Energy management method and system for off-grid system of small hydropower multi-energy complementary microgrid

Abstract

The application discloses a kind of small hydropower's multi-energy complementary microgrid off-grid system energy management method and system, and relates to microgrid energy management technical field, comprising: obtaining microgrid operation data, the microgrid operation data includes head pressure, bus frequency, guide vane change rate sequence, rotor speed sequence and stator current sequence;The drop slope of the bus frequency is calculated, when the drop slope is greater than the preset frequency drop threshold value, the power shortage sequence is calculated based on the guide vane change rate sequence, the head pressure and preset time constant;According to the rotor speed sequence, extract mechanical resonance frequency, and according to the stator current sequence, extract target harmonic phase;Its beneficial effect is: fill in the nonlinear active power shortage generated by water turbine in the initial stage of speed regulation due to water hammer effect, prevent the frequency secondary drop of off-grid system when load is suddenly added, and guarantee the transient power quality of microgrid.

Description

Technical Field

[0001] This invention relates to the field of microgrid energy management technology, and in particular to an energy management method and system for an off-grid system of a multi-energy complementary microgrid powered by small hydropower. Background Technology

[0002] In remote mountainous and forested areas without large power grid coverage, a multi-energy complementary off-grid microgrid architecture of "small hydropower + photovoltaic + energy storage" is typically used for power supply. In this architecture, small hydropower serves as the core supporting power source due to its low operation and maintenance costs and suitability for the water resources of mountainous areas, while photovoltaic and energy storage systems are used to compensate for the output limitations of small hydropower affected by seasonality and climate. However, because the off-grid operation mode is completely independent of the frequency and voltage regulation and power backup support of the infinitely large power grid, the overall rotational inertia of the system is extremely weak. When faced with random load fluctuations or extreme weather, it is prone to problems such as poor power quality or even system oscillation and collapse. Therefore, there is an urgent need for an efficient energy management method to coordinate and dispatch various heterogeneous micro-power sources.

[0003] Existing off-grid microgrid dispatching often focuses only on steady-state power balance, treating hydropower turbines and energy storage as simple, independent response modules. When the microgrid faces large disturbances (such as a sudden increase in heavy load causing a frequency drop), the small hydropower governor urgently opens its guide vanes. Due to the hydrodynamic water hammer effect, the mechanical power of the turbine not only fails to increase in the initial stage but also experiences a momentary nonlinear "reverse regulation" drop. At the same time, this transient hydraulic shock wave can easily excite the inherent mechanical torsional vibration of the long shaft system of the hydropower generator. Traditional energy management strategies rely solely on energy storage for simple linear droop frequency regulation compensation. This not only fails to accurately fill the nonlinear active power deficit during the reverse regulation period, causing a secondary frequency drop, but more dangerously, the inherent electrical harmonics generated by the high-frequency switching of the energy storage converter and photovoltaic inverter in the microgrid, once they coincide with or are close to the frequency of the unsuppressed mechanical torsional vibration of the shaft system, will directly trigger electromechanical coupling subsynchronous resonance, causing irreversible fatigue damage or even breakage of the main shaft of expensive hydropower units in remote areas, seriously threatening the survival and safety of the off-grid system. Summary of the Invention

[0004] In view of the above-mentioned prior art, this application is hereby filed. Embodiments of this application provide an energy management method and system for an off-grid system of a multi-energy complementary microgrid for small hydropower, which compensates for the nonlinear active power deficit caused by water hammer effect in the initial stage of turbine speed regulation, prevents secondary frequency drops in the off-grid system during sudden load increases, and ensures the transient power quality of the microgrid.

[0005] According to one aspect of this application, an energy management method for an off-grid system of a small hydropower multi-energy complementary microgrid is provided, comprising:

[0006] Acquire microgrid operation data, which includes head pressure, bus frequency, guide vane change rate sequence, rotor speed sequence, and stator current sequence;

[0007] Calculate the frequency drop slope of the busbar. When the frequency drop slope is greater than a preset frequency drop threshold, calculate the power deficit sequence based on the guide vane change rate sequence, the water head pressure, and a preset time constant. Extract the mechanical resonant frequency based on the rotor speed sequence and extract the target harmonic phase based on the stator current sequence.

[0008] A reconstructed digital instruction, synthesized by superimposing a first compensation sequence and a second compensation sequence, is generated only when the integral value of the power deficit sequence and the amplitude of the mechanical resonant frequency both satisfy a preset safety boundary; wherein, the first compensation sequence is a mirror compensation sequence generated based on the power deficit sequence, and the second compensation sequence is a compensation sequence with a frequency equal to the mechanical resonant frequency and a phase opposite to the target harmonic.

[0009] Output the reconstructed digital command, obtain the updated bus frequency and updated rotor speed generated in response to the reconstructed digital command, and update the microgrid operation data using the updated bus frequency and the updated rotor speed only when the drop slope of the updated bus frequency is greater than the frequency drop threshold, and recalculate the power deficit sequence and extract the mechanical resonant frequency.

[0010] According to another aspect of this application, an energy management system for a small hydropower multi-energy complementary microgrid off-grid system is provided, comprising:

[0011] The data acquisition module is used to acquire microgrid operation data, which includes head pressure, bus frequency, guide vane change rate sequence, rotor speed sequence, and stator current sequence.

[0012] The calculation and extraction module is used to calculate the frequency drop slope of the bus. When the frequency drop slope is greater than the preset frequency drop threshold, the power deficit sequence is calculated based on the guide vane change rate sequence, the water head pressure, and the preset time constant. The mechanical resonant frequency is extracted according to the rotor speed sequence, and the target harmonic phase is extracted according to the stator current sequence.

[0013] The instruction reconstruction module is used to generate a reconstruction digital instruction synthesized by superimposing a first compensation sequence and a second compensation sequence only when the integral value of the power deficit sequence and the amplitude of the mechanical resonant frequency both meet the preset safety boundary; wherein, the first compensation sequence is a mirror compensation sequence generated based on the power deficit sequence, and the second compensation sequence is a compensation sequence with the same frequency as the mechanical resonant frequency and the opposite phase to the target harmonic.

[0014] The closed-loop control module is used to output the reconfiguration digital command, obtain the updated bus frequency and updated rotor speed generated in response to the reconfiguration digital command, and update the microgrid operation data using the updated bus frequency and updated rotor speed only when the drop slope of the updated bus frequency is greater than the frequency drop threshold, and recalculate the power deficit sequence and extract the mechanical resonant frequency.

[0015] According to another aspect of this application, an electronic device is provided, including a memory and a processor, the memory being used to store computer-executable instructions, and the processor being used to execute the computer-executable instructions, which, when executed by the processor, implement the steps of the method described above.

[0016] According to another aspect of this application, a computer-readable storage medium is provided that stores computer-executable instructions thereon, which, when executed by a processor, implement the steps of the method described above.

[0017] Compared with existing technologies, the energy management method and system for off-grid systems of multi-energy complementary microgrids for small hydropower, according to an embodiment of this application, calculates the frequency drop slope of the bus in real time and calculates the power deficit sequence using the guide vane change rate sequence, head pressure, and preset time constant when a large disturbance is detected. Based on this, a mirrored compensation sequence is generated to guide devices with rapid response capabilities within the microgrid (such as energy storage converters) to perform equal active power offsetting. This mechanism transforms the inertial hysteresis of water flow in hydraulics into quantified digital compensation commands, filling the nonlinear active power deficit caused by water hammer effect in the initial stage of turbine speed regulation, preventing secondary frequency drops in the off-grid system during sudden load increases, and ensuring the transient power quality of the microgrid.

[0018] By cross-analyzing the rotor speed sequence and stator current sequence, the mechanical resonant frequency and corresponding electrical harmonic phase of the unit are extracted. The system then generates a compensation sequence with the opposite harmonic phase. By using power electronic devices to actively inject anti-phase high-frequency damping current into the generator air gap, this method directly offsets the alternating torque acting on the long shaft system at the electrical level, reduces the mechanical vibration caused by transient hydraulic impact, reduces the risk of main shaft fatigue damage to expensive hydroelectric equipment in remote areas, and effectively reduces the long-term operation and maintenance costs of the system.

[0019] By double-checking the integral value of the power deficit sequence and the amplitude of the mechanical resonant frequency within a preset safety boundary, the physical safety of the compensation command output is ensured, preventing equipment overload due to overcompensation. Simultaneously, after outputting the command, the system continuously acquires updated bus frequency and rotor speed for step-by-step verification and cyclic correction. This closed-loop data update feedback mechanism enables the microgrid system to possess dynamic self-calibration capabilities in the face of continuously deteriorating operating conditions, enhancing the survivability of off-grid power supply systems in complex environments.

[0020] Furthermore, the method described in this application operates as a transient compensation layer in a multi-energy complementary off-grid microgrid of "small hydropower-photovoltaics-energy storage," forming a two-layer collaborative architecture with the steady-state energy management layer that allocates power according to hydrological conditions. The compensation commands generated are executed by fast-response converters within the microgrid (e.g., energy storage converters controlled by virtual synchronous generators), enabling the transient active power deficit caused by water hammer effect in small hydropower to be filled by power electronic equipment with millisecond-level response capabilities, without relying on the mechanical adjustment process of the turbine governor itself, thereby improving the speed and controllability of transient compensation. Attached Figure Description

[0021] The above and other objects, features, and advantages of this application will become more apparent from the more detailed description of the embodiments of this application in conjunction with the accompanying drawings. The drawings are provided to further illustrate the embodiments of this application and form part of the specification. They are used together with the embodiments of this application to explain this application and do not constitute a limitation thereof. In the drawings, the same reference numerals generally represent the same components or steps.

