Thermal power generating unit weight distribution method and device based on power grid frequency adjustment and medium

Abstract

The invention relates to the technical field of thermal power generating units, and discloses a thermal power generating unit weight distribution method and device based on power grid frequency adjustment and a medium, and the method comprises the steps: collecting historical frequency change rate data of a power grid and generated power prediction data of new energy; historical operation data of a distributed system in each thermal power generating unit are collected, wherein the operation data comprise pressure deviation and load rate; performing frequency division processing on the historical frequency change rate data through a filter to obtain a frequency division signal; determining an inertia demand of the power grid based on the frequency division signal; obtaining the valve abrasion loss of each thermal power generating unit, and determining the equipment health index of each thermal power generating unit based on the valve abrasion loss; generating a frequency modulation capability coefficient of each thermal power generating unit based on the equipment health index and the historical operation data; and determining a weight distribution coefficient of each thermal power generating unit based on the frequency modulation capability coefficient, the inertia demand and the equipment health index.

Description

Technical Field

[0001] This specification relates to the technical field of thermal power generation units, and in particular to a method, device, and medium for weight distribution of thermal power generation units based on grid frequency adjustment. Background Art

[0002] Thermal power units play a vital role in the stable operation of power systems. As baseload power sources, they not only provide a continuous and stable power supply to the grid but also shoulder the critical task of ensuring grid frequency stability. The regulation capability of thermal power units—that is, their ability to quickly respond to and effectively control grid frequency fluctuations—is crucial for ensuring the safe and economical operation of the grid.

[0003] However, the currently used weight allocation methods for thermal power units often rely on pre-defined regulation parameters. This model lacks flexibility and is difficult to adapt to real-time fluctuations in grid frequency. Because it cannot respond promptly to instantaneous changes in grid frequency, existing weight allocation methods are prone to delayed regulation response, which can lead to grid instability, increase system risks, and affect the economic efficiency of the power system. Summary of the Invention

[0004] One or more embodiments of this specification provide a method, device, and medium for weight distribution of thermal power units based on grid frequency adjustment, which are used to solve the technical problems raised in the background technology.

[0005] One or more embodiments of this specification adopt the following technical solutions:

[0006] One or more embodiments of this specification provide a method for weight allocation of thermal power generation units based on grid frequency adjustment, the method comprising:

[0007] Collect historical frequency change rate data of the power grid and forecast data of new energy power generation;

[0008] Collecting historical operating data of the distributed system in each thermal power unit, the operating data including pressure deviation and load rate;

[0009] Performing frequency division processing on the historical frequency change rate data through a filter to obtain a frequency division signal;

[0010] determining an inertia requirement of the power grid based on the divided frequency signal;

[0011] Obtaining valve wear of each of the thermal power units, and determining an equipment health index of each of the thermal power units based on the valve wear;

[0012] generating a frequency regulation capability coefficient of each of the thermal power units based on the equipment health index and the historical operation data;

[0013] Based on the frequency regulation capability coefficient, the inertia requirement and the equipment health index, a weight distribution coefficient of each of the thermal power units is determined.

[0014] It should be noted that the embodiments of this specification have the following beneficial effects through the above content:

[0015] First, by collecting historical frequency change rate data of the power grid and forecast data of new energy power generation, this method can grasp the operation status of the power grid in real time and provide accurate data support for weight allocation.

[0016] Secondly, by collecting the historical operating data of the distributed system in each thermal power unit, including pressure deviation and load rate, we can fully understand the operating status of the thermal power unit and provide a basis for weight allocation.

[0017] Thirdly, the historical frequency change rate data and the power generation forecast data are frequency-divided by a filter to obtain a frequency-divided signal, which helps to more accurately determine the inertia requirement of the power grid.

[0018] In addition, obtaining the valve wear of each thermal power unit and determining the equipment health index of each thermal power unit based on the valve wear will help evaluate the operating status of the thermal power units and provide strong support for weight allocation.

[0019] Based on the equipment health index and historical operating data, the frequency regulation capability coefficient of each thermal power unit is generated, which can more accurately reflect the regulation capability of the thermal power unit and provide a scientific basis for weight allocation.

[0020] Finally, based on the frequency regulation capability coefficient, inertia demand and equipment health index, the weight distribution coefficient of each thermal power unit is determined, which can realize the dynamic adjustment of the weight of the thermal power unit and improve the response speed and accuracy of the grid frequency adjustment, thereby effectively reducing system risks and improving the economic efficiency of the power system.

[0021] Furthermore, determining the inertia requirement of the power grid based on the divided frequency signal includes:

[0022] Analyzing the frequency-divided signal to determine the influence coefficients of different frequency ranges on the stability of the power grid;

[0023] Determine the inertia requirements of the power grid corresponding to the different frequency ranges according to the influence coefficients.

[0024] Furthermore, analyzing the frequency-divided signal to determine the influence coefficients of different frequency ranges on grid stability includes:

[0025] Performing time domain and frequency domain analysis on the frequency-divided signal to obtain components in different frequency ranges;

[0026] Obtaining pre-set stability indicators;

[0027] The correlation between the components in each frequency range and the stability index is determined, and the correlation is set as the influence coefficient.

[0028] Furthermore, determining the inertia requirements of the power grid corresponding to the different frequency ranges according to the influence coefficient includes:

[0029] Determining, based on the influence coefficient and a preset grid characteristic, a correlation between the inertia requirement and the components of each frequency range, wherein the correlation is a degree of influence of the components of each frequency range on the inertia requirement;

[0030] The inertia requirements of the components in each of the frequency ranges are determined based on the association relationship.

[0031] Furthermore, the obtaining of the valve wear of each thermal power unit and determining the equipment health index of each thermal power unit based on the valve wear include:

[0032] Collecting historical measurement data of the valves of each thermal power unit, wherein the measurement data includes wear depth, wear area and wear rate;

[0033] Analyzing historical measurement data of the valves of each thermal power unit to obtain a corresponding relationship between the historical valve wear amount and the equipment health index of each thermal power unit valve;

[0034] Establishing a relationship model between valve wear and equipment health index based on the corresponding relationship between the historical valve wear and equipment health index of each thermal power unit valve;

[0035] The valve wear amount of each thermal power unit is input into the relationship model to obtain the equipment health index of each thermal power unit.

[0036] Furthermore, generating the frequency regulation capability coefficient of each thermal power unit based on the equipment health index and the historical operation data includes:

[0037] Obtaining a preset mapping relationship between the equipment health index, the historical operating data, and a frequency regulation capability coefficient, and establishing a frequency regulation capability model of the thermal power unit based on the mapping relationship;

[0038] The equipment health index and the historical operation data are input into the frequency regulation capability model of the thermal power unit to generate a frequency regulation capability coefficient of each thermal power unit.

[0039] Furthermore, determining the weight distribution coefficient of each thermal power unit based on the frequency regulation capability coefficient, the inertia requirement and the equipment health index includes:

[0040] determining a first weight coefficient of each of the thermal power generating units according to the frequency regulation capability coefficient;

[0041] determining a second weight coefficient for each of the thermal power units according to the inertia requirement and the inertia that can be provided by each of the thermal power units;

[0042] determining a third weight coefficient for each of the thermal power units according to the equipment health index;

[0043] The first weight coefficient, the second weight coefficient and the third weight coefficient are processed based on a preset weight priority rule to determine a weight allocation coefficient for each of the thermal power generation groups.

[0044] Furthermore, after processing the first weight coefficient, the second weight coefficient, and the third weight coefficient based on a preset weight priority rule to determine the weight distribution coefficient of each thermal power unit, the method further includes:

[0045] Monitor and analyze the real-time operating data of the distributed systems in each thermal power unit in real time to obtain the frequency fluctuation trend;

[0046] Predict the state of the power grid through a pre-trained prediction model to obtain frequency fluctuations and load changes;

[0047] Based on the frequency fluctuation trend and the frequency fluctuation and load change results, the weight distribution coefficient of each thermal power unit is adjusted.

