A distributed photovoltaic power generation operation benefit prediction method and system

Abstract

This invention discloses a method and system for predicting the operational benefits of distributed photovoltaic (PV) power generation. The method includes the following steps: acquiring initial power information including electricity market transaction information, grid operating parameters at the grid connection point, operating information of other energy facilities in the region, local meteorological information, and facility load information; using a battery energy storage system to perform a small power disturbance on the grid and measuring the grid's response information; comprehensively analyzing the initial power information and response information to generate a grid operation limitation risk index including grid congestion probability, voltage over-limit risk, and frequency deviation tolerance; and adjusting the operational benefit prediction results and operating strategies of energy assets based on this risk index. This invention combines indirect inference with direct measurement, enabling proactive perception of grid operation limitations and potential curtailment risks in environments with incomplete external grid information, thus avoiding the problem of operational benefit prediction systems consistently overestimating actual returns.

Description

Technical Field

[0001] This invention relates to the field of distributed energy management technology, and more specifically, to a method and system for predicting the operational benefits of distributed photovoltaic power generation. Background Technology

[0002] In modern energy management, accurately forecasting the operational efficiency of large industrial facilities that rely on distributed photovoltaic (PV) power generation systems and integrate battery energy storage systems is a complex task. These facilities not only need to meet their own electricity demands but also actively participate in the electricity market and provide ancillary services to the grid. Facilities typically continuously collect real-time power output data from PV arrays, monitor internal power consumption patterns, obtain next-day electricity price forecasts from power companies via application programming interfaces (APIs), and integrate local weather forecasts to estimate future PV power generation. Based on this input information, the forecasting system can generate a forecast of the facility's net electricity demand for a predetermined future period and a projection of potential revenue from selling excess electricity to the grid, providing energy managers with a basis for decision-making in developing production plans and optimizing grid interaction strategies.

[0003] As facilities become increasingly sophisticated in their energy management, they typically incorporate large-scale battery storage systems. These systems consist of multiple battery banks, a power conversion system, and a battery management system, integrated with the photovoltaic array and connected to the facility's internal power grid. The design functions of battery storage systems include storing excess photovoltaic power, providing backup power, and performing energy arbitrage operations. With the introduction of battery storage systems, forecasting systems need to adjust charging and discharging strategies in real time, based on forecasted photovoltaic power generation and load, predicted electricity prices, facility power consumption curves, and the operating characteristics of the battery storage system, to maximize overall cost savings and benefits. The complexity of forecasting increases significantly because it must consider the operational constraints of the battery storage system and its impact on the facility's interaction with the grid.

[0004] However, as photovoltaic (PV) and battery storage system infrastructure ages, various sensors within the PV arrays and storage systems begin to exhibit subtle but persistent drift, leading to increasingly inaccurate predictive input data. Simultaneously, the complex communication architecture between PV systems, battery storage systems, local load management, and electricity market platforms introduces variable transmission delays and occasional data synchronization issues, significantly diminishing the effectiveness of short-term scheduling optimization.

[0005] More fundamentally, real-time grid operational constraints and curtailment events significantly impact system performance. Local distribution networks are under increasing pressure due to the proliferation of distributed generation and aging infrastructure, leading grid operators to issue sudden curtailment orders during periods of high photovoltaic output or low load. Due to a lack of real-time, detailed information on these grid operational constraints and the likelihood of curtailment, existing operational benefit forecasting systems consistently overestimate the achievable operational benefits of facilities. Therefore, there is an urgent need for an operational benefit forecasting method that can effectively perceive and predict local grid operational pressures and potential curtailment risks to ensure more reliable and accurate financial forecasts for facility energy assets. Summary of the Invention

[0006] This invention provides a method and system for predicting the operational benefits of distributed photovoltaic power generation, which at least solves the problem that existing operational benefit prediction systems, due to their lack of perception of real-time grid operation limitations and potential power curtailment risks, consistently overestimate actual operational benefits.