[0022] Figure 1 This is a schematic diagram of the overall process of an off-grid energy management method for a multi-energy complementary microgrid system of small hydropower according to the present invention.

[0023] Figure 2 This is a schematic diagram of the logical framework of an off-grid energy management method for a multi-energy complementary microgrid system of small hydropower according to the present invention.

[0024] Figure 3 This is an extended schematic diagram of the energy management method for a multi-energy complementary microgrid off-grid system of small hydropower according to the present invention. Detailed Implementation

[0025] Hereinafter, exemplary embodiments according to this application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of this application, and not all embodiments of this application. It should be understood that this application is not limited to the exemplary embodiments described herein.

[0026] Example 1:

[0027] To address the problem that existing multi-energy complementary off-grid microgrid energy management methods only focus on steady-state power balance and ignore the cross-physical field coupling hazards caused by active power reversal due to water hammer effect and mechanical torsional vibration of unit shafts during transient processes, leading to secondary frequency drops in the system under large disturbances and easy fatigue damage to the main shaft of core hydroelectric equipment due to resonance, this application acquires multi-dimensional microgrid operation data including hydrodynamic and electromechanical characteristics in real time. When the bus frequency drops beyond the limit, the power deficit caused by transient water flow response hysteresis is quantitatively extracted, and the mechanical resonance frequency and target harmonic phase inside the electromechanical system are calculated simultaneously. Based on the above cross-domain characteristics, after verifying the dual safety boundaries of system energy loss and physical stress, the macroscopic active power mirror compensation and the microscopic high-frequency harmonic anti-phase damping are fused and superimposed to generate multi-energy collaborative reconfiguration digital commands for execution. Subsequently, a state step verification and closed-loop feedback mechanism after command response are introduced to continuously track the microgrid frequency evolution trend and dynamically update the compensation strategy.

[0028] This application utilizes a dual-track data processing mechanism across physical fields to transform the transient inertial hysteresis of the turbine into quantified digital compensation commands. When the system faces sudden large disturbances, it accurately fills the nonlinear active power deficit caused by the water hammer effect, avoiding the collapse and instability of the microgrid frequency under conventional control. Simultaneously, based on the characteristic superposition and synthesis of microscopic electrical harmonics and mechanical vibrations, it actively injects anti-phase high-frequency damping current into the generator air gap using power electronic converter equipment. This directly offsets and eliminates the alternating torque acting on the long shaft system at the electrical level, eliminating the hidden fatigue damage risk of the unit's main shaft. This effectively improves the transient power quality of off-grid microgrid systems in remote areas and extends the safe service life of core power generation equipment.

[0029] Reference Figures 1-3 As an embodiment of the present invention, an energy management method for a multi-energy complementary microgrid off-grid system of small hydropower is provided, comprising: S1-S4.

[0030] To facilitate understanding, the following example of a typical off-grid microgrid with multi-energy complementarity in a mountainous area will be used to illustrate Implementation Example 1.

[0031] In this example, the off-grid microgrid is equipped with a small hydropower unit with a rated capacity of 100kW, a photovoltaic array of 50kW, a 100kWh energy storage battery, and a mixed AC / DC load with a total capacity of approximately 80kW. It operates independently from the main power grid, with a bus rated voltage of 380V and a rated frequency of 50Hz.

[0032] During normal operation, the microgrid uses small hydropower as the main power source for voltage and frequency regulation to bear the basic load, photovoltaic power is consumed locally, and energy storage is coordinated for charging and discharging according to different hydrological conditions such as high water, normal water, and low water. The following example 1 addresses the transient process caused by the microgrid encountering large disturbances such as sudden load increases on the basis of the above steady-state scheduling.

[0033] Furthermore, the aforementioned steady-state scheduling and the following transient compensation constitute a two-layer collaborative architecture: the steady-state energy management layer determines the power allocation among small hydropower, photovoltaic, and energy storage based on hydrological conditions and photovoltaic output at a time scale ranging from minutes to hours; the method described in Example 1 below serves as the transient compensation layer, compensating for transient power deficits and torsional vibrations of the unit shaft system caused by large disturbances at a time scale of milliseconds; the time scales of the two do not overlap and can operate in parallel without interfering with each other. The transient compensation is implemented by the fast-response converter within the microgrid, specifically, in this example, by an energy storage converter controlled by a virtual synchronous generator; that is, when a small hydropower station experiences transient active power reversal due to water hammer, the missing active power is filled by the energy storage converter through power injection, rather than relying on the turbine governor to further expand the guide vanes to absorb the loss, thus avoiding the adverse coupling between the governor's action and the inertial hysteresis of the water flow.

[0034] For example, suppose that at a certain moment a high-power irrigation pump unit is put into the microgrid, causing the mixed load on the bus to increase from about 45kW to about 75kW in a step, and the bus frequency subsequently drops from 50Hz, thereby triggering the energy management method of Embodiment 1. In the following steps, wherever specific numerical values ​​are involved, the description will follow the example scenario.

[0035] Figure 1 and Figure 2 The illustration shows an energy management method for an off-grid system of a small hydropower multi-energy complementary microgrid according to an embodiment of this application, specifically including:

[0036] Reference Figure 1 In step S1, microgrid operation data is acquired, including head pressure, bus frequency, guide vane change rate sequence, rotor speed sequence, and stator current sequence.

[0037] In this step, the microgrid operation data consists of multi-dimensional real-time measurement data reflecting the hydraulic and electromechanical characteristics of the off-grid microgrid. Among them, the head pressure is the measured pressure value at the inlet of the turbine unit, reflecting the strength of available hydropower; the bus frequency reflects the real-time balance of active power within the microgrid; the guide vane rate of change sequence is the sequence obtained by differentiating the relative opening of the turbine governor guide vanes with respect to time, reflecting the speed of governor operation; the rotor speed sequence is the sampling sequence of the generator rotor angular velocity, carrying the mechanical vibration information of the unit shaft system; and the stator current sequence is the sampling sequence of the generator stator current, carrying electrical harmonic information.

[0038] Specifically, the aforementioned quantities can be obtained from their respective acquisition terminals: water head pressure is obtained from water level or pressure sensors located on the pressure steel pipe; bus frequency and stator current are obtained from electrical measurement terminals on the bus side; guide vane rate of change is obtained from the position feedback of the speed governor; and rotor speed is obtained from speed sensors at the unit shaft end. The raw data obtained from each acquisition terminal is first filtered by median value to remove instantaneous abnormal sampling points, and then processed by abnormal data removal and time sequence alignment to align all data under a unified time reference for subsequent calculations.

[0039] Discrete sampling period of the acquisition Based on the highest frequency of interest, to ensure that the rotor speed sequence and stator current sequence can carry the unit's shaft torsional vibration and electrical harmonic information without aliasing, the discrete sampling period is determined. The value of satisfies the Nyquist sampling condition, that is:

[0040] ;

[0041] in, The highest frequency (Hz) of interest is selected, which is the larger of the mechanical torsional vibration frequency of the unit shaft system and the electrical harmonic frequency.

[0042] Using the previous example scenario, we take the discrete sampling period. (i.e., sampling frequency of 1kHz) can cover shaft torsional vibration and electrical harmonics in the tens of hertz range without aliasing; it can simultaneously acquire water head pressure (corresponding to rated water head of about 40m), bus frequency (50Hz before disturbance), guide vane rate of change sequence, rotor speed sequence and stator current sequence, and complete filtering and timing alignment.

[0043] return Figure 1 In step S2, the slope of the bus frequency drop is calculated, and the specific formula is as follows:

[0044] ;

[0045] in, This represents the bus frequency drop slope (Hz / s) at the i-th sampling time. and Let represent the bus frequency (Hz) at the i-th and (i-1)-th sampling times, respectively. The discrete sampling period time (s);

[0046] It should be noted that when i=1, i.e., at the initial sampling time or the start of the disturbance, and there is no data for the (i-1)th sampling time, the rated value of the bus frequency (e.g., 50Hz) when the system is in steady state and no disturbance occurs will be used as f(0) for calculation. Similarly, for subsequent forward differential operations involving timing characteristics such as guide vane change rate, if there are missing historical moments due to being located at the sampling boundary, the steady-state reference data cached before the disturbance is triggered will be used as the boundary filling condition by default to ensure the absolute continuity of timing calculation.

[0047] The aforementioned bus frequency sag slope is used to characterize the rate of development of active power imbalance in the microgrid. The reason for using the sag slope rather than the instantaneous value of the frequency deviation as the criterion for large disturbances is that the frequency deviation only reflects the cumulative result of the imbalance, while the sag slope can reflect the development trend of the imbalance before the frequency exceeds the limit, thus allowing for advance compensation. Compared with the criterion based on the frequency deviation threshold, the criterion based on the sag slope is more sensitive to rapid disturbances such as sudden load increases.