[0048] One or more embodiments of this specification provide a thermal power unit weight allocation device based on grid frequency adjustment, including:

[0049] at least one processor; and,

[0050] a memory communicatively connected to the at least one processor; wherein,

[0051] The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can: collect historical frequency change rate data of the power grid and power generation power forecast data of new energy; collect historical operation data of the distributed system in each thermal power unit, and the operation data includes pressure deviation and load rate; perform frequency division processing on the historical frequency change rate data through a filter to obtain a frequency division signal; determine the inertia requirement of the power grid based on the frequency division signal; obtain the valve wear amount of each thermal power unit, and determine the equipment health index of each thermal power unit based on the valve wear amount; generate the frequency regulation capability coefficient of each thermal power unit based on the equipment health index and the historical operation data; and determine the weight distribution coefficient of each thermal power unit based on the frequency regulation capability coefficient, the inertia requirement and the equipment health index.

[0052] One or more embodiments of the present specification provide a non-volatile computer storage medium storing computer-executable instructions, which, when executed by a computer, can achieve the following: collecting historical frequency change rate data of a power grid and power generation power forecast data of new energy sources; collecting historical operating data of a distributed system in each thermal power unit, wherein the operating data includes pressure deviation and load rate; performing frequency division processing on the historical frequency change rate data through a filter to obtain a frequency division signal; determining the inertia requirement of the power grid based on the frequency division signal; obtaining the valve wear amount of each thermal power unit, and determining the equipment health index of each thermal power unit based on the valve wear amount; generating a frequency regulation capability coefficient of each thermal power unit based on the equipment health index and the historical operating data; and determining a weight distribution coefficient of each thermal power unit based on the frequency regulation capability coefficient, the inertia requirement, and the equipment health index.

[0053] At least one of the above technical solutions adopted in the embodiments of this specification can achieve the following beneficial effects:

[0054] First, by collecting historical frequency change rate data of the power grid and forecast data of new energy power generation, this method can grasp the operation status of the power grid in real time and provide accurate data support for weight allocation.

[0055] Secondly, by collecting the historical operating data of the distributed system in each thermal power unit, including pressure deviation and load rate, we can fully understand the operating status of the thermal power unit and provide a basis for weight allocation.

[0056] Thirdly, the historical frequency change rate data and the power generation forecast data are frequency-divided by a filter to obtain a frequency-divided signal, which helps to more accurately determine the inertia requirement of the power grid.

[0057] In addition, obtaining the valve wear of each thermal power unit and determining the equipment health index of each thermal power unit based on the valve wear will help evaluate the operating status of the thermal power units and provide strong support for weight allocation.

[0058] Based on the equipment health index and historical operating data, the frequency regulation capability coefficient of each thermal power unit is generated, which can more accurately reflect the regulation capability of the thermal power unit and provide a scientific basis for weight allocation.

[0059] Finally, based on the frequency regulation capability coefficient, inertia demand and equipment health index, the weight distribution coefficient of each thermal power unit is determined, which can realize the dynamic adjustment of the weight of the thermal power unit and improve the response speed and accuracy of the grid frequency adjustment, thereby effectively reducing system risks and improving the economic efficiency of the power system. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] In order to more clearly illustrate the embodiments of this specification or the technical solutions in the prior art, the following briefly introduces the drawings required for the embodiments or the description of the prior art. Obviously, the drawings described below are only some of the embodiments described in this specification. For those skilled in the art, other drawings can be obtained based on these drawings without inventive work. In the drawings:

[0061] Figure 1 A flow chart of a method for allocating weights of thermal power units based on grid frequency adjustment provided in one or more embodiments of this specification;

[0062] Figure 2 A schematic structural diagram of a thermal power unit weight distribution device based on grid frequency adjustment provided in one or more embodiments of this specification. DETAILED DESCRIPTION

[0063] The embodiments of this specification provide a method, device, and medium for weight distribution of thermal power units based on grid frequency adjustment.

[0064] To help those skilled in the art better understand the technical solutions in this specification, the following will provide a clear and complete description of the technical solutions in the embodiments of this specification, in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of this specification, not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this specification without creative work should fall within the scope of protection of this specification.

[0065] Figure 1This is a flowchart of a method for assigning weights to thermal power units based on grid frequency adjustment, as provided in one or more embodiments of this specification. This process can be executed by a thermal power unit weight assignment system. Certain input parameters or intermediate results in the process can be manually adjusted to help improve accuracy.

[0066] The method steps of the embodiment of this specification are as follows:

[0067] S101, collecting historical frequency change rate data of the power grid and power generation power forecast data of new energy.

[0068] In the embodiments of this specification, the above-mentioned S101 can be implemented by the following specific implementation scheme:

[0069] Deploy high-precision frequency sensors: Install synchronized phasor measurement units (PMUs) at key grid nodes to record grid frequency change rate data in real time and store it in a central database.

[0070] Access to the new energy prediction platform: Integrate the power generation prediction systems of wind farms and photovoltaic power stations to obtain new energy output forecast data for the next 24 hours, including output fluctuation range and confidence level. The new energy prediction platform can be a neural network system.

[0071] Data normalization: Time alignment and normalization of historical frequency change rate data to eliminate noise interference caused by sampling frequency differences.

[0072] S102, collecting historical operating data of the distributed system in each thermal power unit, wherein the operating data includes pressure deviation and load rate.

[0073] Development of distributed system data interface: Through the distributed control system (DCS) interface, real-time collection of parameters such as pressure deviation, load rate, and main steam temperature of each thermal power unit.

[0074] In the embodiments of this specification, the above-mentioned S102 can be implemented by the following specific implementation scheme:

[0075] Data cleaning and storage: Eliminate abnormal values ​​(such as data with pressure deviation exceeding the safety threshold) and store them in a relational database by unit number to form a unit operation status file.

[0076] Historical data retrospective analysis: Extract the load rate change trend over the past year, combine it with the unit start and stop records, and establish a correlation model between load rate and pressure deviation.

[0077] S103 , performing frequency division processing on the historical frequency change rate data through a filter, and correlating the frequency division processed data with the generated power prediction data to obtain a frequency division signal.

[0078] In the embodiments of this specification, the above-mentioned S103 can be implemented by the following specific implementation scheme:

[0079] Multi-band filter design: Use digital filters (such as wavelet analysis or FIR filters) to perform frequency division processing on historical frequency change rate data to separate high-frequency fluctuation (0.1-2Hz) and low-frequency trend (<0.1Hz) signals.

[0080] S104: Determine the inertia requirement of the power grid based on the divided frequency signal.

[0081] In the embodiments of this specification, the above-mentioned S104 can be implemented by the following specific implementation scheme:

[0082] Inertia response model construction: Based on the high-frequency components of the frequency-divided signal, the grid inertia demand parameters are calculated, including the inertia time constant (H) and the frequency change rate (df / dt).

[0083] New energy output superposition verification: Combined with new energy forecast data, simulate the inertia gap under the scenario of sudden change in wind and solar output, and dynamically correct the inertia demand threshold.

[0084] Demand priority classification: Based on the frequency fluctuation amplitude and duration, inertia demand is divided into two categories: emergency frequency regulation (>0.2Hz / s) and regular frequency regulation (≤0.2Hz / s).

[0085] S105 , obtaining the valve wear amount of each thermal power unit, and determining the equipment health index of each thermal power unit based on the valve wear amount.

[0086] In the embodiments of this specification, the above-mentioned S105 can be implemented by the following specific implementation scheme:

[0087] Valve wear monitoring: Vibration sensors and acoustic emission technology are used to collect wear signals during valve opening and closing in real time, and the accumulated wear is calculated.

[0088] Health status quantification model: Establish a nonlinear mapping relationship between wear amount and equipment health index (such as exponential decay model), and set the health index range from 0 (severe wear) to 1 (intact).

[0089] Dynamic threshold warning: When the health index is lower than 0.7, a maintenance warning is triggered and the unit's frequency regulation output is limited to avoid equipment overload and damage.

[0090] S106: Generate a frequency regulation capability coefficient of each thermal power unit based on the equipment health index and the historical operation data.

