grid, and the differences in plant operating conditions across different disturbance events, the comparability and repeatability of test results obtained through passive observation are insufficient. To address these issues, existing technical solutions propose actively constructing periodic frequency change rate signals for inertia response testing without relying on real frequency disturbances in the power grid, and provide basic methods for signal construction, data acquisition, and calculation of response capability indicators. However, this type of solution still faces the following shortcomings in practical engineering applications: On the one hand, the amplitude and period of the test signal are set to fixed values ​​that are preset manually. Different types of new energy power plants vary significantly in terms of virtual inertia control algorithms, hardware response bandwidth, and adjustable capacity. Test signals with fixed parameters may cause some power plants to trigger protection actions due to excessive excitation, while other power plants may have insufficient response characteristics due to insufficient excitation. The effectiveness and discriminative power of the test conclusions are limited, and there is a lack of a closed-loop mechanism that can adaptively adjust the signal parameters according to the real-time response status of the power plant during the test. On the other hand, existing solutions mainly evaluate the response capability index and time-domain amplitude indexes such as linear attenuation coefficient. The evaluation dimension is relatively simple and can only reflect the degree of compliance of the response amplitude. It cannot reveal whether there are degradation modes such as oscillation, delay drift, and asymmetric distortion in the response waveform. It is difficult for operation and maintenance personnel to determine which specific link in the inertia control logic has experienced performance degradation, which is not conducive to accurate diagnosis and targeted optimization. Summary of the Invention

[0004] The purpose of this invention is to provide a method and system for simulating the breathing of rotational inertia at new energy power stations, in order to solve the problems of insufficient adaptability to different types of power stations due to fixed test signal parameters in existing rotational inertia testing schemes, and the inability to accurately locate the root cause of inertia control performance degradation due to a single dimension of response quality assessment.

[0005] To achieve the above objectives, this invention provides a method for simulating the breathing motion of rotational inertia of a new energy power station, comprising the following steps:

[0006] Configure initial test parameters, including initial frequency change rate amplitude, initial test period, and target response capability index value. Generate a rotational inertia breathing simulation test signal based on the signal parameters of the current test period. Inject the test signal into the inertia control interface of the new energy power station, and synchronously acquire the frequency data, active power data, and protection action signals of the grid connection point of the new energy power station at millisecond levels. After the current test period ends, preprocess the acquired data and extract inertia response characteristic parameters, including inertia response start time, maximum active power change, and inertia response duration, and calculate the rotational inertia response capability index accordingly. Perform a multi-dimensional response quality assessment for the current test period, including: calculating the waveform fidelity score based on the normalized cross-correlation coefficient between the actual active power response waveform and the theoretical expected power response waveform, and based on the positive and negative half-cycles of the test signal. The response symmetry index is calculated by comparing the maximum active power change, and the delay consistency index is calculated based on the coefficient of variation of the inertial response start time in multiple test cycles. The rotational inertial response capability index, waveform fidelity score, response symmetry index, and delay consistency index are weighted and fused to obtain a comprehensive response quality score. Anomaly detection is performed on each dimension index based on sliding window statistical process control, and degradation mode classification labels are output. According to the deviation between the rotational inertial response capability index and the target value, the amplitude of the test signal in the next test cycle is adaptively adjusted through a proportional-integral adjustment strategy. The test signal period in the next test cycle is adaptively adjusted according to the sufficiency coefficient formed by the ratio of the inertial response duration to the half-cycle of the test signal, and the adjusted parameters are subjected to safety boundary constraint verification. The above steps are repeated with the updated signal parameters until the planned total number of test cycles is reached or the test is manually terminated by the tester.

[0007] As a preferred embodiment of the present invention, the adaptive adjustment of the test signal amplitude in the next test cycle using a proportional-integral adjustment strategy specifically includes: defining the response capability index deviation. ,in The target response capability index value, For the first The rotational inertia response capability index obtained from actual calculations over each test cycle; according to the formula Calculate the amplitude correction amount, where This is the proportional gain coefficient. This is the integral gain coefficient; update the amplitude for the next cycle to... And apply a hard limiting constraint to the updated amplitude. and rate of change limit constraint This allows the amplitude of the test signal to gradually converge within a safe range to a level that matches the response characteristics of the site.

[0008] As a preferred embodiment of the present invention, the step of adaptively adjusting the test signal period of the next test cycle based on the sufficiency coefficient specifically includes: calculating the sufficiency coefficient. ,in For the first The duration of inertial response per test cycle The current test signal period; when At that time, according to Increase the cycle; when At that time, according to Shorten the cycle; among which Adjust the gain coefficient for the period; apply a hard limit constraint to the updated period. It can automatically match the test cycle length according to the site response time characteristics, avoiding truncation or invalid waiting in the response process.

[0009] As a preferred embodiment of the present invention, the safety boundary constraint verification includes: when a station protection action is triggered in the current test cycle, reducing the amplitude of the next cycle to a preset attenuation ratio of the current amplitude, and locking the amplitude so that it does not increase further, until the protection is not triggered for a consecutive preset number of cycles, after which the value is unlocked; simultaneously checking whether the theoretical active power change corresponding to the test signal in the next cycle meets the requirements. The constraints, among which This is the equivalent inertia control factor. For the rated capacity of the station, This represents the current available active power margin of the power station. This is a safety margin factor. If the value exceeds this constraint, the amplitude will be calculated and truncated to a safe value, ensuring that the test signal does not exceed the safe operating range of the station during the adaptive adjustment process.