[0048] When the drop slope is greater than the preset frequency drop threshold, the power deficit sequence is calculated based on the guide vane change rate sequence, head pressure, and preset time constant. The specific formula is as follows:

[0049] ;

[0050] in, The power deficit sequence element value calculated at the i-th sampling time reflects the transient active power deficit (W) caused by the inertial hysteresis of water flow (water hammer effect). The equivalent rated volumetric flow rate constant of the turbine unit in the microgrid system ( ), Let be the head pressure (Pa) obtained at the i-th sampling time. The preset water flow inertia time constant is used to characterize the physical hysteresis property (s) of the water flow dynamic response. Let be the rate of change of the guide vane at the i-th sampling time, which is physically represented as the time derivative of the guide vane relative opening (1 / s).

[0051] It should be noted that the frequency drop threshold Used to distinguish between normal fluctuations and large disturbances in a microgrid, its value is determined by the maximum allowable transient frequency deviation of the microgrid and the judgment confirmation time window, i.e.:

[0052] ;

[0053] in, This is the frequency drop threshold (Hz / s). This is the maximum allowable transient frequency deviation (Hz) for the microgrid. To determine the confirmation window (s) It can be determined based on the allowable fluctuation range of the off-grid microgrid bus frequency, for example, taking 0.5Hz corresponding to the allowable fluctuation of ±0.5Hz of the bus frequency; To allow for a time margin for the accumulation of frequency deviations, several discrete sampling periods are taken, when At that time, it is determined that a large disturbance has occurred in the microgrid and subsequent calculations are triggered; using the aforementioned example scenario, take =0.5Hz =0.2s, then =2.5Hz / s. After a sudden load increase, the bus frequency drops from 50.0Hz to approximately 49.7Hz within about 0.1s. This can be calculated using the bus frequency drop slope formula. The frequency is approximately 3.0 Hz / s, which is greater than the threshold of 2.5 Hz / s. Based on this, it is determined that a large disturbance has occurred in the microgrid.

[0054] After determining that a large disturbance has occurred, it is necessary to quantify the transient power shortage of the turbine. This is because when the turbine governor opens the guide vanes urgently under a large disturbance, the mechanical power of the turbine does not increase immediately with the opening of the guide vanes due to the inertial hysteresis of the water flow (water hammer effect). Instead, it experiences a momentary reverse decrease, resulting in an active power deficit. The above power deficit formula expresses this active power deficit as the product of the rated volumetric flow rate constant, the head pressure, the inertial time constant of the water flow, and the guide vane change rate. This transforms the inertial hysteresis of the water flow at the hydraulic level into an active power quantity that can be calculated in real time. Compared with the method of obtaining transient power through detailed transfer functions of the turbine or numerical simulation of fluid mechanics, this formula only relies on the guide vane change rate and head pressure, which are easily obtained in real time. The amount of calculation is smaller and the real-time performance is better.

[0055] It should be noted that the power deficit formula described above describes the change in mechanical power caused by the inertial lag of the water flow during the turbine speed regulation process. Its applicable boundary depends on the value of the guide vane change rate: when the guide vane change rate is greater than zero, that is, when the governor opens the guide vane to the maximum and the corresponding load surge occurs, the establishment of the water flow lags behind the change in the guide vane opening. The mechanical power decreases in the initial stage of speed regulation. At this time, the element value of the power deficit sequence is positive, indicating the active power deficit that needs to be compensated by the fast response equipment; when the guide vane change rate is equal to zero, the water flow has entered a quasi-steady state, the element value of the power deficit sequence is zero, and compensation is not triggered; when the guide vane change rate is less than zero, that is, when the governor closes the guide vane to the minimum and the corresponding load shedding occurs, the inertial lag of the water flow causes the mechanical power to surge instantaneously in the initial stage of regulation. At this time, the element value of the power deficit sequence is negative, indicating a transient active power surplus. Correspondingly, the first compensation sequence generated subsequently becomes an indication for the fast response equipment to absorb active power. It can be seen that the above formula holds true under both transient operating conditions of the guide vane opening wide and closing narrow, and the positive or negative value of the formula automatically indicates the direction of compensation.

[0056] First, the water flow inertia time constant preset in the above formula Its value is determined by the geometric and hydraulic parameters of the water diversion system, namely:

[0057] ;

[0058] Where L is the total length (m) of the water diversion pipeline. The velocity of the water flow in the pipeline under rated operating conditions is (m / s), and g is the acceleration due to gravity. ), The rated head of the unit (m);

[0059] Secondly, the equivalent rated volumetric flow rate constant of the turbine unit in the microgrid system. This is the equivalent volumetric flow rate of the turbine unit under rated operating conditions, and its value can be determined by the rated parameters of the unit, i.e.:

[0060] ;

[0061] in, The rated active power output (W) of the generator unit. The density of water ( ), The rated efficiency of the turbine unit (dimensionless).

[0062] Using the previous example scenario, let L = 300m. =1.3m / s =40m, g=9.8 ,but ≈1.0s; take =100kW (i.e., 100,000W) =1000kg / m³ =0.85, then ≈0.30 m³ / s; during periods of large disturbance, the head pressure Governor guide vane change rate Approximately 0.3 / s, substituting into the power deficit formula yields... That is, approximately 35.3 kW, which is equivalent to the sudden increase in load.

[0063] The mechanical resonant frequency is extracted based on the rotor speed sequence, and the target harmonic phase is extracted based on the stator current sequence, including:

[0064] Phase space delay reconstruction is performed on the rotor speed sequence to construct a multidimensional trajectory sequence set. Phase space delay reconstruction refers to reconstructing a multidimensional trajectory that reflects the vibration state of the unit's shaft system from a one-dimensional rotor speed sequence. Specifically, the values ​​of the rotor speed sequence under different delays are used as components of each dimension to construct a multidimensional trajectory sequence set, and the nth trajectory point is represented as:

[0065] ;

[0066] in, It is a set of multidimensional trajectory sequences. This is a rotor speed sequence. Let m be the delay step size, and m be the embedding dimension. The delay corresponding to the first zero crossing of the autocorrelation function of the rotor speed sequence (or the first minimum of the mutual information function) can be taken to ensure that the components of each dimension have appropriate independence; the embedding dimension m can be determined by the spurious nearest neighbor method, and the minimum dimension that makes the proportion of spurious nearest neighbors tend to zero can be taken to ensure that the reconstructed trajectory is fully unfolded without folding.

[0067] The reason for reconstructing the phase space first and then extracting the resonant frequency, instead of directly performing spectral analysis on the rotor speed sequence, is that the rotor speed sequence has non-stationary and nonlinear characteristics during large disturbance transient processes. When performing spectral analysis directly, the dominant vibration mode is easily submerged by noise and adjacent frequency components. After phase space reconstruction, the dominant vibration mode is unfolded in the multidimensional trajectory, which facilitates subsequent extraction.

[0068] Extract the dominant feature vector of the multidimensional trajectory sequence set, transform the dominant feature vector to the frequency domain data space, and determine the corresponding dominant frequency feature value as the mechanical resonant frequency;

[0069] In this step, extracting the dominant feature vector of the multidimensional trajectory sequence set means performing feature decomposition (e.g., singular value decomposition or principal component analysis) on the trajectory matrix composed of multidimensional trajectory points, and taking the feature vector corresponding to the largest singular value (or the largest eigenvalue) as the dominant feature vector. This dominant feature vector corresponds to the vibration mode with the largest energy proportion in the trajectory.

[0070] Transforming the dominant eigenvector into the frequency domain data space involves performing a Discrete Fourier Transform (e.g., Fast Fourier Transform) on the dominant eigenvector. The frequency corresponding to the spectral peak with the largest amplitude in the resulting spectrum is the dominant frequency characteristic value, which is then configured as the mechanical resonant frequency. The amplitude of this spectral peak reflects the strength of the dominant torsional vibration of the shaft system and is denoted as the resonance amplitude. This is used to characterize mechanical stress in the subsequent step S3.

[0071] Using the previous example scenario, with a delay step of τ = 8 sampling points and an embedding dimension of m = 5, the multidimensional trajectory is reconstructed from the rotor speed sequence and singular value decomposition is performed. The dominant eigenvector is then subjected to a fast Fourier transform, yielding a dominant frequency of approximately 23Hz. Based on this, the mechanical resonant frequency is determined. The frequency was set to 23Hz, and the amplitude of this peak was taken as the resonant amplitude. ;

[0072] Generating a quadrature phase-locked loop (PLL) reference sequence based on the mechanical resonant frequency; in this step, generating a quadrature phase-locked loop reference sequence based on the mechanical resonant frequency means using the mechanical resonant frequency as the reference frequency to generate an in-phase reference sequence and a quadrature reference sequence with mutually orthogonal phases, respectively expressed as:

[0073] ;

[0074] ;

[0075] in, As an in-phase reference sequence, It is an orthogonal reference sequence. Where is the mechanical resonant frequency (Hz), and n is the sampling point number. For discrete sampling periods (s), an orthogonal phase-locked reference sequence is generated, which provides a phase reference for subsequently separating the target harmonic components near the mechanical resonant frequency from the stator current sequence;

[0076] Demodulation projection transformation is performed on the stator current sequence and the quadrature phase-locked reference sequence to separate the target harmonic data segment, including:

[0077] The quadrature phase-locked reference sequence is analyzed to separate the in-phase reference sequence and the quadrature reference sequence; a stepwise correlation mapping is performed on the stator current sequence and the in-phase reference sequence to generate the in-phase modulation sequence, and a stepwise correlation mapping is performed on the stator current sequence and the quadrature reference sequence to generate the quadrature modulation sequence.