[0091] In the embodiments of this specification, the above-mentioned S106 can be implemented by the following specific implementation scheme:

[0092] Multi-dimensional parameter fusion: Parameters such as equipment health index, historical pressure deviation mean, and load rate fluctuation are input into the weighted scoring model, and the weight distribution refers to the historical frequency regulation contribution of the unit.

[0093] Dynamic capability grading: For example, frequency modulation capability is divided into three levels based on the scoring results:

[0094] Level A (score ≥ 0.9): Prioritize high-frequency frequency modulation tasks;

[0095] Level B (0.7≤score<0.9): undertake routine frequency modulation tasks;

[0096] Level C (score < 0.7): only participates in low-frequency auxiliary frequency modulation.

[0097] Real-time update mechanism: The frequency modulation capacity coefficient is recalculated every 15 minutes to ensure data timeliness.

[0098] S107: Determine a weight distribution coefficient for each of the thermal power units based on the frequency regulation capability coefficient, the inertia requirement, and the equipment health index.

[0099] In the embodiments of this specification, the above-mentioned S107 can be implemented by the following specific implementation scheme:

[0100] Multi-objective optimization framework: The allocation model can be constructed with the frequency regulation capability coefficient as the primary weight (50%), the inertia demand matching as the secondary weight (30%), and the equipment health index as the constraint condition (20%).

[0101] The dynamic priority policy can be as follows:

[0102] In emergency frequency regulation scenarios, priority is given to units with high frequency regulation capability coefficients and health indexes ≥ 0.8;

[0103] In conventional frequency regulation scenarios, power is evenly distributed to Class A and Class B units, taking into account the life of the equipment.

[0104] Visualization of allocation results: The weight coefficient of each unit is displayed through the scheduling system interface, and manual intervention and adjustment are supported.

[0105] It should be noted that the embodiments of this specification have the following beneficial effects through the above content:

[0106] First, by collecting historical frequency change rate data of the power grid and forecast data of new energy power generation, this method can grasp the operation status of the power grid in real time and provide accurate data support for weight allocation.

[0107] Secondly, by collecting the historical operating data of the distributed system in each thermal power unit, including pressure deviation and load rate, we can fully understand the operating status of the thermal power unit and provide a basis for weight allocation.

[0108] Thirdly, the historical frequency change rate data and the power generation forecast data are frequency-divided by a filter to obtain a frequency-divided signal, which helps to more accurately determine the inertia requirement of the power grid.

[0109] In addition, obtaining the valve wear of each thermal power unit and determining the equipment health index of each thermal power unit based on the valve wear will help evaluate the operating status of the thermal power units and provide strong support for weight allocation.

[0110] Based on the equipment health index and historical operating data, the frequency regulation capability coefficient of each thermal power unit is generated, which can more accurately reflect the regulation capability of the thermal power unit and provide a scientific basis for weight allocation.

[0111] Finally, based on the frequency regulation capability coefficient, inertia demand and equipment health index, the weight distribution coefficient of each thermal power unit is determined, which can realize the dynamic adjustment of the weight of the thermal power unit and improve the response speed and accuracy of the grid frequency adjustment, thereby effectively reducing system risks and improving the economic efficiency of the power system.

[0112] Furthermore, when the embodiment of this specification determines the inertia requirement of the power grid based on the frequency division signal, the frequency division signal can be first analyzed to determine the influence coefficient of different frequency ranges on the stability of the power grid; and then, based on the influence coefficient, the inertia requirement of the power grid corresponding to the different frequency ranges is determined.

[0113] In the examples of this specification, the above content can be implemented by the following specific implementation schemes:

[0114] 1. Analysis of the impact of frequency division signal on power grid stability

[0115] 1. Multi-band division and feature extraction

[0116] The frequency-divided signal is divided into three typical frequency bands: high-frequency fluctuations (>0.5Hz), medium-frequency fluctuations (0.1-0.5Hz), and low-frequency trends (<0.1Hz). The amplitude, energy distribution, and duration characteristics of the signals in each frequency band are extracted. For example, high-frequency fluctuations primarily reflect sudden changes in renewable energy output and load switching disturbances, while low-frequency trends correspond to system inertial response characteristics.

[0117] 2. Impact coefficient quantification model

[0118] High-frequency impact coefficient: By statistically analyzing the instantaneous power change rate of high-frequency signals and combining it with historical frequency collapse cases of the power grid, a mapping relationship between the high-frequency fluctuation amplitude and the probability of transient instability of the system is established.

[0119] Intermediate frequency influence coefficient: Based on the correlation between the duration of intermediate frequency fluctuations and the system damping ratio, it quantifies their impact on oscillation stability (such as the damping ratio drop rate).

[0120] Low-frequency impact coefficient: Assess the degree to which the low-frequency signal weakens the inertia support capacity based on the relationship between the low-frequency signal energy ratio and the system inertia time constant.

[0121] 3. Dynamic weight allocation

[0122] The weight of each frequency band's impact coefficient is dynamically adjusted based on real-time operating conditions (such as renewable energy penetration and tie-line power ratio). For example, in scenarios with high wind power penetration, the weight of the high-frequency impact coefficient is increased to 60%.

[0123] 2. Determination of inertia requirements by levels

[0124] 1. Parameterization of frequency band inertia requirements

[0125] High-frequency inertia requirement: With the goal of suppressing the rate of change of frequency (RoCoF), the minimum required inertia support capacity is calculated based on the amplitude and duration of high-frequency fluctuations.

[0126] Medium-frequency inertia requirement: The medium-frequency inertia compensation value required to suppress oscillation is inferred through the damping ratio threshold constraint.

[0127] Low-frequency inertia requirement: Calculates the minimum inertia reference value required to maintain steady-state frequency based on the relationship between the system inertia time constant and the frequency drop depth.

[0128] 2. Multi-objective integration and optimization

[0129] Using a weighted superposition method, the inertia requirements of each frequency band are combined into a total inertia requirement based on the weight of the influence coefficient. Equipment health constraints (such as valve wear and frequency modulation response delay) are introduced to dynamically adjust the total inertia requirement.

[0130] 3. Real-time feedback and adaptive adjustment

[0131] Frequency fluctuations are monitored in real time through PMUs (synchronized phasor measurement units), which update the frequency-division signal signature library and trigger inertia demand recalculation. An inertia safety domain model is established, with upper and lower thresholds defined. When these thresholds are exceeded, backup units or energy storage compensation are automatically activated.

[0132] It should be noted that the embodiments of this specification have the following beneficial effects through the above content:

[0133] Improve grid stability: Accurately assessing and rationally allocating grid inertia requirements can effectively improve the stability of the power system and reduce the risk of power outages caused by grid fluctuations.

[0134] Optimize resource allocation: Rationally allocate inertia resources in the power system based on inertia requirements in different frequency ranges to improve resource utilization efficiency and reduce operating costs.

[0135] Supporting the development of the power market: Accurately assessing the grid inertia demand will help power market participants better understand the operation of the power system and provide strong support for power market transactions.

[0136] Furthermore, when analyzing the obtained divided frequency signal to determine the influence coefficient of different frequency ranges on the stability of the power grid, the divided frequency signal can be first analyzed in the time domain and frequency domain to obtain the components of the different frequency ranges; then a pre-set stability index is obtained; the correlation between the components of each frequency range and the stability index is determined, and the correlation is set as the influence coefficient.

[0137] In the examples of this specification, the above content can be implemented by the following specific implementation schemes:

[0138] 1. Joint Time-Frequency Analysis of Frequency-Divided Signals

[0139] 1. Multi-band time domain feature extraction

[0140] Using a sliding time window technique (with the window width dynamically adjusted by frequency band), we perform time-domain waveform analysis on the high-frequency (>0.5Hz), mid-frequency (0.1-0.5Hz), and low-frequency (<0.1Hz) components of the frequency-divided signal, extracting the instantaneous amplitude fluctuation rate and duration ratio of each frequency band signal (e.g., the peak-to-peak value of high-frequency fluctuations and the duration of low-frequency trends). Using the transient energy flow method, we can quantify the distribution characteristics of the energy in each frequency band along the time axis.