[0007] In a first aspect, the present invention provides a method for predicting the operational benefits of distributed photovoltaic power generation, comprising the following steps: Acquire initial power information, which includes power market transaction information, grid operation parameters at the grid connection point, operation information of other energy facilities in the region, local meteorological information, and facility load information; A small power disturbance is applied to the power grid using a battery energy storage system, and the power grid's response information is measured. The initial power information and the response information are analyzed comprehensively to generate power grid operation limitation risk indicators; Adjust the operational benefit forecast results of energy assets based on the aforementioned power grid operation constraint risk indicators; Adjust the operating strategy of energy assets based on the aforementioned grid operation limitation risk indicators.

[0008] Optionally, the step of using a battery energy storage system to perform a small power disturbance on the power grid and measuring the power grid's response information includes: Continuously collect grid voltage and current waveform data at the grid connection point; Spectral analysis is performed on the power grid voltage and current waveform data to identify and quantify low-amplitude harmonic or interharmonic components within a specific frequency range, and to extract the amplitude, frequency, and phase characteristics of the background signal. Based on the amplitude, frequency, and phase characteristics of the background signal, a composite power disturbance signal is generated. The composite power disturbance signal includes a preset micro-power probe test signal and a cancellation signal that is opposite in phase and matches the amplitude of the background signal. The composite power disturbance signal is injected into the power grid, and the voltage and current waveform data of the power grid at the grid connection point are collected. The voltage and frequency responses caused by the micro-power probe test signal are extracted, and the instantaneous carrying capacity of the power grid is evaluated.

[0009] Optionally, the step of performing spectrum analysis on the grid voltage and current waveform data at the grid connection point includes: The voltage and current waveform data of the power grid are processed using short-time Fourier transform or wavelet transform to obtain a time-frequency distribution map; Identify the frequency components and amplitudes of background harmonics from the time-frequency distribution diagram; The phase characteristics of the background harmonics are quantified to establish a background signal feature database.

[0010] Optionally, the step of comprehensively analyzing the initial power information and the response information to generate power grid operation limitation risk indicators includes: The electricity market transaction information, power grid operation parameters, other energy facility operation information, local meteorological information, and facility load information in the initial power information are input into a pre-established comprehensive analysis model; The response information is input into the comprehensive analysis model; The comprehensive analysis model outputs power grid operation limitation risk indicators, which include power grid congestion probability, voltage over-limit risk, and frequency deviation tolerance.

[0011] Optionally, the comprehensive analysis model is established in the following way: Collect historical power information samples and historical response information samples; Feature extraction and normalization are performed on the historical power information samples and historical response information samples; Machine learning algorithms are used to train the processed samples to establish a mapping relationship between input features and power grid operation limitation risk indicators.

[0012] Optionally, the step of adjusting the operational benefit forecast results of energy assets based on the power grid operation constraint risk indicators includes: Obtain initial operational benefit forecasts for energy assets; The initial operating benefit prediction result is corrected based on the power grid operation limitation risk index. Specifically, when the power grid operation limitation risk index shows that the probability of power grid congestion is higher than a preset threshold, the expected benefit prediction value is reduced. Output the revised operational benefit forecast results.

[0013] Optionally, the step of adjusting the operating strategy of energy assets based on the grid operation constraint risk index includes: Acquire the current operating strategy of energy assets; Based on the aforementioned grid operation limitation risk indicators, adjust the power output plan and energy storage system charging and discharging strategy in the current operation strategy; When the grid operation limitation risk indicator shows that the voltage over-limit risk is higher than the preset threshold, the charging and discharging strategy of the energy storage system is adjusted first to support the grid voltage.

[0014] Optionally, the method further includes: Real-time monitoring of changes in power grid operation limitation risk indicators; When the aforementioned power grid operation constraint risk indicators change significantly, a reassessment of the operational benefit forecast results and operational strategies is triggered.

[0015] Optionally, the step of obtaining initial power information further includes: The acquired initial power information is preprocessed, including timestamp alignment, data format conversion, and data integrity checks.