[0078] In this step, the element-wise correlation mapping refers to multiplying the stator current sequence and the reference sequence one by one according to the sampling points; multiplying the stator current sequence with the in-phase reference sequence and the quadrature reference sequence one by one respectively, to obtain the in-phase modulation sequence and the quadrature modulation sequence, which are expressed as follows:

[0079] ;

[0080] ;

[0081] in, It is an in-phase modulation sequence. It is an orthogonal modulation sequence. The stator current sequence (A) is essentially a series of components multiplied one by one. The stator current is orthogonally demodulated with the mechanical resonant frequency as the center, so that the components of the stator current with frequencies close to the mechanical resonant frequency are shifted to low frequencies, while the remaining components are shifted to high frequencies.

[0082] The preset low-pass filter truncation parameters are called to perform convolution smoothing on the in-phase modulation sequence and the quadrature modulation sequence respectively, and extract the corresponding in-phase baseband feature sequence and quadrature baseband feature sequence.

[0083] In this step, convolution smoothing refers to performing convolution on the in-phase modulation sequence and the quadrature modulation sequence using a low-pass filter kernel, filtering out the high-frequency components and retaining only the target components shifted to low frequencies, thus obtaining the in-phase baseband feature sequence and the quadrature baseband feature sequence, respectively represented as:

[0084] ;

[0085] ;

[0086] in, The characteristic sequence of the in-phase baseband is (A). Let A be the orthogonal baseband feature sequence, and k be the discrete sliding summation step size variable in the convolution operation, with a value ranging from 0 to 1. positive integers, where The effective sequence length of the configured low-pass filter kernel.

[0087] This indicates that the elements within the sliding window are summed. This is a dimensionless low-pass filter core. The low-pass filter cutoff parameter is the cutoff frequency of the low-pass filter core, and its value is less than the mechanical resonant frequency. And it is greater than the frequency of the target component after demodulation, so as to filter out the high-frequency components generated by demodulation and retain the target component; in engineering, it can be taken as a fraction of the mechanical resonant frequency;

[0088] Specifically, the cutoff frequency of the low-pass filter core can be taken as one-fifth to one-tenth of the mechanical resonant frequency. Within this range, it can effectively filter out the high-frequency components near the second harmonic of the mechanical resonant frequency generated by demodulation, while completely preserving the target components shifted to the low frequency by demodulation, and also taking into account the convergence speed of phase estimation. In addition, considering that the subsequent steps in this process require the target harmonic phase to be obtained from the in-phase baseband characteristic sequence and the quadrature baseband characteristic sequence, the low-pass filter core should preferably adopt a linear phase finite impulse response structure. Its group delay is a constant value in the passband, equal to half a sampling point of the effective sequence length of the low-pass filter core minus one. It introduces the same time delay to the in-phase and quadrature paths, which cancel each other out when the phase is obtained by the ratio of the two, thus not introducing additional phase estimation bias. If a phase nonlinear infinite impulse response structure is used, the phase shift of each frequency component in the passband will be different, which will introduce frequency-related estimation bias into the target harmonic phase, so it is not advisable to use it.

[0089] The in-phase baseband feature sequence and the orthogonal baseband feature sequence are vectorized and concatenated to construct the target harmonic data segment; the phase offset attribute of the target harmonic data segment is extracted and configured as the target harmonic phase;

[0090] In this step, the target harmonic data segment is constructed by vectorizing and splicing the in-phase baseband characteristic sequence and the orthogonal baseband characteristic sequence. and The complex baseband data is constructed; the phase offset attribute of the target harmonic data segment is extracted and configured as the target harmonic phase, which means obtaining the target harmonic phase from the arctangent of the ratio of the quadrature component to the in-phase component, i.e.:

[0091] ;

[0092] in, For the target harmonic phase (rad), arctan2 is the four-quadrant arctangent operation. During the arctangent operation, the orthogonal baseband characteristic sequence is... The data is mapped to the imaginary part of the complex field operation as input, and the in-phase baseband characteristic sequence is used. The input is mapped to the real part of the complex field operations;

[0093] The reason for using orthogonal demodulation to extract the target harmonic phase, instead of directly performing spectral analysis on the stator current sequence to read the phase, is that the mechanical resonant frequency has already been determined by the aforementioned steps. Orthogonal demodulation of this known frequency can obtain a relatively stable phase estimate within a shorter data window, and it has a certain ability to suppress noise and adjacent frequency components. Thus, the mechanical resonant frequency is obtained from the rotor speed sequence. and resonance amplitude The target harmonic phase is obtained from the stator current sequence. This provides a basis for generating the compensation sequence in the subsequent step S3.

[0094] return Figure 1 In step S3, a reconstructed digital instruction synthesized by superimposing the first compensation sequence and the second compensation sequence is generated only when the integral value of the power deficit sequence and the amplitude of the mechanical resonant frequency both meet the preset safety boundary; wherein, the first compensation sequence is a mirror compensation sequence generated based on the power deficit sequence, and the second compensation sequence is a compensation sequence with the same frequency as the mechanical resonant frequency and the opposite phase to the target harmonic.

[0095] In this step, the first compensation sequence is a mirror compensation sequence generated based on the power deficit sequence. It is equal in magnitude and opposite in direction to the power deficit sequence, i.e., the power deficit sequence is negative. This sequence is used to guide fast-response devices such as energy storage converters to inject an equal amount of active power to compensate for the active power deficit caused by the water hammer effect. The second compensation sequence has a frequency equal to the mechanical resonant frequency. And the phase is the same as the target harmonic phase. The opposite compensation sequence is used to guide the power electronic equipment to inject anti-phase high-frequency damping current into the generator air gap, thereby offsetting the alternating torque acting on the shaft system at the electrical level.

[0096] The reason for performing a dual safety boundary check using the integral value of the power deficit sequence and the amplitude of the mechanical resonant frequency before generating the reconfiguration digital command is that the former reflects the energy required for compensation, the energy and overload capacity of the corresponding fast response equipment, and the latter reflects the strength of the shaft torsional vibration and the mechanical bearing capacity of the corresponding unit shaft system. The two respectively constrain the energy boundary and mechanical boundary of the compensation, and the compensation command is only output when both are within the safety boundary, which can avoid the fast response equipment from exceeding the limit overload or the shaft stress from exceeding the limit due to overcompensation.

[0097] Furthermore, the generation of reconstructed digital instructions includes:

[0098] Extract the energy loss parameter corresponding to the integral value of the power deficit sequence, and extract the physical stress parameter corresponding to the amplitude of the mechanical resonant frequency.

[0099] In this step, the energy loss parameter corresponding to the integral value of the power deficit sequence refers to the energy obtained by integrating the power deficit sequence over time within the perturbation time window, expressed as:

[0100] ;

[0101] in, The energy loss parameter is (J). The power deficit sequence element value (W) is calculated at the i-th sampling time. The discrete sampling period is (s). This means that within the set disturbance time window, the discrete energy blocks corresponding to all discrete sampling points i are summed and calculated. The physical stress parameter corresponding to the amplitude of the mechanical resonant frequency is taken as the aforementioned resonant amplitude. It reflects the strength of the dominant torsional vibration of the shaft system and is positively correlated with the alternating mechanical stress borne by the shaft system;

[0102] The energy loss parameter and the physical stress parameter are combined to construct a two-dimensional state coordinate vector;

[0103] In this step, the energy loss parameter and the physical stress parameter are combined to construct a two-dimensional state coordinate vector, that is, the energy loss parameter is used as the basis for the two-dimensional state coordinate vector. Using one-dimensional coordinates and physical stress parameters This represents another dimension of coordinates, forming a state point in a two-dimensional state space;

[0104] The reason for combining the two parameters into a two-dimensional state coordinate vector and verifying them in a two-dimensional state space, rather than setting scalar thresholds separately, is that the physical meaning and dimensions of the energy loss parameter (with dimensions of J) and the physical stress parameter (characterizing the torsional amplitude of the shaft system) are different, and the two cannot be converted into a single scalar for comparison. Placing them in a two-dimensional state space and characterizing the safety boundary with a two-dimensional feasible region can simultaneously constrain the energy boundary and the mechanical boundary. Setting independent scalar thresholds for the two separately is insufficient to reflect the coupling relationship between the two constraints.

[0105] Invoke the envelope data matrix corresponding to the preset security boundary, perform spatial topology mapping of the two-dimensional state coordinate vector relative to the envelope data matrix, and generate compliance triggering signaling, including:

[0106] In this step, the envelope data matrix is ​​a matrix composed of the coordinates of the boundary nodes of the safe and feasible region in the aforementioned two-dimensional state space arranged in order, with each row representing the coordinates of a node on the boundary of the feasible region. The values ​​of the envelope data matrix are determined by the safety limits of the equipment: the boundary in the energy dimension is determined by the maximum energy that a fast-response device (such as an energy storage converter) can call upon within the allowable overload time, and the boundary in the mechanical dimension is determined by the maximum torsional amplitude allowed by the unit's shaft system (corresponding to its fatigue limit). Discretizing the safety limits in each of the above directions in the two-dimensional state space and connecting them in order constitutes the boundary of the safe and feasible region described by the envelope data matrix.