[0141] 2. Frequency domain characteristic mapping

[0142] Perform power spectral density analysis on each frequency band signal, plotting the frequency-energy distribution to identify the dominant frequency components and their harmonic distribution. Combined with amplitude response characteristics (such as gain margin and attenuation coefficient), a correlation model is established between frequency band energy concentration and the system's dynamic response capability. For example, high-frequency energy concentration may indicate the risk of transient instability caused by sudden changes in renewable energy output.

[0143] 2. Dynamic Matching of Stability Indicators

[0144] 1. Construction of multi-dimensional indicator library

[0145] Integrate core stability indicators of the power system:

[0146] Static stability: damping ratio threshold (e.g., critical mode damping ratio ≥ 3%), static stability margin (e.g., BIBO stability boundary).

[0147] Dynamic stability: phase margin, oscillation mode frequency offset.

[0148] Anti-interference capability: signal-to-noise ratio threshold, transient energy flow direction.

[0149] 2. Dynamic configuration of indicator priority

[0150] Dynamically adjust indicator weights based on the real-time operating status of the power grid (such as the penetration rate of new energy and the load rate of the main grid tie line):

[0151] In high wind power penetration scenarios, focus on monitoring the low-frequency damping ratio and mid-frequency phase margin;

[0152] In heavy-load interconnection line scenarios, the priority of the transient energy flow direction judgment criteria in the high-frequency band is increased.

[0153] 3. Correlation Modeling and Influence Coefficient Generation

[0154] 1. Multivariable coupling analysis

[0155] The grey correlation analysis method is used to calculate the correlation coefficient between the characteristic parameters of the signal in each frequency band (such as high-frequency amplitude fluctuation rate and low-frequency energy ratio) and the stability index (such as damping ratio and phase margin). For example:

[0156] The negative correlation between the high-frequency amplitude fluctuation rate and the transient stability margin (the greater the fluctuation, the lower the stability margin);

[0157] The duration of low-frequency energy is positively correlated with the static damping ratio (low-frequency energy accumulation reflects inertial support capacity).

[0158] 2. Dynamic calibration of influence coefficient

[0159] A three-dimensional influence coefficient matrix is ​​established, and the dimensions include frequency band (high / medium / low), indicator type (static / dynamic / anti-interference), and time scale (seconds / minutes).

[0160] For example, the coefficient matrix is ​​updated every 5 minutes through the transient energy flow online calculation platform:

[0161] High frequency band influence coefficient = 0.6×(transient stability correlation) + 0.4×(anti-interference correlation);

[0162] Low-frequency band influence coefficient = 0.7×(static stability correlation) + 0.3×(dynamic response correlation).

[0163] 3. Abnormal frequency band closed-loop feedback

[0164] When the influence coefficient of a certain frequency band exceeds the preset threshold (for example, the high-frequency band coefficient is >0.8), the multi-band stabilizer (PSS4B) adjustment strategy is automatically triggered: the problem frequency band signal is isolated through a bandpass filter; the excitation system gain parameters are adjusted to suppress high-frequency oscillation energy.

[0165] It should be noted that the embodiments of this specification have the following beneficial effects through the above content:

[0166] Improve the accuracy of grid stability analysis: By performing time and frequency domain analysis on the frequency-divided signal, it is possible to carefully identify components within different frequency ranges, thereby more accurately assessing the impact of these components on grid stability, providing a more accurate data basis for grid stability assessment.

[0167] Optimize stability index setting: By determining the correlation between different frequency range components and pre-set stability indicators, the selection and setting of stability indicators can be optimized, making the stability indicators more consistent with the operating characteristics of the actual power grid.

[0168] Dynamically adjust the regulation strategy: Based on the influence coefficient of different frequency ranges, the regulation strategy of the thermal power unit can be dynamically adjusted, making the regulation measures more targeted and able to more effectively respond to the challenges of different frequency components to the stability of the power grid.

[0169] Enhance the grid's ability to resist interference: By analyzing the impact coefficients of different frequency ranges, the frequency components that have the greatest impact on grid stability can be identified, so that corresponding measures can be taken to enhance the grid's resistance to these interferences and improve the overall stability of the grid.

[0170] Furthermore, when determining the inertia requirements of the power grid corresponding to the different frequency ranges based on the influence coefficient, a correlation between the inertia requirement and the components of each frequency range can be determined based on the influence coefficient and pre-set power grid characteristics, where the correlation represents the degree of influence of the components of each frequency range on the inertia requirement; and the inertia requirements of the components of each frequency range are determined based on the correlation.

[0171] In the examples of this specification, the above content can be implemented by the following specific implementation schemes:

[0172] 1. Multi-band division and characteristic mapping

[0173] 1. Frequency band division and feature extraction

[0174] The frequency-divided signal is divided into three typical frequency bands: high frequency (>0.5Hz), medium frequency (0.1-0.5Hz), and low frequency (<0.1Hz). The energy proportion, amplitude fluctuation rate, and duration characteristics of each frequency band are extracted. For example, the high frequency band reflects sudden changes in renewable energy output (such as short-term fluctuations in wind power), while the low frequency band corresponds to system inertia decay (such as the loss of inertia caused by the retirement of traditional units).

[0175] 2. Integration of grid characteristic parameters

[0176] Obtain real-time grid parameters: renewable energy penetration rate, proportion of synchronous units, and load dynamic characteristics (such as motor load proportion) as boundary conditions for calculating the impact coefficient.

[0177] Preset stability indicators: rate of change of frequency (RoCoF) threshold, damping ratio lower limit (for example, requiring the key mode damping ratio to be ≥3%).

[0178] 2. Construction of frequency band-inertia correlation model

[0179] 1. Dynamic calibration of influence coefficient

[0180] The grey correlation analysis method is used to calculate the correlation between the characteristics of each frequency band and the grid stability index;

[0181] High-frequency impact coefficient: Quantifies the threat to transient stability through the statistical relationship between the amplitude of high-frequency fluctuations and the number of RoCoF violations (for example, for every 10% increase in high-frequency fluctuations, the probability of RoCoF violations increases by 15%).

[0182] Low-frequency impact coefficient: Based on the inverse relationship between the low-frequency energy fraction and the equivalent inertia time constant, it assesses the degree to which the low-frequency energy fraction reduces the system inertia (for example, for every 5% increase in the low-frequency energy fraction, the equivalent inertia decreases by 0.2s).

[0183] Dynamically adjust weights based on grid characteristics: for example, when the new energy penetration rate is >30%, the high-frequency impact coefficient weight is increased to 60%.

[0184] 2. Matrixing of association relationships

[0185] Construct a frequency band-inertia requirement association matrix (three-dimensional table structure):

[0186] Dimension 1: frequency band (high / medium / low);

[0187] Dimension 2: grid characteristic parameters (new energy penetration rate, load type, etc.);

[0188] Dimension 3: Inertia demand impact weight (0-1 scale).

[0189] For example: when the new energy penetration rate is 40%, the inertia demand weight of the high-frequency band = 0.6, the low-frequency band = 0.3, and the medium-frequency band = 0.1.

[0190] 3. Hierarchical calculation and dynamic correction of inertia requirements

[0191] 1. Calculation of layered inertia requirements

[0192] High-frequency inertia requirement: To suppress RoCoF, the minimum virtual inertia compensation required is calculated based on the amplitude and duration of high-frequency fluctuations (for example, every 1 Hz / s of RoCoF requires 0.5 seconds of equivalent inertia).

[0193] Low-frequency inertia requirement: Based on the relationship between the low-frequency energy ratio and the system inertia time constant, the minimum inertia reference value required to maintain steady-state frequency is calculated (for example, if the low-frequency energy ratio is 20%, an additional 1.2s of energy storage inertia is required).

[0194] Mid-frequency inertia requirement: Combined with the damping ratio requirement, the inertia compensation is calculated using the mid-frequency phase margin indicator (for example, for every 5° decrease in phase margin, 0.3s of inertia needs to be added).