[0010] As a preferred embodiment of the present invention, the calculation of the waveform fidelity score includes: based on the nominal equivalent inertia constant of the station. Rated capacity and the frequency change rate signal injected in the current period ,according to Construct the theoretically expected power response waveform; calculate the actual response waveform. Compared with the theoretically expected waveform Normalized cross-correlation coefficients between The waveform fidelity score is defined as follows: , The value ranges from zero to one; the closer it is to one, the better the actual response matches the theoretical expectation.

[0011] As a preferred embodiment of the present invention, the calculation of the response symmetry index includes: […]. Each test cycle is divided into a positive half-cycle and a negative half-cycle according to the polarity of the test signal, and the maximum active power change in the positive half-cycle is extracted respectively. and the maximum active power change during the negative half-cycle ,according to Calculate the response symmetry index; the calculation of the delay consistency index includes: before collection Calculate the coefficient of variation of the inertial response startup time series for each test cycle. ,according to Calculate the latency consistency index, where This is the sensitivity adjustment coefficient.

[0012] As a preferred technical solution of the present invention, the anomaly detection of each dimension index based on sliding window statistical process control includes: maintaining a sliding window of a preset length for each dimension index, calculating the mean and standard deviation within the sliding window, and marking the dimension as abnormal when the value of a certain dimension index deviates from the window mean by more than a preset control limit multiple of the standard deviation in the current test period; outputting degradation mode classification labels according to the combination of abnormal dimensions, including waveform distortion degradation, asymmetric response degradation, time delay drift degradation, inertia capability decay degradation, and comprehensive performance degradation, and outputting corresponding diagnostic suggestions, which can accurately locate the degradation link of inertia control performance based on the abnormal combination of multi-dimensional indexes.

[0013] As a preferred technical solution of the present invention, the method further includes switching between adaptive mode and locking mode: when the fluctuation of the comprehensive response quality score for a consecutive preset number of cycles is less than a preset threshold, the system is automatically switched from adaptive mode to locking mode, the current signal parameters are locked and no longer adjusted, and the system enters the steady-state test stage. The test data in the steady-state test stage is used for final evaluation and report generation. The system can automatically switch to steady-state test after the signal parameters converge, thereby improving the consistency of the final evaluation data.

[0014] This invention also provides a simulation test system for the rotational inertia breathing of a new energy power station, comprising: an adaptive test signal generation module, used to generate a simulation test signal for rotational inertia breathing in each test cycle based on preset initial parameters and closed-loop feedback information; the adaptive test signal generation module includes a signal waveform construction unit, an amplitude adaptive adjustment unit, a cycle adaptive scaling unit, and a safety boundary constraint unit; an inertia response triggering interface module, used to inject the test signal into the inertia control interface of the new energy power station through a communication protocol; the inertia response triggering interface module includes a protocol adaptation unit and a signal injection synchronization unit; and a millisecond-level data acquisition module, used to synchronously acquire frequency data, active power data, and protection action signals at the grid connection point of the new energy power station at the millisecond level; the millisecond-level data acquisition module includes a high-speed... The system includes a synchronous acquisition unit and a data preprocessing unit; a multi-dimensional inertial response quality assessment module, used for multi-dimensional feature extraction, quality scoring, and abnormal pattern identification of the inertial active power response for each test cycle, comprising a basic feature extraction unit, a waveform fidelity scoring unit, a response symmetry assessment unit, a delay consistency assessment unit, and a comprehensive quality scoring and abnormal pattern identification unit; a closed-loop feedback coordination module, used to transmit the evaluation results of the multi-dimensional inertial response quality assessment module to the adaptive test signal generation module to drive the adaptive adjustment of signal parameters, comprising a feedback data packaging unit and an adaptive enable and mode switching unit; and a test report generation module, used to summarize all test cycle data and generate a structured test report after the test is completed.

[0015] As a preferred embodiment of the present invention, the amplitude adaptive adjustment unit employs a proportional-integral adjustment strategy to calculate the amplitude correction based on the deviation of the rotational inertia response capability index, and applies hard amplitude limiting constraints and rate of change amplitude limiting constraints; the period adaptive scaling unit determines whether the current test cycle is sufficient based on the sufficiency coefficient formed by the ratio of the inertia response duration to the half-cycle of the test signal, and adjusts the signal cycle of the next cycle accordingly and applies hard amplitude limiting constraints; the safety boundary constraint unit performs compliance verification on the signal amplitude and cycle before each parameter update, including amplitude reduction locking logic after protection action is triggered and amplitude truncation logic based on available active power margin, which can achieve safe and stable convergence of test signal parameters.

[0016] As a preferred embodiment of the present invention, the waveform fidelity scoring unit obtains the waveform fidelity score by constructing the theoretically expected power response waveform and calculating the normalized cross-correlation coefficient between the actual response waveform and the theoretically expected waveform; the response symmetry evaluation unit calculates the response symmetry index by comparing the maximum active power change in the positive and negative half-cycles of the test signal; the delay consistency evaluation unit obtains the delay consistency index by calculating the coefficient of variation of the inertial response start-up time series of multiple test cycles; the comprehensive quality scoring and anomaly pattern recognition unit calculates the comprehensive response quality score by weighted fusion of multi-dimensional indicators, and applies the sliding window control chart criterion to each dimension indicator for anomaly determination, outputting degradation mode classification labels and diagnostic suggestions based on the combination of anomaly dimensions, thereby achieving an upgrade in evaluation from response amplitude compliance rate to response process quality profile.