[0107] Using the previous example scenario, the values ​​of the envelope data matrix can be determined as follows: In the energy dimension, the rated power of the energy storage converter is 100kW, its allowable short-term overload multiple is 1.2 times, and its allowable overload time is 1s. Therefore, the maximum energy that the device can access within the allowable overload time is 1.2 × 100kW × 1s, i.e. J represents the physical limit of the equipment; after taking into account factors such as battery state of charge fluctuations, equipment aging, and the capacity occupancy of the converter by other control components, a certain safety margin is adopted. J serves as the tuning boundary in the energy dimension. In the mechanical dimension, the maximum allowable alternating torque of the unit's shaft system is taken as 30% of the rated torque (this value is determined based on the fatigue limit of the shaft system material). Then, based on the torsional stiffness of the shaft system and the unit's moment of inertia, this allowable alternating torque is converted into the torsional amplitude of the rotor speed at the mechanical resonant frequency. The resulting amplitude is the maximum allowable value of the resonant amplitude in the mechanical dimension. Discretizing the limits in the two directions in the two-dimensional state space and connecting them sequentially yields the specific values ​​of the envelope data matrix in this example scenario.

[0108] Performing a spatial topological mapping of the two-dimensional state coordinate vector relative to the envelope data matrix, i.e., determining whether the state point is within the aforementioned safe and feasible region, and generating compliant triggering signaling accordingly, the specific process includes the following steps:

[0109] Extract the sequence of outer topological boundary nodes from the envelope data matrix;

[0110] Determine the shortest spatial projection vector of the two-dimensional state coordinate vector relative to the sequence of external topological boundary nodes;

[0111] Construct a test ray vector that starts from the two-dimensional state coordinate vector and extends along the shortest spatial projection vector, and determine the cross-traversal dataset of the test ray vector and the sequence of external topological boundary nodes;

[0112] Extract the parity mapping attributes from the cross-traversal dataset, and structurally concatenate the parity mapping attributes with the modulus data of the shortest spatial projection vector to generate compliant triggering signaling;

[0113] In this step, the process of determining whether a state point is within the feasible region is performed using the parity of ray intersections: starting from the state point corresponding to the two-dimensional state coordinate vector, a test ray vector is constructed along the direction of the shortest spatial projection vector. The intersections of this test ray with the boundary formed by the sequence of topological boundary nodes outside the feasible region are counted. When the number of intersections is odd, it indicates that the state point is within the feasible region and is deemed compliant; when the number of intersections is even, it indicates that the state point is outside the feasible region and is deemed non-compliant. The parity mapping attribute of the cross-traversal dataset is extracted, that is, the parity of the number of intersections is extracted.

[0114] The parity mapping attribute and the magnitude data of the shortest spatial projection vector are structurally concatenated to generate compliance triggering signal because the parity mapping attribute characterizes whether the state point is compliant, while the magnitude of the shortest spatial projection vector characterizes the distance from the state point to the safety boundary, i.e., the margin of the current compensation intensity relative to the safety boundary. Together, they constitute the compliance triggering signal, which indicates whether the output compensation command is allowed and carries margin information for subsequent reference. The reason for using ray cross parity discrimination instead of edge-by-edge cross product discrimination is that the boundary of the aforementioned safe feasible region may be non-convex, and ray cross parity discrimination is applicable to both convex and non-convex boundaries.

[0115] In response to a compliance trigger signaling, the first compensation sequence is mapped to the direct axis channel of a preset orthogonal coordinate system to construct a direct axis data component, and the second compensation sequence is mapped to the cross axis channel of a preset orthogonal coordinate system to construct a cross axis data component.

[0116] In this step, the orthogonal coordinate system is a two-phase rotating coordinate system that rotates synchronously with the generator's electrical angular velocity, containing mutually orthogonal direct-axis and quadrature-axis channels. The positioning of this orthogonal coordinate system is determined by the phase reference provided by the aforementioned orthogonal phase-locked reference sequence, i.e., the phase obtained from phase-locking is used as the orientation angle of the rotating coordinate system. The first compensation sequence is mapped to the direct-axis channel, and the second compensation sequence is mapped to the quadrature-axis channel because the first compensation sequence corresponds to the active power component that fills the active power deficit and is associated with the direct axis, while the second compensation sequence corresponds to the anti-phase high-frequency damping component and is associated with the quadrature axis.

[0117] The reason for mapping the two compensation paths to mutually orthogonal direct and quadrature axis channels is that the direct axis components and quadrature axis components are decoupled from each other and can be synthesized and executed by the same power electronic device in a unified rotating coordinate system without interfering with each other. This allows the active power compensation at the macro level and the harmonic anti-phase damping at the micro level to be achieved in the same reconfiguration digital instruction.

[0118] Regarding the execution equipment, the aforementioned direct-axis channel corresponds to the active power control channel of the fast-response converter, used to carry the active power mirror compensation corresponding to the first compensation sequence; the aforementioned quadrature-axis channel corresponds to the harmonic damping current control channel of the corresponding equipment, used to carry the anti-phase alternating damping current corresponding to the second compensation sequence, with a frequency equivalent to the mechanical resonant frequency. This damping current differs from the steady-state reactive power regulation in both frequency scale and control objective. In the implementation of the fast-response converter using virtual synchronous generator control, the first compensation sequence is superimposed on the active power reference quantity of the virtual synchronous generator control in the form of power increments, participating in its active-frequency loop calculation; the second compensation sequence is superimposed on the current inner loop reference quantity of the virtual synchronous generator control, and is directly tracked and executed by the current inner loop. Since the virtual synchronous generator control itself provides virtual inertia and damping support for the off-grid microgrid under steady-state and quasi-steady-state conditions, and the compensation described in this application superimposes feedforward water hammer deficit filling and shaft torsional vibration damping on top of it, the two are implemented in layers within the same converter equipment and do not conflict with each other.

[0119] Perform vector orthogonal synthesis transformation on the direct-axis data components and the quadrature-axis data components to generate reconstructed digital instructions;

[0120] In this step, performing a vector orthogonal synthesis transformation on the direct-axis and quadrature-axis data components means synthesizing the two components from the direct-axis and quadrature-axis paths using an inverse transformation of the rotating coordinate system into a three-phase control command that the executing device can receive. Specifically, the direct-axis data component is converted from the active power mirror compensation corresponding to the first compensation sequence into a direct-axis current reference, i.e.:

[0121] ;

[0122] in, For direct-axis current reference (A). The active mirror compensation power (W, which is numerically equal to the active compensation amount corresponding to the first compensation sequence). The direct-axis voltage component (V) and the quadrature-axis current reference Iq are determined by the amplitude of the second compensation sequence, which is used to inject anti-phase high-frequency damping current. The amplitude of the second compensation sequence is based on the amplitude of the mechanical resonant frequency. Determine, for example, take as AND The amplitude is proportional to the aforementioned compliance trigger signaling, and is limited by the safety margin of this amplitude, so that the injected reverse high-frequency damping current increases with the increase of shaft torsional vibration but does not exceed the mechanical boundary allowed by the unit shaft system, referenced by the direct-axis current. The three-phase current reference is synthesized from the quadrature-axis current reference Iq through an inverse transformation in the rotating coordinate system, that is:

[0123] ;

[0124] ;

[0125] ;

[0126] in, , , For three-phase current reference (A). Iq is the direct-axis current reference (A), and Iq is the quadrature-axis current reference. The three-phase current reference is the reconfigured digital command, which is the electrical phase angle (rad) determined by the aforementioned phase reference.

[0127] Using the previous example scenario, at the moment of large disturbance, the power deficit is approximately 35.3kW. Taking the disturbance time window as 0.5s and the average power deficit within the disturbance time window as approximately 30kW, the energy loss parameter is... J; The physical stress parameter is taken as the resonant amplitude at 23Hz. Suppose that the boundary of the envelope data matrix in the energy dimension corresponds to approximately... J. The boundary of the mechanical dimension corresponds to the allowable torsional amplitude of the shaft system, then the state point ( J, If the ray crossover parity determination finds the data to fall within the safe and feasible region, and the number of crossover points is odd, it is deemed compliant, and a compliance trigger signal is generated. Based on this, the first compensation sequence guides the energy storage converter to inject approximately 35kW of reverse active power to fill the gap, and the second compensation sequence guides the power electronic equipment to inject power at a frequency of 23Hz, with a phase phase consistent with the target harmonic. The opposite high-frequency damping currents are mapped to the direct axis and quadrature axis channels respectively. After vector orthogonal synthesis, they are combined into a three-phase current reference and used as the reconstructed digital command output.

[0128] return Figure 1 In step S4, a reconstructed digital command is output, and the updated bus frequency and updated rotor speed generated in response to the reconstructed digital command are obtained. Only when the drop slope of the updated bus frequency is greater than the frequency drop threshold, the microgrid operation data is updated using the updated bus frequency and updated rotor speed, and the power deficit sequence and mechanical resonant frequency are recalculated.

[0129] In this step, outputting the reconfiguration digital instruction means sending the three-phase current reference generated in step S3 to the current tracking control loop of fast response devices such as energy storage converters and photovoltaic inverters. The fast response devices track and execute the three-phase current reference, thereby injecting the active power corresponding to the first compensation sequence and the anti-phase high-frequency damping current corresponding to the second compensation sequence into the microgrid, and simultaneously compensating for the active power deficit caused by the water hammer effect and the alternating torque acting on the unit shaft system.