[0195] 2. Dynamic adjustment and safety verification

[0196] By introducing the concept of frequency change extreme time (TFM), the inertia requirement can be updated every 5 minutes. When the TFM is shortened (for example, from 8s to 5s), it means that the system inertia is decreasing and the inertia weight of the low-frequency band needs to be increased. When the high-frequency fluctuation amplitude suddenly increases, the virtual inertia control module (such as the additional power adjustment of the wind turbine) is triggered to respond first.

[0197] Safety zone verification: Set upper and lower thresholds for inertia demand (e.g. total inertia not less than 2s). When the limit is exceeded, the backup unit or energy storage compensation is started.

[0198] 4. Coordinated Allocation of Multi-Source Inertia Resources

[0199] 1. Resource Prioritization

[0200] Dynamic allocation can be performed based on device type, as follows:

[0201] Synchronous units: give priority to low-frequency inertia requirements (depending on physical rotational inertia);

[0202] Virtual inertia source (such as wind power, energy storage): responds to rapid fluctuations in high frequency bands (leveraging the speed control advantage of converters);

[0203] Motor load: Participates in mid-frequency damping adjustment through the equivalent inertia model.

[0204] 2. Economic optimization matching

[0205] The inertia cost-benefit model is established as follows:

[0206] Virtual inertia cost: charged according to the number of adjustments (e.g., 0.2 yuan / kWh for each second of inertia provided by energy storage);

[0207] Cost of synchronous units: taking into account fuel loss and equipment wear (e.g. the inertia support cost of coal-fired power units is 0.1 yuan / kWh).

[0208] The inertia cost-benefit model can dynamically select the combination of inertia sources (such as high-frequency fluctuations are responded to by energy storage and low-frequency fluctuations are supported by synchronous units) with the goal of minimizing total cost.

[0209] It should be noted that the embodiments of this specification have the following beneficial effects through the above content:

[0210] Accurate inertia requirement assessment: By combining the influence coefficient and pre-set grid characteristics, the correlation between inertia requirement and the components in each frequency range can be accurately determined, thereby achieving accurate assessment of grid inertia requirement.

[0211] Strong adaptability: This method can flexibly adjust the inertia configuration of the power grid according to the degree of influence of different frequency range components on the inertia demand, so as to adapt to the changes of different frequency components in the operation of the power grid and improve the adaptability of the power grid.

[0212] Optimize resource allocation: Based on a precise understanding of inertia requirements, the configuration of inertia resources can be optimized to ensure sufficient inertia support within the critical frequency range of the power grid, thereby improving the stability and reliability of the power grid.

[0213] Improved regulation response speed: By quickly determining the inertia requirements for each frequency range, thermal power units and other regulation resources can respond more quickly to changes in grid frequency, reducing frequency fluctuations and improving the dynamic stability of the grid.

[0214] Furthermore, when obtaining the valve wear of each thermal power unit and determining the equipment health index of each thermal power unit based on the valve wear, historical measurement data of the valves of each thermal power unit can be collected, and the measurement data include wear depth, wear area and wear rate; the historical measurement data of the valves of each thermal power unit are analyzed to obtain the corresponding relationship between the historical valve wear of the valves of each thermal power unit and the equipment health index; based on the corresponding relationship between the historical valve wear of the valves of each thermal power unit and the equipment health index, a relationship model between the valve wear and the equipment health index is established; the valve wear of each thermal power unit is input into the relationship model to obtain the equipment health index of each thermal power unit.

[0215] In the examples of this specification, the above content can be implemented by the following specific implementation schemes:

[0216] 1. Data Collection and Standardization

[0217] 1. Multi-dimensional wear parameter collection

[0218] Measurement tool deployment: Laser scanners and optical microscopes are installed at key valve locations (such as valve cores, seats, and sealing surfaces) in thermal power units to collect real-time data on wear depth, wear area, and wear rate. This data is then connected to the unit's DCS system to obtain operating parameters such as valve opening, opening and closing frequency, and medium pressure / temperature.

[0219] Data preprocessing: Use a low-pass filter to eliminate high-frequency noise interference (such as vibration signal interference), normalize the wear depth data, and eliminate the dimensional differences between different sensors.

[0220] 2. Historical data integration

[0221] A valve lifecycle database is built to integrate wear data from the past three years with equipment health records (such as maintenance logs and fault shutdown records) for the corresponding time period. This database is categorized by valve model and material (such as martensitic steel / austenitic steel), and correlated with operating parameters such as unit load rate and number of starts and stops.

[0222] 2. Historical Data Correlation Analysis

[0223] 1. Wear pattern recognition

[0224] Cluster analysis (such as K-means algorithm) can be used to divide the wear types as follows:

[0225] Uniform wear: Wear depth and area increase linearly, and the corresponding equipment health index decreases steadily;

[0226] Local erosion: The wear rate at a specific location changes suddenly (such as the valve seat sealing surface causing exponential wear due to high pressure difference erosion), and the health index drops sharply.

[0227] Compare the wear rate differences of valves made of different materials (such as cemented carbide and tungsten carbide coating) and establish a material hardness-wear resistance correlation table.

[0228] 2. Quantification of health index

[0229] Define the device health index (0-100%):

[0230] 100%: wear depth <0.1mm and no local erosion;

[0231] 70%: uniform wear depth 0.3mm or local erosion area >5%;

[0232] 50%: Triggering maintenance threshold (such as ultra-high pressure valve wear rate > 0.05mm / thousand hours).

[0233] Based on historical failure cases, a mapping relationship between the health index and the remaining life can be established (for example, for every 10% decrease in the health index, the remaining life is reduced by 2,000 hours).

[0234] 3. Relationship Model Construction and Verification

[0235] 1. Multivariate regression modeling

[0236] The key variables were selected: wear depth (primary variable), wear area (auxiliary variable), and medium pressure (modulating variable). Multiple linear regression analysis was used to calculate weight coefficients (e.g., wear depth contribution 60%, pressure fluctuation contribution 25%).

[0237] Introducing interaction analysis: Under high-pressure conditions (>10 MPa), the nonlinear effect of wear rate on health index needs to be described by a piecewise function.

[0238] 2. Dynamic baseline calibration

[0239] Establish a benchmark model for each unit type (such as supercritical / subcritical): Due to the high temperature and high pressure characteristics of supercritical units, the health index decays 30% faster than that of subcritical units.

[0240] The model parameters are dynamically updated using real-time data streams (e.g., collected every 15 minutes), and the impact of seasonal operating load fluctuations is eliminated through a sliding window algorithm.

[0241] It should be noted that the embodiments of this specification have the following beneficial effects through the above content:

[0242] Real-time equipment status monitoring: By collecting measurement data such as wear depth, wear area, and wear rate of thermal power unit valves, real-time monitoring of equipment wear can be achieved, and equipment wear problems can be discovered in a timely manner.

[0243] Equipment health assessment: Based on historical measurement data, the corresponding relationship between valve wear and equipment health index is established, which can quantitatively evaluate the health status of thermal power unit equipment and provide a scientific basis for equipment maintenance.

[0244] Improve equipment reliability: By continuously monitoring and evaluating equipment health indexes, we can ensure that key components of thermal power units are in good condition, thereby improving equipment reliability and operating life.

[0245] Data-driven decision-making: By establishing a relationship model between wear and equipment health index, data support can be provided for the operation and decision-making of thermal power units, making decisions more scientific and reasonable.

[0246] Furthermore, when generating the frequency regulation capability coefficient of each thermal power unit based on the equipment health index and the historical operating data, a mapping relationship between the pre-set equipment health index, the historical operating data, and the frequency regulation capability coefficient can be obtained, and a frequency regulation capability model of the thermal power unit can be established based on the mapping relationship; the equipment health index and the historical operating data are input into the frequency regulation capability model of the thermal power unit to generate the frequency regulation capability coefficient of each thermal power unit.