[0017] Compared with the prior art, the present invention has the following beneficial effects:

[0018] This invention introduces an adaptive closed-loop signal conditioning mechanism based on real-time response feedback, enabling the amplitude and period of the test signal to automatically adjust according to the current response state of the power station. This avoids the problems of excessive or insufficient excitation caused by fixed parameters, allowing the same test system to be adapted to different types and capacities of new energy power stations, thus improving test effectiveness and cross-station comparability. This invention constructs a multi-dimensional response quality assessment framework including waveform fidelity scoring, response symmetry index, and delay consistency index, achieving an upgrade in assessment from response amplitude compliance rate to response process quality profiling. It can locate degradation modes such as response oscillation, delay drift, and asymmetric distortion, providing a diagnostic basis for optimizing the power station's inertia control logic. Adaptive signal conditioning and multi-dimensional quality assessment form a closed loop. Adaptive signal conditioning dynamically optimizes test conditions based on the multi-dimensional assessment results, while multi-dimensional quality assessment obtains more comprehensive performance differentiation through the multi-condition data created by adaptive conditioning. Together, they improve the engineering adaptability and assessment accuracy of the test scheme. Attached Figure Description

[0019] Figure 1 This is a schematic diagram of the architecture of the new energy power station rotational inertia breathing simulation test system provided in an embodiment of the present invention;

[0020] Figure 2 A flowchart illustrating the method for simulating the breathing motion of a new energy power station in an embodiment of the present invention.

[0021] The module comprises: 100, Adaptive Test Signal Generation Module; 101, Signal Waveform Construction Unit; 102, Amplitude Adaptive Adjustment Unit; 103, Period Adaptive Scaling Unit; 104, Safety Boundary Constraint Unit; 200, Inertia Response Trigger Interface Module; 201, Protocol Adaptation Unit; 202, Signal Injection Synchronization Unit; 300, Millisecond-Level Data Acquisition Module; 301, High-Speed ​​Synchronous Acquisition Unit; 302, Data Preprocessing Unit; 400, Multidimensional Inertia Response Quality Assessment Module; 401, Basic Feature Extraction Unit; 402, Waveform Fidelity Scoring Unit; 403, Response Symmetry Assessment Unit; 404, Delay Consistency Assessment Unit; 405, Comprehensive Quality Scoring and Abnormal Pattern Recognition Unit; 500, Closed-Loop Feedback Coordination Module; 501, Feedback Data Packaging Unit; 502, Adaptive Enable and Mode Switching Unit; and 600, Test Report Generation Module. Detailed Implementation

[0022] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention.

[0023] like Figure 1 As shown, the new energy power station rotational inertia breathing simulation test system provided by this invention includes an adaptive test signal generation module 100, an inertia response trigger interface module 200, a millisecond-level data acquisition module 300, a multi-dimensional inertia response quality assessment module 400, a closed-loop feedback coordination module 500, and a test report generation module 600. The modules establish a collaborative working relationship through data flow and control flow. The output of the adaptive test signal generation module 100 is injected into the power station under test via the inertia response trigger interface module 200. The millisecond-level data acquisition module 300 collects the response data from the power station's grid connection point and transmits it to the multi-dimensional inertia response quality assessment module 400 for analysis. The closed-loop feedback coordination module 500 feeds back the assessment results from the multi-dimensional inertia response quality assessment module 400 to the adaptive test signal generation module 100, forming a complete closed loop of "signal generation—injection—acquisition—assessment—feedback adjustment". The test report generation module 600 summarizes the data and generates a structured report after all test cycles are completed.

[0024] The adaptive test signal generation module 100 is responsible for generating a rotational inertia breathing simulation test signal for each test cycle based on preset initial parameters and closed-loop feedback information. This module internally includes a signal waveform construction unit 101, an amplitude adaptive adjustment unit 102, a cycle adaptive scaling unit 103, and a safety boundary constraint unit 104. The signal waveform construction unit 101 constructs a frequency change rate signal based on the signal parameters of the current cycle; the signal waveform can be sinusoidal.

[0025]

[0026] in, For the first The amplitude of the rate of change of frequency over each test period For the first The signal period of each test cycle. In some test scenarios that require more explicit excitation boundaries, the signal waveform construction unit 101 can also adopt a square wave form:

[0027]

[0028] Sinusoidal signals have a pure spectrum and are suitable for scenarios requiring detailed analysis of frequency response characteristics; square wave signals maintain a constant amplitude in both positive and negative directions and are suitable for scenarios requiring a clear assessment of the station's sustained response capability under stable excitation. Testers can choose the signal waveform type according to their actual testing needs.

[0029] The amplitude adaptive adjustment unit 102 calculates the amplitude correction amount for the next cycle based on the response capability index deviation after each test cycle. This unit defines the response capability index deviation as:

[0030]

[0031] in, The preset target response capability index value typically ranges from 0.85 to 0.95. For the first The rotational inertia response capability index is obtained from actual calculations over one test cycle. The amplitude correction is calculated using a proportional-integral adjustment strategy.