[0130] Obtaining the updated bus frequency and updated rotor speed generated by the response reconfiguration digital command means that after the fast response device executes the reconfiguration digital command and goes through a control cycle, the acquisition terminal in step S1 re-acquires the bus frequency and rotor speed. The acquired bus frequency and rotor speed are used as the updated bus frequency and updated rotor speed, respectively, to characterize the latest operating status of the microgrid under this compensation action.

[0131] Specifically, for the updated bus frequency, the bus frequency drop slope is recalculated using the same method as in step S2 to obtain the updated bus frequency drop slope. This updated bus frequency drop slope is then compared with the frequency drop threshold. If the updated bus frequency drop slope is not greater than the frequency drop threshold, it indicates that the frequency drop has been contained after this compensation and the microgrid is stabilizing. In this case, the current iteration ends and the current compensation output is maintained. If the updated bus frequency drop slope is still greater than the frequency drop threshold, it indicates that the microgrid's operating conditions continue to deteriorate and a single compensation is insufficient to contain the frequency drop. In this case, the next round of regulation begins.

[0132] When entering the next round of regulation, the microgrid operation data is updated using the updated bus frequency and updated rotor speed. That is, the corresponding quantities in the microgrid operation data are replaced with the updated bus frequency, updated rotor speed, and the synchronously re-acquired guide vane change rate sequence, stator current sequence, etc., as the input for the new round of calculation. Then, the power deficit sequence is recalculated according to step S2, and the mechanical resonant frequency and the target harmonic phase are re-extracted according to the updated rotor speed in step S2. After passing the double safety boundary verification in step S3, the reconstructed digital command is regenerated and output. This process of "output command - re-acquisition - calculate update drop slope - comparison and judgment - update data - regenerate command" is repeated round by round until the drop slope of the updated bus frequency is not greater than the frequency drop threshold, thus forming a closed-loop step correction of the compensation strategy.

[0133] It should be noted that, in order to avoid non-convergence of the iteration under extremely deteriorating operating conditions, the maximum number of iteration rounds or the minimum adjustment step size can be configured as supplementary iteration termination conditions. When the supplementary termination conditions are met, the compensation that has passed the safety boundary verification is output to balance the effectiveness of the compensation and the safety of the equipment.

[0134] After outputting the reconstructed digital command, the compensation effect needs to be verified to determine whether to continue adjusting. The reason for obtaining the updated bus frequency and updated rotor speed generated in response to the command after outputting the command, and using whether the drop slope of the updated bus frequency is still greater than the frequency drop threshold as the criterion for whether to continue iterating, is that a single feedforward compensation may not be sufficient to curb the frequency drop when the operating conditions continue to deteriorate. By re-judging with the updated data after compensation, and recalculating the power deficit sequence and extracting the mechanical resonant frequency using the updated data when the limit is still exceeded, a gradually approximating closed-loop correction can be formed, so that the compensation strategy is updated rollingly as the operating conditions evolve. Compared with the open-loop feedforward method with fixed gain and no compensation before judgment, this closed-loop step method has self-calibration capability when the operating conditions continue to deteriorate.

[0135] Using the aforementioned example scenario, after outputting the above-mentioned reconstructed digital instructions, the bus frequency rises from approximately 49.7Hz to approximately 49.9Hz, and the updated bus frequency drop slope decreases to approximately 1.2Hz / s, which is less than the frequency drop threshold of 2.5Hz / s. Based on this, the current iteration ends. If the updated bus frequency drop slope is still greater than the threshold, the microgrid operating data is updated with the updated bus frequency and the updated rotor speed. The power deficit sequence and mechanical resonant frequency are recalculated, and the above compensation process is repeated.

[0136] Based on the aforementioned compensation generated from the power deficit sequence, since the power deficit sequence is calculated by feedforward using the aforementioned hydrodynamic analytical formula, there may be a discrepancy between the active power degradation trend reflected by it and the active power degradation trend reflected by the measured bus frequency. To enhance compensation in a timely manner when the system approaches the edge of nonlinear collapse, the power deficit sequence is further modified after calculation, as follows: Figure 3 As shown, it also includes:

[0137] The power deficit sequence is differentially processed according to a preset time window to extract the power deficit change rate sequence. The specific formula is as follows:

[0138] ;

[0139] in, Let be the rate of change of power deficit at the i-th sampling time, which characterizes the acceleration (W / s) of the deterioration of active power in the microgrid system. This represents the total number of discrete sampling points included within a preset time window. The discrete sampling period time (s) is given. The power deficit sequence element value (W) is calculated at the i-th sampling time. For the first The power deficit sequence element value (W) corresponding to each sampling time (i.e., the starting time of a preset time window shifted forward from the current time);

[0140] Furthermore, considering the sliding delay of the preset time window, when the current sampling time number At this point, since the system has not yet experienced any disturbances or the accumulated data has not exceeded the length of a single window, the program directly records the corresponding historical moment that has exceeded the limit. The power deficit sequence element values ​​are configured to zero (i.e., 0W) for extraction;

[0141] In this step, the power deficit sequence is differentially processed according to a preset time window. The resulting power deficit change rate sequence reflects how fast the power deficit deteriorates over time, i.e., the acceleration of active power deterioration. The preset time window contains the total number of discrete sampling points. It is determined by the inertial time constant of the water flow and the discrete sampling period, that is:

[0142] ;

[0143] in, The preset water flow inertia time constant, The discrete sampling period is used to match the time span of the differential window with the transient time scale of the water hammer effect, so that the power deficit rate sequence can reflect the power degradation trend within the water hammer effect range. The reason for introducing the power deficit rate sequence is to provide an active power degradation trend quantity corresponding to the frequency degradation trend when comparing it with the bus frequency drop slope later.

[0144] The power deficit rate of change sequence is time-aligned with the bus frequency drop slope, and the numerical difference at the same time point is calculated to construct the difference sequence. The specific formula is as follows:

[0145] ;

[0146] in, The numerical difference (W / s) is constructed at the same time point. Let be the rate of change of power deficit at the i-th sampling time. The equivalent active power-frequency damping transformation coefficient characterizes the dynamic change in active power in a microgrid corresponding to a unit frequency sag slope. The slope of the bus frequency drop at the i-th sampling time (Hz / s) represents the bus frequency drop slope.

[0147] In this step, the power deficit change rate sequence is time-aligned with the bus frequency drop slope and the numerical difference at the same time node is calculated. The resulting difference sequence reflects the degree of deviation between the active power degradation rate calculated based on the feedforward of the water flow physical parameters and the active power degradation rate reflected by the actual bus frequency.

[0148] Equivalent active-frequency damping transformation coefficient It is used to convert the bus frequency drop slope into an active power quantity with the same dimension as the power deficit change rate. Its value is equal to the dynamic change of active power in the microgrid (W / Hz) when the bus frequency changes by a unit amount. The value of can be obtained from the historical small disturbance operation data of the microgrid. For example, a least-squares fit can be performed on multiple sets of bus frequency changes and corresponding active power changes, and the fitting slope can be taken as . .

[0149] The reason for constructing the above difference sequence is that when the active power degradation rate predicted by the hydraulic model feedforward deviates from the active power degradation rate reflected by the measured bus frequency, it indicates that the actual response of the system has deviated from the steady-state feedforward prediction trajectory, that is, the system is approaching the edge of nonlinear collapse. Using the difference between the two as the basis for subsequent additional compensation allows for timely enhancement of compensation when the deviation widens. Alternative methods include proportional or proportional-integral corrections based solely on the bus frequency deviation.

[0150] Based on a pre-defined numerical mapping relationship, the weight values ​​corresponding to the difference sequence are determined, including:

[0151] Extract the deviation magnitude of the difference sequence and determine the offset characteristic parameter of the deviation magnitude relative to the preset reference tolerance. The specific formula is as follows:

[0152] ;

[0153] in, Let be the offset feature parameter at the i-th sampling time. The magnitude of the deviation of the difference sequence (i.e. (absolute value) Preset reference tolerance (W / s);

[0154] In this step, the deviation magnitude (i.e., the absolute value of the difference) of the difference sequence is extracted, and its ratio with respect to the preset reference tolerance is used as the offset feature parameter. The resulting offset feature parameter reflects the multiple of the current deviation degree relative to the allowable deviation degree.

[0155] Preset reference tolerance The tolerance range of the difference sequence during normal operation of the microgrid is characterized by its value, which can be determined by the statistical fluctuation of the difference sequence during steady-state operation. For example, the standard deviation of the difference sequence during steady-state operation can be used as the tolerance range. (W / s); When the offset characteristic parameter is greater than 1, it indicates that the current divergence has exceeded the allowable fluctuation range;

[0156] The offset feature parameters are transformed using a preset composite mapping rule to determine the weight values. The composite mapping rule includes a proportional adjustment term and an exponential penalty term. The exponential penalty term uses the natural constant as the base and the offset feature parameter as the exponential mapping factor corresponding to the natural constant. The weight values ​​are constructed by jointly mapping the offset feature parameters using the proportional adjustment term and the exponential penalty term, as shown in the following formula:

[0157] ;

[0158] in, The weight value is determined at the i-th sampling time. This is the coefficient for the proportional adjustment term. For the exponential penalty term, Let be the offset feature parameter at the i-th sampling time.