[0016] Secondly, the present invention provides a distributed photovoltaic power generation operation benefit prediction system, the system comprising: The information acquisition module is used to acquire initial power information, which includes power market transaction information, power grid operation parameters at the grid connection point, operation information of other energy facilities in the region, local meteorological information, and facility load information. The disturbance measurement module is used to perform small-power disturbances on the power grid using a battery energy storage system and measure the power grid's response information. The risk analysis module is used to comprehensively analyze the initial power information and the response information to generate power grid operation limitation risk indicators; The forecasting and adjustment module is used to adjust the forecast results of the operating benefits of energy assets based on the power grid operation limitation risk indicators. The strategy adjustment module is used to adjust the operation strategy of energy assets based on the power grid operation limitation risk indicators.

[0017] Compared with the prior art, the present invention has the following beneficial technical effects: By actively collecting and analyzing multi-source initial power information, including electricity market transaction information, grid operating parameters, operating information of other energy facilities in the region, local meteorological information, and facility load information, and comprehensively utilizing battery energy storage systems to perform small power disturbances and measure grid response information to directly obtain the instantaneous physical characteristics of grid connection points, this approach combines indirect inference with direct measurement to generate grid operation limitation risk indicators. These risk indicators are then integrated into operational benefit predictions and operational strategy adjustments. This enables proactive perception and quantification of grid operation limitations and potential power curtailment risks in complex environments with incomplete external grid information and high grid operation uncertainty, avoiding the problem of continuously overestimating operational benefits. By introducing spectrum analysis to identify background harmonic characteristics and generate composite power disturbance signals containing cancellation signals, the effectiveness and accuracy of probe testing are improved. By establishing a comprehensive analysis model that integrates multi-source information, a multi-dimensional comprehensive assessment of grid congestion probability, voltage over-limit risk, and frequency deviation tolerance is achieved.

[0018] Details of one or more embodiments of the present invention are set forth in the following drawings and description, so that other features, objects and advantages of the invention will be more readily understood. Attached Figure Description

[0019] Figure 1 This is a flowchart of the distributed photovoltaic power generation operation benefit prediction method in Example 1.

[0020] Figure 2 This is a flowchart of step S2 in Example 1, in which a small power disturbance is performed using a battery energy storage system and the response information is measured.

[0021] Figure 3 This is a flowchart of the spectrum analysis in step S22 of Example 1.

[0022] Figure 4 This is a flowchart of step S3 in Example 1, which involves the comprehensive analysis and generation of power grid operation limitation risk indicators.

[0023] Figure 5 This is a flowchart of adjusting the operational benefit forecast results in step S4 of Example 1.

[0024] Figure 6 This is a flowchart of the comprehensive analysis model establishment in step S41 of Example 1.

[0025] Figure 7 This is a block diagram of the distributed photovoltaic power generation operation benefit prediction system in Example 2. Detailed Implementation

[0026] To make the objectives, technical solutions, and advantages of this application clearer, the application is described and illustrated below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application. All other embodiments obtained by those skilled in the art based on the embodiments provided in this application without inventive effort are within the scope of protection of this application.

[0027] It should be noted that the term "comprising" and any variations thereof in the specification and claims of this invention are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or device that comprises a series of steps or modules is not necessarily limited to those steps or modules that are explicitly listed, but may include other steps or modules that are not explicitly listed or that are inherent to such processes, methods, products, or devices.