[0032]

[0033] in, This is the proportional gain coefficient. This is the integral gain coefficient. The proportional term provides an immediate response to the current deviation, enabling the amplitude to quickly track changes in the station's response state; the integral term eliminates steady-state bias, ensuring that the amplitude converges after several cycles to a level that brings the response capability index close to the target value. The amplitude is updated for the next cycle. And apply a hard limiting constraint to the updated amplitude:

[0034]

[0035] Simultaneously apply a rate-of-change constraint:

[0036]

[0037] in, and These are the minimum and maximum allowable values ​​for the amplitude of the rate of change of frequency, for example... Hz / s Hz / s, This represents the maximum permissible amplitude change per cycle, for example, 0.05 Hz / s. Hard limiting ensures the signal amplitude remains within the safe range acceptable to the station, while rate-of-change limiting prevents drastic amplitude jumps between adjacent cycles, ensuring smooth adjustment. If the updated amplitude change exceeds the rate-of-change limit, it is truncated to the limit value.

[0038] The periodic adaptive scaling unit 103 determines whether the current test cycle is sufficient based on the station response time characteristics, and adjusts the signal cycle for the next cycle accordingly. This unit calculates the... Duration of inertial response within each test cycle With the current test signal half cycle The ratio is defined as the sufficiency coefficient:

[0039]

[0040] when For example This indicates that the current half-cycle is insufficient to accommodate the complete response process, and the response may be truncated during half-cycle switching. In this case, the cycle needs to be increased.

[0041]

[0042] when For example This indicates that the current half-cycle is much longer than the response time, resulting in a large number of invalid waiting periods. The cycle can be shortened to improve testing efficiency.

[0043]

[0044] in, The gain coefficient is adjusted periodically. A hard limiting constraint is also applied to the updated period.

[0045]

[0046] in, and These are the minimum and maximum allowed values ​​for the test period, for example... s、 The minimum constraint prevents the station inertia control from not having enough time to complete a full response process due to an excessively short cycle, while the maximum constraint prevents the test time from becoming uneconomical due to an excessively long cycle.

[0047] Before each parameter update, the safety boundary constraint unit 104 performs compliance verification on the signal amplitude and period. Its verification logic comprises two levels. At the protection action level, the safety boundary constraint unit 104 checks whether the current active power response has triggered a station protection action, using the protection signal flag bit returned by the millisecond-level data acquisition module 300 for determination. If a protection action is triggered in the current period, the amplitude in the next period is reduced to a preset attenuation ratio of the current amplitude, for example, attenuated to 70% of the current amplitude, and the amplitude is locked to prevent further increase until continuous... Unlocked after each cycle failed to trigger protection, for example At the power margin level, the safety boundary constraint unit 104 checks whether the change in theoretical active power corresponding to the equivalent inertia trigger signal exceeds the current available active power margin of the power station. The criterion is as follows:

[0048]

[0049] in, This is the equivalent inertia control factor. For the rated capacity of the station, This represents the current available active power margin of the power station (obtained in real time by the millisecond-level data acquisition module 300). This is the safety margin factor, typically taken as 0.8. If the calculated result exceeds this constraint, the calculation is reversed and truncated. The test signal must reach a safe value that satisfies this constraint. A dual-layer verification mechanism ensures that the test signal never exceeds the safe operating range of the station during the adaptive adjustment process.

[0050] The inertia response trigger interface module 200 is responsible for inputting the test signal generated by the adaptive test signal generation module 100 to the inertia control interface or virtual inertia control interface of the new energy power station. This module internally includes a protocol adaptation unit 201 and a signal injection synchronization unit 202. The protocol adaptation unit 201 automatically encapsulates the test signal data frame according to the communication protocol type of the target power station control system. Supported communication protocols include, but are not limited to, Modbus TCP, IEC 61850 MMS, and OPCUA. The signal injection synchronization unit 202 sends a precise timestamp to the millisecond-level data acquisition module 300 at the signal injection moment, ensuring millisecond-level alignment between the signal injection moment and the data acquisition moment. This time alignment accuracy is crucial for the accurate extraction of the subsequent inertia response start time.

[0051] The millisecond-level data acquisition module 300 is responsible for synchronously acquiring frequency data, active power data, and protection action signals at the grid connection point of the new energy power station at the millisecond level. This module internally includes a high-speed synchronous acquisition unit 301 and a data preprocessing unit 302. The high-speed synchronous acquisition unit 301 has a sampling rate of no less than 1kHz, and its acquisition channels include grid connection point frequency, active power, reactive power, and power station protection action flags. The reactive power channel is used to assist in analyzing whether reactive power coupling exists during the power station's inertial response process. All acquisition channels achieve millisecond-level synchronization based on a unified GPS / BeiDou timing signal or the IEEE 1588 precise time protocol, ensuring the time consistency of data from each channel. The data preprocessing unit 302 performs outlier removal and low-pass filtering on the raw acquired data. Outlier removal adopts the 3σ criterion, identifying and removing sampling points that deviate from the local mean by more than three standard deviations. The cutoff frequency of the low-pass filter is configurable, with a typical value of 50Hz, used to filter out high-frequency measurement noise. After preprocessing, clean millisecond-level time-series data is output for subsequent analysis.

[0052] The multidimensional inertial response quality assessment module 400 is one of the core innovative modules of this invention, responsible for multidimensional feature extraction, quality scoring, and abnormal pattern recognition of the inertial active power response for each test cycle. This module internally includes a basic feature extraction unit 401, a waveform fidelity scoring unit 402, a response symmetry assessment unit 403, a time delay consistency assessment unit 404, and a comprehensive quality scoring and abnormal pattern recognition unit 405.