[0159] In this step, a composite mapping rule containing a proportional adjustment term and an exponential penalty term is invoked to transform the offset feature parameter to determine the weight value. The proportional adjustment term makes the weight increase linearly with the offset feature parameter, corresponding to a gentle adjustment when the deviation is small. The exponential penalty term uses the natural constant as the base and the offset feature parameter as the exponent, making the weight increase exponentially with the offset feature parameter, corresponding to a rapid enhancement when the deviation is large.

[0160] Proportional adjustment term coefficient This is used to adjust the proportion of the proportional adjustment term relative to the exponential penalty term. Its value can be tuned according to the required compensation sensitivity. For example, when the offset characteristic parameter is equal to 1 (the deviation just reaches the reference tolerance), the contribution of the proportional adjustment term and the exponential penalty term to the weights reaches the desired proportion. Based on this, the calculation can be performed inversely. ;

[0161] The reason for introducing an exponential penalty term with a base of the natural constant in addition to the proportional adjustment term is that when only linear proportional adjustment is used, the increase in weight is insufficient to enhance compensation in time when the deviation is large. The exponential penalty term makes the weight change approximately linearly and smoothly when the deviation is small, and increases exponentially with the offset characteristic parameter when the deviation is large, so as to provide rapid support for the compensation when the system is close to the edge of nonlinear collapse.

[0162] Furthermore, the segmented action intervals of the aforementioned composite mapping rule can be determined by the relative magnitudes of the growth rates of the proportional adjustment term and the exponential penalty term: when the offset characteristic parameter is less than the natural logarithm of the proportional adjustment term coefficient, the incremental contribution of the proportional adjustment term to the weight value is greater than that of the exponential penalty term, and the weight value increases approximately linearly and gradually with the offset characteristic parameter, corresponding to an operating interval with a small degree of deviation and no need for reinforcement compensation; when the offset characteristic parameter is greater than the natural logarithm of the proportional adjustment term coefficient, the incremental contribution of the exponential penalty term exceeds that of the proportional adjustment term and quickly takes over, and the weight value increases sharply with the offset characteristic parameter, corresponding to an operating interval with a large degree of deviation, where the system is approaching the edge of nonlinear collapse and requires rapid reinforcement. Therefore, by tuning the proportional adjustment term coefficient, the boundary position between the two action intervals can be adjusted.

[0163] The difference sequence is weighted using weight values ​​to generate the first corrected sequence, as shown in the following formula:

[0164] ;

[0165] in, The element values ​​of the first correction sequence generated at the i-th sampling time represent the feedforward active power compensation (W) that must be forcibly added due to the system being on the verge of nonlinear collapse. The weight value is determined at the i-th sampling time. The numerical difference (W / s) is constructed at the same time point. This represents the preset basic compensation response time constant (s);

[0166] In this step, the difference sequence is weighted using weight values ​​and multiplied by the basic compensation response time constant to generate the first correction sequence, which represents the amount of feedforward active power compensation that must be forcibly added when the system approaches the edge of nonlinear collapse; the preset basic compensation response time constant... The equivalent response time required for a fast-response device to complete one basic compensation cycle is characterized by the equivalent bandwidth of the compensation control loop, i.e.:

[0167] ;

[0168] in, To compensate for the equivalent bandwidth (Hz) of the control loop;

[0169] Then, the first corrected sequence and the power deficit sequence are summarized to update the power deficit sequence;

[0170] In this step, the first correction sequence and the power deficit sequence are summarized to update the power deficit sequence. That is, the aforementioned forced additional feedforward active power compensation is superimposed on the original power deficit sequence, so that the updated power deficit sequence increases accordingly when the system approaches the edge of nonlinear collapse. Consequently, the first compensation sequence generated by the aforementioned steps is also enhanced. Through the above correction of the power deficit sequence, the compensation is added when the feedforward calculation trend based on hydraulic characteristics deviates from the measured trend of bus frequency, thereby enhancing the filling of active power deficit when the system approaches the edge of nonlinear collapse.

[0171] Furthermore, it should be noted that the method described in this application does not depend on the absolute size of the unit capacity. Parameters such as the water flow inertia time constant, equivalent rated volumetric flow rate constant, mechanical resonant frequency, and envelope data matrix involved in the method are all taken according to the actual parameters of the unit and equipment. Therefore, this method is also applicable to small test platforms or micro-power stations with capacities smaller than those in the example scenarios described above, such as multi-energy complementary off-grid microgrids composed of 15kW hydropower units, 5kW photovoltaic arrays, and kilowatt-level wind turbines. Moreover, the smaller the unit capacity, the smaller its equivalent inertia time constant is usually, and the greater the bus frequency drop slope caused by the same active power deficit. The proportion of active power reversal caused by water hammer effect in the transient process increases accordingly, thus making the suppression effect of this method on secondary frequency drops more significant. Correspondingly, in addition to photovoltaics, other distributed power sources such as wind power can also be connected to the microgrid. Their connection does not change the implementation method of this method; only the adjustable capacity of the corresponding power source needs to be included in the compensation capability of the aforementioned fast-response equipment.

[0172] In summary, by acquiring microgrid operation data, quantifying the power deficit caused by water hammer effect, extracting the mechanical resonant frequency and target harmonic phase, generating and outputting a reconstructed digital command synthesized by active power mirror compensation and anti-phase high-frequency damping under dual safety boundary verification, and performing closed-loop step correction on the compensation effect, the nonlinear active power deficit in the initial stage of turbine speed regulation and the suppression of the secondary frequency drop of the off-grid microgrid are filled, while the alternating torque caused by transient hydraulic shock on the unit shaft system is reduced. This helps to improve the transient power quality of the off-grid microgrid and the safe service of the core power generation equipment.

[0173] Example 2:

[0174] This is one embodiment of the present invention, which differs from the previous embodiment in that:

[0175] An off-grid energy management system for a multi-energy complementary microgrid system of small hydropower includes:

[0176] The data acquisition module is used to acquire microgrid operation data, which includes head pressure, bus frequency, guide vane change rate sequence, rotor speed sequence, and stator current sequence.

[0177] The calculation and extraction module is used to calculate the frequency drop slope of the bus. When the drop slope is greater than the preset frequency drop threshold, the power deficit sequence is calculated based on the guide vane change rate sequence, water head pressure and preset time constant. The mechanical resonant frequency is extracted according to the rotor speed sequence and the target harmonic phase is extracted according to the stator current sequence.

[0178] The instruction reconstruction module is used to generate a reconstructed digital instruction composed of a first compensation sequence and a second compensation sequence only when the integral value of the power deficit sequence and the amplitude of the mechanical resonant frequency both meet the preset safety boundary; wherein, the first compensation sequence is a mirror compensation sequence generated based on the power deficit sequence, and the second compensation sequence is a compensation sequence with the same frequency as the mechanical resonant frequency and the opposite phase to the target harmonic.

[0179] The closed-loop control module is used to output reconfiguration digital commands, obtain the updated bus frequency and updated rotor speed generated in response to the reconfiguration digital commands, and update the microgrid operation data only when the drop slope of the updated bus frequency is greater than the frequency drop threshold, and recalculate the power deficit sequence and extract the mechanical resonant frequency.

[0180] Example 3:

[0181] In one embodiment of the present invention, which differs from the previous embodiment, the electronic device includes one or more processors and a memory.

[0182] A processor can be a central processing unit (CPU) or other form of processing unit with data processing and / or instruction execution capabilities, and can control other components in an electronic device to perform desired functions.

[0183] The memory may include one or more computer program products, which may include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. Volatile memory may include, for example, random access memory (RAM) and / or cache memory. Non-volatile memory may include, for example, read-only memory (ROM), hard disk, flash memory, etc.

[0184] In one example, the electronic device may also include input devices and output devices, which are interconnected via a bus system and / or other forms of connection mechanisms (not shown). In addition, depending on the specific application, the electronic device may include any other suitable components.

[0185] Example 4:

[0186] Embodiments of this application may also be computer-readable storage media storing computer program instructions thereon, which, when executed by a processor, cause the processor to perform the steps described in the "Embodiment 1" section of this specification according to the various embodiments of this application.

[0187] Computer-readable storage media may take the form of any combination of one or more readable media. A readable medium may be a readable signal medium or a readable storage medium. A readable storage medium may, for example, include, but is not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatuses, or devices, or any combination thereof. More specific examples of readable storage media (a non-exhaustive list) include: electrical connections having one or more wires, portable disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0188] The basic principles of this application have been described above with reference to specific embodiments. However, it should be noted that the advantages, benefits, and effects mentioned in this application are merely examples and not limitations, and should not be considered as essential features of each embodiment of this application. Furthermore, the specific details disclosed above are for illustrative and facilitative purposes only, and are not limitations. These details do not restrict the application from being implemented using the specific details described above.

[0189] The block diagrams of devices, apparatuses, devices, and systems involved in this application are merely illustrative examples and are not intended to require or imply that they must be connected, arranged, or configured in the manner shown in the block diagrams. As those skilled in the art will recognize, these devices, apparatuses, devices, and systems can be connected, arranged, and configured in any manner. Words such as “comprising,” “including,” “having,” etc., are open-ended terms meaning “including but not limited to,” and are used interchangeably with them. The terms “or” and “and” as used herein refer to the terms “and / or,” and are used interchangeably with them unless the context clearly indicates otherwise. The term “such as” as used herein refers to the phrase “such as but not limited to,” and is used interchangeably with it.