[0247] In the examples of this specification, the above content can be implemented by the following specific implementation schemes:

[0248] 1. Data Preparation and Feature Integration

[0249] 1. Standardization of equipment health index

[0250] The Equipment Health Index (EHI) assessment system integrates parameters such as valve wear depth, wear rate, and maintenance records for thermal power units, quantifying their health status on a uniform scale of 0-100%. Using a logistic regression model, the health index is correlated with failure rates to define different health levels (e.g., A for a health index ≥80%, B for 50%-80%, and C for <50%).

[0251] 2. Cleaning and classification of historical operation data

[0252] Collect historical operating data on thermal power units, including pressure deviation, load rate, and start / stop times, and remove outliers (e.g., data where pressure deviation exceeds safety thresholds). Categorize and store data by frequency modulation scenario, for example: high frequency modulation (response time <5 seconds), medium frequency modulation (5 seconds to 1 minute), and low frequency modulation (>1 minute).

[0253] 2. Frequency Modulation Capability Mapping Relationship Modeling

[0254] 1. Multi-dimensional parameter correlation analysis

[0255] The grey correlation analysis method is used to calculate the correlation between the equipment health index and indicators such as frequency modulation response rate and frequency modulation accuracy. For example:

[0256] The health index was positively correlated with FM response speed (for every 10% decrease in health index, the response delay increased by 15%);

[0257] The load factor is associated with the upper limit of the frequency regulation capacity (when the load factor is > 90%, the frequency regulation capacity decreases by 20%).

[0258] Combined with the frequency modulation control logic, key mapping dimensions are defined: health index-frequency modulation response time, pressure deviation-frequency modulation stability, and load rate-frequency modulation capacity.

[0259] 2. Dynamic Weight Allocation Model

[0260] According to the fuzzy control principle, the dynamic weight distribution rules are designed, which can be as follows:

[0261] High-frequency FM scenario: FM response speed accounts for 70% of the weight, and health index accounts for 30% of the weight;

[0262] Low-frequency frequency modulation scenario: The frequency modulation capacity weight accounts for 60%, and the pressure deviation weight accounts for 40%.

[0263] Introduce marginal substitution analysis to optimize the real-time economic efficiency of weight distribution (such as giving priority to units with high health index to undertake high-frequency frequency regulation tasks).

[0264] 3. Frequency Modulation Capability Model Construction

[0265] 1. Machine Learning Model Training

[0266] Based on the simulation model optimization method, random forest or neural network algorithm is used to input parameters such as equipment health index, pressure deviation mean, load rate fluctuation rate, etc., and output the frequency regulation capability coefficient (0-1 scale).

[0267] Divide the training set and validation set:

[0268] Training set: historical data from the past three years (80%), including typical frequency modulation events (such as frequency mutations and fluctuations in renewable energy output);

[0269] Validation set: data from the past 6 months (20%), used to verify the generalization ability of the model.

[0270] 2. Model dynamic calibration mechanism

[0271] The unit operation data is collected in real time through the SCADA system, and the model input parameters are updated every 15 minutes.

[0272] When the health index drops by more than 5% or the load rate suddenly changes by more than 10%, model retraining is triggered.

[0273] 4. Generation and Verification of Frequency Modulation Capacity Coefficient

[0274] 1. Real-time coefficient generation and grading

[0275] The current equipment health index and operating data are input into the model, and the frequency regulation capability coefficient is output, which is graded according to the following rules:

[0276] Level 1 frequency modulation capability (coefficient ≥ 0.9): Prioritizes high-frequency frequency modulation tasks (e.g., frequency fluctuation > 0.2 Hz / s);

[0277] Secondary frequency regulation capability (0.7≤coefficient<0.9): Participates in conventional frequency regulation (such as load following);

[0278] Level 3 frequency regulation capability (coefficient <0.7): only used as a backup frequency regulation resource.

[0279] The rationality of the coefficient is verified through the power adjustment formula (such as the linear matching of the frequency regulation capability coefficient and the power adjustment amount).

[0280] 2. Closed-loop feedback and optimization

[0281] Based on the actual frequency modulation effect (such as frequency recovery time and frequency modulation instruction execution deviation), the model parameters are reversed to form a dynamic optimization closed loop. The model prediction accuracy is evaluated regularly (e.g., monthly). If the deviation exceeds 10%, the feature weights or algorithm parameters are readjusted.

[0282] It should be noted that the embodiments of this specification have the following beneficial effects through the above content:

[0283] Optimize frequency regulation resource allocation: By generating frequency regulation capacity coefficients, frequency regulation resources can be allocated more efficiently, ensuring that the power grid has sufficient frequency regulation response capacity when needed.

[0284] Improve frequency regulation response speed: The frequency regulation capability model established based on equipment health index and historical operating data can quickly evaluate the frequency regulation capability of thermal power units, thereby accelerating frequency regulation response speed and enhancing grid stability.

[0285] Reducing unnecessary frequency regulation interventions: By pre-setting the mapping relationship between the equipment health index and the frequency regulation capability coefficient, unnecessary frequency regulation interventions for thermal power units in poor health can be avoided, reducing resource waste.

[0286] Improving frequency regulation quality: The generation of frequency regulation capability coefficients helps ensure the efficiency and accuracy of frequency regulation operations, thereby improving the frequency regulation quality of the power grid.

[0287] Furthermore, when determining the weight distribution coefficient of each of the thermal power groups based on the frequency regulation capability coefficient, the inertia requirement and the equipment health index, a first weight coefficient of each of the thermal power groups can be determined according to the frequency regulation capability coefficient; a second weight coefficient of each of the thermal power groups can be determined according to the inertia requirement and the inertia that each of the thermal power groups can provide; a third weight coefficient of each of the thermal power groups can be determined according to the equipment health index; and the first weight coefficient, the second weight coefficient and the third weight coefficient can be processed based on a pre-set weight priority rule to determine the weight distribution coefficient of each of the thermal power groups.

[0288] In the examples of this specification, the above content can be implemented by the following specific implementation schemes:

[0289] 1. Weight coefficient definition and data preparation

[0290] 1. Weight coefficient classification

[0291] The first weight coefficient (frequency regulation capability weight): Based on the frequency regulation capability coefficient, it reflects the unit's dynamic regulation capability in response to frequency fluctuations.

[0292] The second weight coefficient (inertia support weight) is determined based on the ratio of the unit's actual inertia provision capability to the system's total inertia requirement, and represents its support priority for grid stability.

[0293] The third weight coefficient (health status weight): is standardized to a scale of 0-1 through the Equipment Health Index (EHI). The higher the health index, the greater the weight, which reduces the risk of failure.

[0294] 2. Data standardization

[0295] Frequency regulation capability coefficient: normalized according to the maximum frequency regulation capacity of the unit (e.g. 6%-10% of the rated capacity).

[0296] Inertia requirement matching: Calculate the percentage of inertia provided by each unit to the total system demand (e.g., inertia percentage of a unit = unit inertia / total inertia demand).

[0297] Equipment Health Index: Integrates parameters such as valve wear depth and maintenance records, and quantifies them uniformly on a 0-100% scale.

[0298] 2. Dynamic calculation of weight coefficient

[0299] 1. Calculation of the first weight coefficient

[0300] Dynamic adjustment based on frequency modulation capability coefficient can be as follows:

[0301] High-frequency frequency modulation scenario (response time < 5 seconds): The weighting emphasizes frequency modulation speed. The coefficient = frequency modulation capacity coefficient × 0.7 + load factor × 0.3.

[0302] Low-frequency frequency modulation scenario (response time > 1 minute): The weight focuses on frequency modulation capacity. The coefficient = frequency modulation capacity coefficient × 0.6 + inertia matching × 0.4.

[0303] 2. Calculation of the second weight coefficient

[0304] Dynamic calibration based on the inertia supply and demand relationship can be performed as follows:

[0305] When the system inertia demand is urgent (such as when the new energy penetration rate is >30%), the coefficient = unit inertia ratio × emergency correction factor (1.2-1.5).

[0306] In conventional scenarios, the coefficient = unit inertia ratio × health index attenuation factor (the coefficient decreases by 0.1 for every 10% decrease in the health index).