[0053] The basic feature extraction unit 401 extracts basic inertial response feature parameters from the millisecond-level time-series data of each test cycle. Inertial response start-up time. Defined as the period from the moment the test signal is injected until the change in active power first exceeds a preset threshold. The time interval, where It can be set to 0.5% of the station's rated capacity. Maximum active power variation. This represents the absolute value of the maximum deviation of active power from the reference value within this test period. Maximum active power change rate. The maximum rate of change of the active power time-series curve within this test period is obtained by numerical differentiation of the filtered power data. Inertial response duration. Defined as the first time the change in active power exceeds a threshold. Until the last drop to the threshold The following time intervals. The basic feature extraction unit 401 also calculates the theoretical active power change based on the injected signal parameters of the current period and the station's rated capacity. Then, the rotational inertia response capability index is calculated. .

[0054] The waveform fidelity scoring unit 402 is used to evaluate the degree of agreement between the actual active power response waveform of the power station and the theoretically expected waveform. It is one of the key innovative sub-modules that distinguishes this invention from existing technologies. This unit first assesses the degree of agreement between the power station's nominal equivalent inertia constant and the actual active power response waveform. Rated capacity Construct the theoretically expected power response waveform using the actual frequency change rate signal injected in the current period:

[0055]

[0056] Then calculate the actual response waveform. Compared with the theoretically expected waveform Normalized cross-correlation coefficients between them:

[0057]

[0058] in, and These are the average values ​​of the actual waveform and the theoretical waveform over the test period, respectively, and the summation range covers all sampling points within the test period. The waveform fidelity score is defined as follows: The value ranges from zero to one; the closer it is to one, the closer the actual response is to the theoretical expectation, and the higher the waveform fidelity. When When the value is lower than the preset threshold (e.g., 0.7), it indicates that there may be abnormal control logic or parameter mismatch at the station.

[0059] The response symmetry evaluation unit 403 is used to evaluate whether the active power response of the station during the positive and negative half-cycles of the frequency change rate signal is symmetrical. The positive half-cycle corresponds to the stage of analog frequency increase, and the negative half-cycle corresponds to the stage of analog frequency decrease. This evaluation is of great significance for identifying whether the station's inertia control capability is consistent in both the frequency increase and decrease directions. This unit will... Each test cycle is divided into positive half-cycles according to the positive and negative polarities of the signal. Phase) and negative half-cycle ( (Stages), extracting the maximum active power change during the positive half-cycle respectively. and the maximum active power change during the negative half-cycle Calculate the response symmetry index:

[0060]

[0061] The value ranges from zero to one; the closer it is to one, the more symmetrical the response is in both positive and negative directions. A value significantly lower than one may indicate that the active power margin of the station is insufficient in a certain direction, or that the inertia control parameters are inconsistent in the frequency increase and decrease directions.

[0062] The latency consistency evaluation unit 404 is used to evaluate the consistency of the inertial response start-up time of the site over multiple consecutive test cycles. This unit collects prior data. Inertial response startup time series for each test cycle Calculate its standard deviation and coefficient of variation:

[0063]

[0064] in, For the front The average startup time of each cycle. The latency consistency index is defined as:

[0065]

[0066] in, This is the sensitivity adjustment coefficient, typically set to 10. The closer the value is to 1, the more consistent the startup latency and the better the response timing stability. A drop in latency may indicate jitter in the control system's communication link or instability in the algorithm's execution cycle.

[0067] The comprehensive quality scoring and anomaly pattern recognition unit 405 integrates the above-mentioned indicators into a comprehensive response quality score, and performs anomaly detection on each dimension based on statistical process control methods. This unit defines a six-dimensional indicator vector:

[0068]

[0069] in, This is an indicator of rotational inertia response capability. Normalized attenuation coefficient ( (After taking the value, subtract one to make the larger the value, the better the performance). As of the date The cumulative correct action rate over the period, The waveform fidelity is scored. In response to the symmetry index, This is the latency consistency index. The overall response quality score is calculated as follows:

[0070]

[0071] in, The weights for each dimension are set, with equal weights by default (1 / 6 for each dimension). Testers can also customize the weights for each dimension according to their specific assessment focus. The value ranges from zero to one, with higher values ​​indicating better overall quality.

[0072] The anomaly pattern recognition algorithm maintains a sliding window for each dimension indicator, with a window length of... The typical value is the average of the last 5 test periods. The mean is calculated for the data within the sliding window. and standard deviation When the first For example, the value of a certain dimension indicator in a test period deviates from the window mean by more than a preset control limit (e.g., When this occurs, the dimension is marked as an anomaly. Based on the combination of anomaly dimensions, the system outputs a degradation mode classification label: if only Abnormal and If the waveform decreases, it is marked as "waveform distortion degradation," and it is recommended to check the inertia control parameter matching; if only Abnormal and If the response decreases, it is marked as "asymmetric response degradation," and it is recommended to check the power margin or control gain consistency of the station in both the positive and negative regulation directions; if only... Abnormal and If the degradation decreases, it is marked as "delay drift-type degradation," and it is recommended to check the control system communication link or algorithm execution timer; if and If all aspects are abnormal, it is marked as "inertia capacity degradation type" and it is recommended to check the energy storage charge status or the available output of wind turbines and photovoltaics; if multiple dimensions are abnormal at the same time, it is marked as "comprehensive performance degradation" and it is recommended to conduct a comprehensive site-level test.