[0190] It should also be noted that in the apparatus, equipment, and methods of this application, the components or steps can be disassembled and / or recombined. These disassemblies and / or recombinations should be considered as equivalent solutions of this application.

[0191] The above description of the disclosed aspects is provided to enable any person skilled in the art to make or use this application. Various modifications to these aspects will be readily apparent to those skilled in the art, and the general principles defined herein can be applied to other aspects without departing from the scope of this application. Therefore, this application is not intended to be limited to the aspects shown herein, but rather to be accorded the widest scope consistent with the principles and novel features disclosed herein.

[0192] The above description has been given for purposes of illustration and description. Furthermore, this description is not intended to limit the embodiments of this application to the forms disclosed herein. Although numerous exemplary aspects and embodiments have been discussed above, those skilled in the art will recognize certain variations, modifications, alterations, additions, and sub-combinations thereof.

Claims

1. An energy management method for an off-grid system of a multi-energy complementary microgrid in small hydropower, characterized in that, include: Acquire microgrid operation data, which includes head pressure, bus frequency, guide vane change rate sequence, rotor speed sequence, and stator current sequence; Calculate the frequency drop slope of the busbar. When the frequency drop slope is greater than a preset frequency drop threshold, calculate the power deficit sequence based on the guide vane change rate sequence, the water head pressure, and a preset time constant. Extract the mechanical resonant frequency based on the rotor speed sequence and extract the target harmonic phase based on the stator current sequence. A reconstructed digital instruction, synthesized by superimposing a first compensation sequence and a second compensation sequence, is generated only when the integral value of the power deficit sequence and the amplitude of the mechanical resonant frequency both satisfy a preset safety boundary; wherein, the first compensation sequence is a mirror compensation sequence generated based on the power deficit sequence, and the second compensation sequence is a compensation sequence with a frequency equal to the mechanical resonant frequency and a phase opposite to the target harmonic. The generation of the reconstructed digital instructions includes: Extract the energy loss parameter corresponding to the integral value of the power deficit sequence, and extract the physical stress parameter corresponding to the amplitude of the mechanical resonant frequency; The energy loss parameter and the physical stress parameter are combined to construct a two-dimensional state coordinate vector; The envelope data matrix corresponding to the preset security boundary is invoked, and the spatial topology mapping of the two-dimensional state coordinate vector relative to the envelope data matrix is ​​performed to generate compliance triggering signaling; In response to the compliance trigger signaling, the first compensation sequence is mapped to the direct axis channel of the preset orthogonal coordinate system to construct a direct axis data component, and the second compensation sequence is mapped to the cross axis channel of the preset orthogonal coordinate system to construct a cross axis data component; Perform a vector orthogonal synthesis transformation on the direct-axis data components and the quadrature-axis data components to generate the reconstructed digital instructions; Output the reconstructed digital command, obtain the updated bus frequency and updated rotor speed generated in response to the reconstructed digital command, and update the microgrid operation data using the updated bus frequency and the updated rotor speed only when the drop slope of the updated bus frequency is greater than the frequency drop threshold, and recalculate the power deficit sequence and extract the mechanical resonant frequency.

2. The energy management method for a small hydropower multi-energy complementary microgrid off-grid system according to claim 1, characterized in that, After calculating the power deficit sequence, the following is also included: The power deficit sequence is subjected to differential operation according to a preset time window to extract the power deficit change rate sequence; The power deficit change rate sequence is time-aligned with the bus frequency drop slope, and the numerical difference at the same time point is calculated to construct a difference sequence. Based on a preset numerical mapping relationship, the weight values ​​corresponding to the difference sequence are determined; The difference sequence is weighted using the weight values ​​to generate a first corrected sequence. The first corrected sequence is then combined with the power deficit sequence to update the power deficit sequence.

3. The energy management method for a small hydropower multi-energy complementary microgrid off-grid system according to claim 2, characterized in that, Determining the weight values ​​corresponding to the difference sequence includes: Extract the deviation magnitude of the difference sequence and determine the offset characteristic parameter of the deviation magnitude relative to a preset reference tolerance; The offset feature parameter is transformed by calling a preset composite mapping rule to determine the weight value. The composite mapping rule includes a proportional adjustment term and an exponential penalty term. The exponential penalty term is configured with a natural constant as the base and the offset feature parameter is configured as the exponential mapping factor corresponding to the natural constant. The weight value is constructed by jointly mapping the offset feature parameter with the proportional adjustment term and the exponential penalty term.

4. The energy management method for a small hydropower multi-energy complementary microgrid off-grid system according to claim 1, characterized in that, The generation of compliance triggering signaling includes: Extract the sequence of outer topological boundary nodes from the envelope data matrix; Determine the shortest spatial projection vector of the two-dimensional state coordinate vector relative to the sequence of external topological boundary nodes; Construct a test ray vector that starts from the two-dimensional state coordinate vector and extends along the shortest spatial projection vector, and determine the cross-traversal dataset of the test ray vector and the sequence of external topological boundary nodes; Extract the parity mapping attribute of the cross-traversal dataset, and structurally concatenate the parity mapping attribute with the modulus data of the shortest spatial projection vector to generate the compliance trigger signaling.

5. The energy management method for a small hydropower multi-energy complementary microgrid off-grid system according to claim 1, characterized in that, The extraction of the mechanical resonant frequency and the extraction of the target harmonic phase include: Perform phase space delay reconstruction on the rotor speed sequence to construct a multidimensional trajectory sequence set; Extract the dominant feature vector of the multidimensional trajectory sequence set, and transform the dominant feature vector into the frequency domain data space to determine the corresponding dominant frequency feature value as the mechanical resonant frequency; A quadrature phase-locked reference sequence is generated based on the mechanical resonant frequency; Demodulation projection transformation is performed on the stator current sequence and the orthogonal phase-locked reference sequence to separate the target harmonic data segment; Extract the phase offset attribute of the target harmonic data segment and configure it as the target harmonic phase.

6. The energy management method for a small hydropower multi-energy complementary microgrid off-grid system according to claim 5, characterized in that, The separated target harmonic data segment includes: The orthogonal phase-locked reference sequence is analyzed to separate the in-phase reference sequence and the orthogonal reference sequence; A stepwise correlation mapping is performed between the stator current sequence and the in-phase reference sequence to generate an in-phase modulation sequence, and a stepwise correlation mapping is performed between the stator current sequence and the orthogonal reference sequence to generate an orthogonal modulation sequence. By calling preset low-pass filter truncation parameters, convolution smoothing is performed on the in-phase modulation sequence and the quadrature modulation sequence respectively, and the corresponding in-phase baseband feature sequence and quadrature baseband feature sequence are extracted. The in-phase baseband feature sequence and the orthogonal baseband feature sequence are vectorized and concatenated to construct the target harmonic data segment.

7. An energy management system for a small hydropower multi-energy complementary microgrid off-grid system, characterized in that, include: The data acquisition module is used to acquire microgrid operation data, which includes head pressure, bus frequency, guide vane change rate sequence, rotor speed sequence, and stator current sequence. The calculation and extraction module is used to calculate the frequency drop slope of the bus. When the frequency drop slope is greater than the preset frequency drop threshold, the power deficit sequence is calculated based on the guide vane change rate sequence, the water head pressure, and the preset time constant. The mechanical resonant frequency is extracted according to the rotor speed sequence, and the target harmonic phase is extracted according to the stator current sequence. The instruction reconstruction module is used to generate a reconstruction digital instruction synthesized by superimposing a first compensation sequence and a second compensation sequence only when the integral value of the power deficit sequence and the amplitude of the mechanical resonant frequency both meet the preset safety boundary; wherein, the first compensation sequence is a mirror compensation sequence generated based on the power deficit sequence, and the second compensation sequence is a compensation sequence with the same frequency as the mechanical resonant frequency and the opposite phase to the target harmonic. The generation of the reconstructed digital instructions includes: Extract the energy loss parameter corresponding to the integral value of the power deficit sequence, and extract the physical stress parameter corresponding to the amplitude of the mechanical resonant frequency; The energy loss parameter and the physical stress parameter are combined to construct a two-dimensional state coordinate vector; The envelope data matrix corresponding to the preset security boundary is invoked, and the spatial topology mapping of the two-dimensional state coordinate vector relative to the envelope data matrix is ​​performed to generate compliance triggering signaling; In response to the compliance trigger signaling, the first compensation sequence is mapped to the direct axis channel of the preset orthogonal coordinate system to construct a direct axis data component, and the second compensation sequence is mapped to the cross axis channel of the preset orthogonal coordinate system to construct a cross axis data component; Perform a vector orthogonal synthesis transformation on the direct-axis data components and the quadrature-axis data components to generate the reconstructed digital instructions; The closed-loop control module is used to output the reconfiguration digital command, obtain the updated bus frequency and updated rotor speed generated in response to the reconfiguration digital command, and update the microgrid operation data using the updated bus frequency and updated rotor speed only when the drop slope of the updated bus frequency is greater than the frequency drop threshold, and recalculate the power deficit sequence and extract the mechanical resonant frequency.

8. An electronic device comprising a memory and a processor, characterized in that: The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions, which, when executed by the processor, implement the steps of the method as described in any one of claims 1 to 6.

9. A computer-readable storage medium storing computer-executable instructions thereon, characterized in that: When the computer-executable instructions are executed by a processor, they implement the steps of the method as described in any one of claims 1 to 6.

Images