[0307] 3. Calculation of the third weight coefficient

[0308] The health index classification mapping can be specifically as follows:

[0309] Level A (health index ≥ 80%): Coefficient = 1.0, high-load tasks are assigned first.

[0310] Level B (50%-80%): Coefficient = 0.6, limiting participation in high frequency modulation.

[0311] Level C (<50%): Coefficient = 0.3, only used as a backup resource.

[0312] 3. Weight Priority Rules and Fusion Strategies

[0313] 1. Priority rule setting

[0314] Emergency frequency modulation scenario: first weight (frequency modulation capability) > second weight (inertia) > third weight (health).

[0315] Steady-state operation scenario: the third weight (health) > the first weight (frequency regulation capability) > the second weight (inertia).

[0316] Economic optimization scenario: Introduce cost factors (such as virtual inertia cost of 0.2 yuan / kWh) and dynamically adjust weight priorities.

[0317] 2. Multi-objective fusion algorithm

[0318] Adopting adaptive weight matrix, the specific method is as follows:

[0319] A three-dimensional weight matrix is ​​constructed, and the dimensions include scenario type (emergency / steady state / economic), unit type (supercritical / subcritical), and time scale (seconds / minutes).

[0320] Calculate the comprehensive weight by grey relational analysis:

[0321] Comprehensive weight = α·W1+β·W2+γ·W3

[0322] Among them, α, β, and γ are dynamically adjusted parameters (such as α=0.6, β=0.3, and γ=0.1 in an emergency scenario), and W1, W2, and W3 are the first weight, the second weight, and the third weight, respectively.

[0323] 3. Dynamic correction mechanism

[0324] Real-time data feedback: PMU monitors frequency fluctuations and unit status, and can update weight coefficients every 5 minutes.

[0325] Safety domain verification: Set upper and lower limits for weight distribution (e.g., total weight coefficient ≥ 0.7). When the limit is exceeded, the backup unit or energy storage compensation is triggered.

[0326] It should be noted that the embodiments of this specification have the following beneficial effects through the above content:

[0327] Optimize resource utilization: By determining the weight allocation coefficients of thermal power units, frequency regulation resources can be allocated more reasonably, ensuring that the power grid can obtain frequency regulation services from the most appropriate thermal power units when needed.

[0328] Improve frequency regulation efficiency: Assigning weights based on the frequency regulation capability coefficient, inertia requirement, and equipment health index helps improve the efficiency of frequency regulation operations and reduce unnecessary frequency regulation actions.

[0329] Enhanced grid stability: By prioritizing thermal power units with strong frequency regulation capabilities, as well as units that can provide sufficient inertia and are in good equipment health, grid stability can be enhanced.

[0330] Furthermore, after processing the first weight coefficient, the second weight coefficient and the third weight coefficient based on the pre-set weight priority rule to determine the weight allocation coefficient of each of the thermal power groups, the real-time operation data of the distributed system in each thermal power group can be monitored and analyzed in real time to obtain the frequency fluctuation trend; the state of the power grid is predicted through a pre-trained prediction model to obtain frequency fluctuation and load change results; and the weight allocation coefficient of each of the thermal power groups is adjusted based on the frequency fluctuation trend and the frequency fluctuation and load change results.

[0331] In the examples of this specification, the above content can be implemented by the following specific implementation schemes:

[0332] 1. Real-time collection and monitoring of multi-source data

[0333] 1. Distributed system operation data collection

[0334] Acquisition parameters: Real-time acquisition of core parameters such as active power, frequency deviation, frequency modulation response time, fuel quantity command, and main steam pressure for each thermal power unit. By deploying high-precision sensors and PMUs (synchronized phasor measurement units), high-frequency data capture can be achieved at a sampling frequency of 12.8kHz.

[0335] Data preprocessing: The sliding window method is used to screen the data, distinguish between steady-state and dynamic processes, and eliminate noise interference through low-pass filtering.

[0336] 2. Frequency fluctuation trend analysis

[0337] Real-time spectrum calculation: Scan the voltage, current, and power signals in the full frequency range of 0-2500Hz to identify the dominant frequency components.

[0338] Trend prediction: Based on the calculation of frequency deviation based on the synchronized phasor algorithm, combined with historical frequency data and current unit load, the fluctuation trend in the next 5-15 minutes can be predicted.

[0339] 2. Prediction Model Construction and Grid State Prediction

[0340] 1. Multi-model prediction framework design

[0341] Model selection: A local transfer function model is established for different scenarios (such as high-energy penetration and load mutation) and converted into a CARIMA model.

[0342] Prediction input: Integrates real-time load data, inertia demand, and equipment health index as model inputs, and outputs key indicators such as frequency fluctuation amplitude, load change rate, and frequency regulation margin.

[0343] 2. Dynamic weight fusion prediction

[0344] Improved particle swarm optimization algorithm: introduce multi-objective optimal factors and priority coefficients to optimize the weight distribution of the prediction model.

[0345] Scenario adaptation: Dynamically adjust prediction model parameters according to the grid status (emergency frequency regulation, steady-state operation), for example, prioritize predicting high-frequency fluctuation risks in emergency scenarios.

[0346] 3. Dynamic Adjustment of Weight Distribution Coefficient

[0347] 1. Weight priority rule setting

[0348] Emergency frequency modulation scenario: The first weight (frequency modulation capability) accounts for 60%, the second weight (inertia support) accounts for 30%, and the third weight (equipment health) accounts for 10%.

[0349] Steady-state operation scenario: The third weight (equipment health) accounts for 50%, the first weight (frequency regulation capability) accounts for 30%, and the second weight (inertia support) accounts for 20%.

[0350] 2. Dynamic correction based on prediction results

[0351] Frequency fluctuation exceeding the limit: If the predicted frequency deviation is greater than 0.2 Hz / s, the priority of the first weight coefficient will be increased, and units with strong frequency regulation capabilities will be activated first.

[0352] Load mutation response: The load change rate is analyzed through a sliding time window. If a load mutation is predicted, such as greater than 10% of the rated power, the allocation ratio of the second weight coefficient (inertia support) can be dynamically increased.

[0353] 3. Closed-loop feedback mechanism

[0354] Frequency modulation effect evaluation: Real-time monitoring of frequency modulation instruction execution deviation. If the deviation persists, for example, greater than 5%, the weight distribution model parameters can be corrected in reverse.

[0355] Model online calibration: The prediction model weights can be updated every 15 minutes based on the latest data, using the random moment quadratic gradient optimization algorithm.

[0356] It should be noted that the embodiments of this specification have the following beneficial effects through the above content:

[0357] Real-time response capability: By real-time monitoring and analysis of the operating data of the distributed system in the thermal power units, it can quickly respond to frequency fluctuations and improve the adaptability and stability of the power grid to frequency fluctuations.

[0358] Predictive maintenance: Real-time monitoring of frequency fluctuation trends helps predict potential equipment failures or maintenance needs, allowing preventive maintenance to be implemented and reduce unplanned downtime.

[0359] Dynamic weight adjustment: Based on the prediction results of frequency fluctuations and load changes, the weight allocation coefficients of thermal power units can be dynamically adjusted to ensure that the power grid can quickly adjust resources when load changes or frequency fluctuations occur.

[0360] Figure 2 A structural schematic diagram of a thermal power unit weight distribution device based on grid frequency adjustment provided for one or more embodiments of this specification includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to: collect historical frequency change rate data of the power grid and power generation power forecast data of new energy; collect historical operation data of the distributed system in each thermal power unit, the operation data including pressure deviation and load rate; perform frequency division processing on the historical frequency change rate data through a filter to obtain a frequency division signal; determine the inertia requirement of the power grid based on the frequency division signal; obtain the valve wear amount of each thermal power unit, and determine the equipment health index of each thermal power unit based on the valve wear amount; generate a frequency regulation capability coefficient of each thermal power unit based on the equipment health index and the historical operation data; and determine the weight distribution coefficient of each thermal power unit based on the frequency regulation capability coefficient, the inertia requirement and the equipment health index.