[0073] The closed-loop feedback coordination module 500 is responsible for coordinating the transmission of evaluation results from the multidimensional inertial response quality assessment module 400 to the parameter feedback of the adaptive test signal generation module 100. This module internally includes a feedback data packaging unit 501 and an adaptive enable and mode switching unit 502. At the end of each test cycle, the feedback data packaging unit 501 packages the basic feature parameters extracted by the basic feature extraction unit 401 (especially...) and The waveform fidelity scoring unit 402, response symmetry evaluation unit 403, and delay consistency evaluation unit 404 calculate the following: , , The data is packaged into a structured feedback data packet and transmitted to the adaptive test signal generation module 100. The adaptive enable and mode switching unit 502 controls the start / stop and mode switching of the adaptive adjustment; the system supports two operating modes. In adaptive mode, the adaptive test signal generation module 100 adjusts the signal parameters in real time based on the feedback data. This mode is suitable for initial testing or scenarios where the field performance is unknown. In locked mode, when continuous... One cycle (e.g.) Overall quality score When the fluctuation is less than the preset threshold, the system automatically locks the current signal parameters and stops adjusting, entering the steady-state testing phase. The test data from this phase is used for final evaluation and report generation. The mode switching criterion is:

[0074]

[0075] in, This is the threshold for fluctuation in the overall quality score, typically set at 0.05.

[0076] The test report generation module 600 is responsible for summarizing all test cycle data and generating a structured test report after the test is completed. The report includes adaptive adjustment process curves (reflecting the trajectory of signal parameters from initial to convergent values ​​in each cycle), six-dimensional indicator radar charts for each cycle (visualizing the quality distribution in each dimension), a comprehensive quality score trend curve (reflecting the overall performance of the station's inertial response stability), and a list of abnormal pattern recognition results along with corresponding degradation mode classification labels and diagnostic suggestions. The test report can be output to a higher-level system or stored locally for use in grid connection acceptance, operational evaluation, or ancillary service capability verification.

[0077] like Figure 2 As shown, the complete workflow of the new energy power station rotational inertia breathing simulation test method provided by the present invention is as follows.

[0078] Testers configure initial test parameters, including the initial frequency change rate amplitude, through the test system's human-machine interface. Initial testing period Target response capability indicators Proportional-integral regulator gain and Safety boundary parameters ( , , , , , ), Anomaly detection control limit multiple and total number of planned test cycles The protocol adaptation unit 201 of the inertial response trigger interface module 200 completes interface initialization and link establishment testing according to the communication protocol type of the target site control system. The high-speed synchronous acquisition unit 301 of the millisecond-level data acquisition module 300 completes clock synchronization calibration and confirms that each acquisition channel is working normally. After the system initialization is completed, it enters the ready-to-trigger state.

[0079] After the system enters the first test cycle, the signal waveform construction unit 101 of the adaptive test signal generation module 100 constructs the signal waveform according to the initial parameters. and The rotational inertia breathing simulation test signal for the first test cycle is constructed. After protocol encapsulation by the protocol adaptation unit 201 of the inertia response trigger interface module 200, the signal is injected into the inertia control interface of the station under precise timestamp marking through the signal injection synchronization unit 202. At the same time, the millisecond-level data acquisition module 300 begins to synchronously acquire grid connection point frequency data, active power data, and protection signal flag bits at the millisecond level.

[0080] In the current testing cycle Within the time window, the high-speed synchronous acquisition unit 301 continuously acquires data from each channel of the grid connection point, and the data preprocessing unit 302 performs outlier removal and low-pass filtering on the raw data based on the 3σ criterion, outputting clean millisecond-level time-series data and storing it in the data buffer.

[0081] After the current test cycle ends, the basic feature extraction unit 401 reads the data of that cycle from the data buffer and extracts the inertial response start time. Maximum active power change Maximum active power change rate and inertial response duration The theoretical change in active power is calculated based on the injection signal parameters of the current cycle and the rated capacity of the power station. Then, the rotational inertia response capability index is calculated. .

[0082] After the basic feature extraction is completed, each evaluation unit of the multidimensional inertial response quality assessment module 400 performs evaluation calculations for its respective dimension. The waveform fidelity scoring unit 402 constructs the theoretically expected power response waveform based on the current periodic injected signal and the station's nominal parameters, calculates the normalized cross-correlation coefficient between the actual waveform and the theoretical waveform, and obtains the waveform fidelity score. The response symmetry evaluation unit 403 divides the current period data into positive and negative half-cycles of the signal, extracts the maximum power change in each half-cycle, and calculates the response symmetry index. The latency consistency evaluation unit 404 adds the start time of the current cycle to the historical sequence, calculates the coefficient of variation, and outputs the latency consistency index. The integrated quality scoring and abnormal pattern recognition unit 405 combines the six-dimensional indicators into a vector. The overall quality score is calculated through weighted fusion. At the same time, a sliding window control chart criterion is applied to each dimension to determine whether there are any anomalies. If an anomaly is found, the corresponding degradation mode classification label and diagnostic suggestions are output.

[0083] The feedback data packaging unit 501 of the closed-loop feedback coordination module 500 will... , , , , The protection action flag and other information are packaged into a feedback data packet and transmitted to the adaptive test signal generation module 100. The safety boundary constraint unit 104 of the adaptive test signal generation module 100 first performs a safety boundary check to see if a protection action is triggered in the current cycle. If triggered, the protection amplitude reduction logic is initiated. Subsequently, the amplitude adaptive adjustment unit 102 adjusts the data according to the protection action flag and other information. The deviation is calculated and updated using a proportional-integral control strategy to determine the amplitude correction. Hard limiting and rate-of-change limiting constraints are applied. The periodic adaptive scaling unit 103 is based on the sufficiency coefficient. Determine if the cycle needs adjustment and update. Apply a periodic hard limiting constraint. If the adaptive enable and the mode switching criterion of the mode switching unit 502 are satisfied, i.e., continuous... Overall quality score for each cycle Fluctuation less than the threshold If the signal is switched to locked mode, the signal parameters for subsequent cycles will remain fixed.