[0361] One or more embodiments of the present specification provide a non-volatile computer storage medium storing computer-executable instructions, which, when executed by a computer, can achieve the following: collecting historical frequency change rate data of a power grid and power generation power forecast data of new energy sources; collecting historical operating data of a distributed system in each thermal power unit, wherein the operating data includes pressure deviation and load rate; performing frequency division processing on the historical frequency change rate data through a filter to obtain a frequency division signal; determining the inertia requirement of the power grid based on the frequency division signal; obtaining the valve wear amount of each thermal power unit, and determining the equipment health index of each thermal power unit based on the valve wear amount; generating a frequency regulation capability coefficient of each thermal power unit based on the equipment health index and the historical operating data; and determining a weight distribution coefficient of each thermal power unit based on the frequency regulation capability coefficient, the inertia requirement, and the equipment health index.

[0362] The various embodiments in this specification are described in a progressive manner. Similar portions between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from the other embodiments. In particular, the device, apparatus, and non-volatile computer storage medium embodiments are generally similar to the method embodiments, so their descriptions are relatively simplified. For relevant details, refer to the descriptions of the method embodiments.

[0363] The various embodiments in this specification are described in a progressive manner. Similar parts between the various embodiments can be referred to in conjunction with each other. Each embodiment focuses on the differences from other embodiments. In particular, the device embodiments are generally similar to the method embodiments, so the description is relatively simple. For relevant parts, refer to the description of the method embodiments.

[0364] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0365] In the embodiments provided in this application, it should be understood that the disclosed devices / network equipment and methods can be implemented in other ways. For example, the device / network equipment embodiments described above are merely illustrative. For example, the division of the modules or units is merely a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0366] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0367] In addition, the functional units in the various embodiments of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above units may be implemented in the form of hardware or software.

[0368] If the integrated module / unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present application implements all or part of the process in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and when the computer program is executed by the processor, it can implement the steps of the above-mentioned various method embodiments. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electric carrier signal, telecommunication signal and software distribution medium. It should be noted that the content contained in the computer-readable medium can be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, computer-readable media do not include electric carrier signals and telecommunication signals.

Claims

1. A method for weight allocation of thermal power units based on grid frequency adjustment, characterized in that: include: Collect historical frequency change rate data of the power grid and forecast data of new energy power generation; Collecting historical operating data of the distributed system in each thermal power unit, the operating data including pressure deviation and load rate; Performing frequency division processing on the historical frequency change rate data through a filter to obtain a frequency division signal; determining an inertia requirement of the power grid based on the divided frequency signal; Obtaining valve wear of each of the thermal power units, and determining an equipment health index of each of the thermal power units based on the valve wear; generating a frequency regulation capability coefficient of each of the thermal power units based on the equipment health index and the historical operation data; Based on the frequency regulation capability coefficient, the inertia requirement and the equipment health index, a weight distribution coefficient of each of the thermal power units is determined.

2. The method according to claim 1, characterized in that The determining the inertia requirement of the power grid based on the divided frequency signal includes: Analyzing the frequency-divided signal to determine the influence coefficients of different frequency ranges on the stability of the power grid; Determine the inertia requirements of the power grid corresponding to the different frequency ranges according to the influence coefficients.

3. The method according to claim 2, characterized in that The analyzing the frequency division signal to determine the influence coefficients of different frequency ranges on the grid stability includes: Performing time domain and frequency domain analysis on the frequency-divided signal to obtain components in different frequency ranges; Obtaining pre-set stability indicators; The correlation between the components in each frequency range and the stability index is determined, and the correlation is set as the influence coefficient.

4. The method according to claim 2, characterized in that The determining, according to the influence coefficient, the inertia requirements of the power grid corresponding to the different frequency ranges includes: Determining, based on the influence coefficient and a preset grid characteristic, a correlation between the inertia requirement and the components of each frequency range, wherein the correlation is a degree of influence of the components of each frequency range on the inertia requirement; The inertia requirements of the components in each of the frequency ranges are determined based on the association relationship.

5. The method according to claim 1, wherein The obtaining of the valve wear amount of each thermal power unit and determining the equipment health index of each thermal power unit based on the valve wear amount includes: Collecting historical measurement data of the valves of each thermal power unit, wherein the measurement data includes wear depth, wear area and wear rate; Analyzing historical measurement data of the valves of each thermal power unit to obtain a corresponding relationship between the historical valve wear amount and the equipment health index of each thermal power unit valve; Establishing a relationship model between valve wear and equipment health index based on the corresponding relationship between the historical valve wear and equipment health index of each thermal power unit valve; The valve wear amount of each thermal power unit is input into the relationship model to obtain the equipment health index of each thermal power unit.

6. The method according to claim 1, characterized in that Generating the frequency regulation capability coefficient of each thermal power unit based on the equipment health index and the historical operation data includes: Obtaining a preset mapping relationship between the equipment health index, the historical operating data, and a frequency regulation capability coefficient, and establishing a frequency regulation capability model of the thermal power unit based on the mapping relationship; The equipment health index and the historical operation data are input into the frequency regulation capability model of the thermal power unit to generate a frequency regulation capability coefficient of each thermal power unit.

7. The method according to claim 1, characterized in that The determining of the weight distribution coefficient of each thermal power unit based on the frequency regulation capability coefficient, the inertia requirement, and the equipment health index includes: determining a first weight coefficient of each of the thermal power generating units according to the frequency regulation capability coefficient; determining a second weight coefficient for each of the thermal power units according to the inertia requirement and the inertia that can be provided by each of the thermal power units; determining a third weight coefficient for each of the thermal power units according to the equipment health index; The first weight coefficient, the second weight coefficient and the third weight coefficient are processed based on a preset weight priority rule to determine a weight allocation coefficient for each of the thermal power generation groups.

8. The method according to claim 7, characterized in that After processing the first weight coefficient, the second weight coefficient, and the third weight coefficient based on a preset weight priority rule to determine the weight distribution coefficient of each thermal power unit, the method further includes: Monitor and analyze the real-time operating data of the distributed systems in each thermal power unit in real time to obtain the frequency fluctuation trend; Predict the state of the power grid through a pre-trained prediction model to obtain frequency fluctuations and load changes; Based on the frequency fluctuation trend and the frequency fluctuation and load change results, the weight distribution coefficient of each thermal power unit is adjusted.

9. A thermal power unit weight distribution device based on grid frequency adjustment, characterized in that: include: at least one processor; as well as, a memory communicatively connected to the at least one processor; wherein, The memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to: Collect historical frequency change rate data of the power grid and forecast data of new energy power generation; Collecting historical operating data of the distributed system in each thermal power unit, the operating data including pressure deviation and load rate; Performing frequency division processing on the historical frequency change rate data through a filter to obtain a frequency division signal; determining an inertia requirement of the power grid based on the divided frequency signal; Obtaining valve wear of each of the thermal power units, and determining an equipment health index of each of the thermal power units based on the valve wear; generating a frequency regulation capability coefficient of each of the thermal power units based on the equipment health index and the historical operation data; Based on the frequency regulation capability coefficient, the inertia requirement and the equipment health index, a weight distribution coefficient of each of the thermal power units is determined.

10. A non-volatile computer storage medium, characterized in that The computer-executable instructions are stored, and when the computer-executable instructions are executed by a computer, they can achieve: Collect historical frequency change rate data of the power grid and forecast data of new energy power generation; Collecting historical operating data of the distributed system in each thermal power unit, the operating data including pressure deviation and load rate; Performing frequency division processing on the historical frequency change rate data through a filter to obtain a frequency division signal; determining an inertia requirement of the power grid based on the divided frequency signal; Obtaining valve wear of each of the thermal power units, and determining an equipment health index of each of the thermal power units based on the valve wear; generating a frequency regulation capability coefficient of each of the thermal power units based on the equipment health index and the historical operation data; Based on the frequency regulation capability coefficient, the inertia requirement and the equipment health index, a weight distribution coefficient of each of the thermal power units is determined.

Images