[0084] The system uses updated parameters and Entering the next test cycle, repeat the above process of signal generation and injection, data acquisition and preprocessing, feature extraction, multidimensional evaluation, closed-loop feedback and parameter update, until the planned total number of test cycles is reached. Or it can be manually terminated by the tester.

[0085] After all test cycles are completed, the test report generation module 600 summarizes the signal parameter change trajectories, six-dimensional index data, comprehensive quality score sequence, and abnormal pattern recognition results for all cycles, generating a structured test report. The report includes an adaptive adjustment convergence curve, six-dimensional index radar charts for each cycle, a comprehensive quality score trend line, a list of abnormal events, and corresponding degradation mode classification labels and diagnostic suggestions. The test report is output to the host system or stored locally for use in grid connection acceptance, operational evaluation, or ancillary service capability verification.

[0086] It should be noted that the values ​​of various parameters involved in the above embodiments (such as...) The range of values, , , , , , , , , , , , , The values ​​(etc.) are typical engineering recommended values. Those skilled in the art can make reasonable adjustments based on the type, capacity, and control strategy characteristics of the tested site. The adjusted parameter values ​​still fall within the protection scope of this invention. Similarly, the waveform of the test signal is not limited to sine waves and square waves. Other periodic waveforms such as triangular waves or trapezoidal waves used by those skilled in the art according to test needs are also equivalent alternative implementations of this invention. The choice of communication protocol is not limited to Modbus TCP, IEC 61850 MMS, and OPC UA listed above. Any standard industrial communication protocol that can meet the real-time transmission requirements of the test signal is applicable.

[0087] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for simulating the breathing motion of rotational inertia of a new energy power station, characterized in that, Includes the following steps: Configure initial test parameters, including initial frequency change rate amplitude, initial test period and target response capability index value, and generate rotational inertia breathing simulation test signal based on the signal parameters of the current test period; The test signal is injected into the inertia control interface of the new energy power station to synchronously collect the frequency data, active power data and protection action signals of the grid connection point of the new energy power station at the millisecond level. After the current test cycle ends, the collected data is preprocessed and inertial response characteristic parameters are extracted. The inertial response characteristic parameters include inertial response start time, maximum active power change and inertial response duration, and the rotational inertial response capability index is calculated accordingly. A multi-dimensional response quality assessment is performed on the current test cycle, including: calculating the waveform fidelity score based on the normalized cross-correlation coefficient between the actual active power response waveform and the theoretical expected power response waveform; calculating the response symmetry index based on the comparison of the maximum active power change in the positive and negative half-cycles of the test signal; and calculating the delay consistency index based on the coefficient of variation of the inertial response start-up time in multiple test cycles. The rotational inertia response capability index, waveform fidelity score, response symmetry index, and time delay consistency index are weighted and fused to obtain a comprehensive response quality score. Anomaly detection is performed on each dimension index based on sliding window statistical process control, and degradation mode classification labels are output. Based on the deviation between the rotational inertia response capability index and the target value, the amplitude of the test signal in the next test cycle is adaptively adjusted using a proportional-integral adjustment strategy. The test signal cycle in the next test cycle is also adaptively adjusted based on the sufficiency coefficient formed by the ratio of the inertia response duration to the half-cycle of the test signal. Finally, safety boundary constraint verification is performed on the adjusted parameters. Repeat the above steps with the updated signal parameters until the planned total number of test cycles is reached or the test is manually terminated by the tester.

2. The method for simulating the breathing motion of rotational inertia of a new energy power station according to claim 1, characterized in that, The adaptive adjustment of the test signal amplitude for the next test cycle using a proportional-integral adjustment strategy specifically includes: Define the response capability index deviation ,in The target response capability index value, For the first The rotational inertia response capability index obtained from the actual calculation of each test cycle. According to the formula Calculate the amplitude correction amount, where This is the proportional gain coefficient. This is the integral gain coefficient; Update the amplitude for the next cycle to And apply a hard limiting constraint to the updated amplitude. and rate of change limit constraint .

3. The method for simulating the breathing motion of rotational inertia of a new energy power station according to claim 1, characterized in that, The adaptive adjustment of the test signal period for the next test cycle based on the sufficiency coefficient specifically includes: Calculate the sufficiency coefficient ,in For the first The duration of inertial response per test cycle This is the current test signal period; when At that time, according to Increase the cycle; when At that time, according to Shorten the cycle; among which The gain coefficient is adjusted periodically. Apply hard limiting constraints to the updated period .

4. The method for simulating the breathing motion of rotational inertia of a new energy power station according to claim 1, characterized in that, The safety boundary constraint verification includes: When the current test cycle triggers the field protection action, the amplitude of the next cycle is reduced to the preset attenuation ratio of the current amplitude, and the amplitude is locked so that it will not increase further until the protection is not triggered for a preset number of consecutive cycles, after which the protection is unlocked. Check whether the theoretical active power change corresponding to the test signal in the next cycle meets the requirements. The constraints, among which This is the equivalent inertia control factor. For the rated capacity of the station, This represents the current available active power margin of the power station. This is the safety margin factor. If the value exceeds this constraint, the magnitude will be calculated and truncated to the safe value.

5. The method for simulating the breathing motion of rotational inertia of a new energy power station according to claim 1, characterized in that, The calculation of the waveform fidelity score includes: According to the station's nominal equivalent inertia constant Rated capacity and the frequency change rate signal injected in the current period ,according to Construct the theoretically expected power response waveform; Calculate the actual response waveform Compared with the theoretically expected waveform Normalized cross-correlation coefficients between ; Waveform fidelity score is defined as .

6. The method for simulating the breathing motion of rotational inertia of a new energy power station according to claim 1, characterized in that: The calculation of the response symmetry index includes: […]. Each test cycle is divided into a positive half-cycle and a negative half-cycle according to the polarity of the test signal, and the maximum active power change in the positive half-cycle is extracted respectively. and the maximum active power change during the negative half-cycle ,according to Calculate the response symmetry index; The calculation of the latency consistency index includes: before collection Calculate the coefficient of variation of the inertial response startup time series for each test cycle. ,according to Calculate the latency consistency index, where This is the sensitivity adjustment coefficient.

7. The method for simulating the breathing motion of rotational inertia of a new energy power station according to claim 1, characterized in that, The anomaly detection of indicators across various dimensions based on sliding window statistical process control includes: For each dimension indicator, maintain a sliding window of a preset length, calculate the mean and standard deviation within the sliding window, and mark the dimension as abnormal when the value of a certain dimension indicator in the current testing period deviates from the window mean by more than a preset control limit multiple of the standard deviation. The degradation mode classification labels are output based on the combination of abnormal dimensions. The degradation mode classification labels include waveform distortion degradation, asymmetric response degradation, time delay drift degradation, inertia capability decay degradation, and overall performance degradation.

8. The method for simulating the breathing motion of rotational inertia of a new energy power station according to claim 1, characterized in that, The method further includes: When the fluctuation of the comprehensive response quality score for a consecutive preset number of cycles is less than the preset threshold, the system will automatically switch from adaptive mode to locked mode, lock the current signal parameters and no longer adjust them, and enter the steady-state test phase.

9. A simulation test system for the rotational inertia breathing of a new energy power station, characterized in that, include: An adaptive test signal generation module (100) is used to generate a rotational inertia breathing simulation test signal for each test cycle based on preset initial parameters and closed-loop feedback information. The adaptive test signal generation module (100) includes a signal waveform construction unit (101), an amplitude adaptive adjustment unit (102), a cycle adaptive scaling unit (103), and a safety boundary constraint unit (104). The inertia response trigger interface module (200) is used to inject the test signal into the inertia control interface of the new energy power station through a communication protocol. The inertia response trigger interface module (200) includes a protocol adaptation unit (201) and a signal injection synchronization unit (202). The millisecond-level data acquisition module (300) is used to synchronously acquire frequency data, active power data and protection action signals at the grid connection point of new energy power plants at the millisecond level. The millisecond-level data acquisition module (300) includes a high-speed synchronous acquisition unit (301) and a data preprocessing unit (302). The multidimensional inertial response quality assessment module (400) is used to perform multidimensional feature extraction, quality scoring and abnormal pattern recognition on the inertial active power response of each test cycle. The multidimensional inertial response quality assessment module (400) includes a basic feature extraction unit (401), a waveform fidelity scoring unit (402), a response symmetry assessment unit (403), a time delay consistency assessment unit (404), and a comprehensive quality scoring and abnormal pattern recognition unit (405). The closed-loop feedback coordination module (500) is used to transmit the evaluation results of the multidimensional inertial response quality evaluation module (400) to the adaptive test signal generation module (100) to drive the adaptive adjustment of signal parameters. The closed-loop feedback coordination module (500) includes a feedback data packaging unit (501) and an adaptive enable and mode switching unit (502). The test report generation module (600) is used to summarize all test cycle data and generate a structured test report after the test is completed.

10. The new energy power station rotational inertia breathing simulation test system according to claim 9, characterized in that, The amplitude adaptive adjustment unit (102) uses a proportional-integral adjustment strategy to calculate the amplitude correction based on the deviation of the rotational inertia response capability index, and applies hard amplitude limiting constraints and rate of change amplitude limiting constraints; the period adaptive scaling unit (103) determines whether the current test cycle is sufficient based on the sufficiency coefficient formed by the ratio of the inertia response duration to the half-cycle of the test signal, and adjusts the signal cycle of the next cycle accordingly and applies hard amplitude limiting constraints; the safety boundary constraint unit (104) performs compliance verification on the signal amplitude and cycle before each parameter update, including the amplitude reduction locking logic after the protection action is triggered and the amplitude truncation logic based on the available active power margin.

11. The new energy power station rotational inertia breathing simulation test system according to claim 9, characterized in that, The waveform fidelity scoring unit (402) obtains the waveform fidelity score by constructing the theoretically expected power response waveform and calculating the normalized cross-correlation coefficient between the actual response waveform and the theoretically expected waveform; the response symmetry evaluation unit (403) calculates the response symmetry index by comparing the maximum active power change in the positive half-cycle and negative half-cycle of the test signal; the time delay consistency evaluation unit (404) obtains the time delay consistency index by calculating the coefficient of variation of the inertial response start-up time series of multiple test cycles; the comprehensive quality scoring and abnormal pattern recognition unit (405) calculates the comprehensive response quality score by weighted fusion of multi-dimensional indicators, and applies the sliding window control chart criterion to each dimension indicator to determine abnormalities, and outputs degradation mode classification labels and diagnostic suggestions based on the combination of abnormal dimensions.

Images