[0028] Example 1 Embodiment 1 of the present invention provides a method for predicting the operational benefits of distributed photovoltaic power generation. Figure 1 This is a flowchart of the distributed photovoltaic power generation operation benefit prediction method in Example 1. Figure 1 As shown, the method includes the following steps: S1, Obtain initial power information, wherein the initial power information includes power market transaction information, power grid operation parameters at the grid connection point, operation information of other energy facilities in the region, local meteorological information, and facility load information; In this step, multiple parallel data acquisition channels are used to acquire multi-source initial power information required by distributed photovoltaic power generation facilities in real time. Regarding electricity market transaction information, transaction data from the regional wholesale electricity market, including real-time electricity prices, day-ahead prices, and clearing prices in the ancillary services market, is captured in real time through public application programming interfaces (APIs) provided by electricity market operators. When an unexpected drop in electricity prices or a sustained low price is detected within a short period, it is marked as an indirect signal of increased grid absorption pressure. Regarding grid operating parameters, high-precision power quality monitoring equipment, such as power quality analyzers conforming to international standards, is installed at the facility's grid connection point to monitor and record instantaneous values ​​of grid frequency and voltage in real time with a high-frequency sampling rate. When the grid frequency is consistently higher than the nominal value or local voltage shows abnormal increases, it is identified as a direct signal of increased grid operating pressure. Regarding the operating information of other energy facilities in the region, operating data trends of other large-scale distributed energy sources or large-scale power-consuming facilities within the same distribution area are obtained through public channels. Regarding local meteorological information, local meteorological forecast data, including total solar radiation, ambient temperature, and wind speed information, is integrated. Regarding facility load information, real-time electricity load curves and the operating status of each electrical device are obtained through the facility's internal energy management system.

[0029] S2, using a battery energy storage system to perform a small power disturbance on the power grid and measuring the power grid's response information; In this step, a probe testing module is integrated into the facility's energy management system. This module sends commands to the battery energy storage system during preset periods of relatively stable grid load and minimal impact on the overall system, such as nighttime load low periods or periods of stable photovoltaic output, instructing it to perform brief charge and discharge operations at extremely low power. In a typical test, a 10 MW battery energy storage system might be discharged at 0.5% to 1% of its rated power for approximately 5-10 seconds. During the test, high-precision power quality monitoring equipment installed at the grid connection point accurately measures and records the changes in voltage and frequency before and after the micro-disturbance. By calculating voltage sensitivity and frequency sensitivity, the instantaneous impedance characteristics and carrying capacity of the grid are evaluated. The voltage sensitivity Sv is calculated as follows: The formula for calculating frequency sensitivity Sf is: In the formula, This represents the change in grid connection point voltage before and after the micro-disturbance. The change in frequency This refers to the amount of power change injected or absorbed. If these ratios increase abnormally, for example, the ratio of voltage change to power change exceeds a preset threshold, it indicates that the power grid is sensitive to power changes and its carrying capacity may be weak. Through this proactive, real-time testing, the instantaneous physical characteristics of the power grid connection points can be directly obtained, providing the most direct and reliable feedback for assessing the health of the power grid.

[0030] S3, comprehensively analyze the initial power information and the response information to generate power grid operation limitation risk indicators; In this step, the various public power grid signal characteristics collected in step S1 are comprehensively analyzed together with the power grid response data obtained through probe testing in step S2, combined with local weather forecasts, the facilities themselves, and inferable regional load change trends. During the comprehensive analysis, a scenario matching mechanism is used to identify potential power grid stress scenarios in the current and short-term future. For example, if multiple conditions are met simultaneously—weather forecasts indicating high irradiance and high temperatures suggesting high photovoltaic output, a continuous decline in regional wholesale electricity prices or a persistently high grid frequency, probe testing showing abnormal sensitivity of the grid connection point voltage to small power injections, and inference from the behavior patterns of neighboring facilities indicating a general decline in the output of other photovoltaic power plants in the region—it is determined that there is a risk of overlapping high power generation and low grid absorption capacity. Based on the comprehensive analysis results of these multi-source information, a quantitative power grid operation constraint risk index is generated. This index comprehensively reflects multiple dimensions such as grid congestion probability, voltage over-limit risk, and frequency deviation tolerance, providing a basis for subsequent benefit prediction adjustments and operational strategy optimization.

[0031] S4. Adjust the operational benefit forecast results of energy assets based on the aforementioned power grid operation limitation risk indicators; In this step, the initial operational benefit forecast of the energy asset under normal conditions is obtained. This forecast is calculated based on photovoltaic power generation, load, electricity price, and the planned charging and discharging strategy of the energy storage system. Then, the grid operation constraint risk index generated in step S3 is used as a correction factor to adjust the initial operational benefit forecast. Specifically, when the grid congestion probability indicated by the grid operation constraint risk index is higher than a preset risk tolerance threshold, the expected revenue forecast is reduced by a correction coefficient proportional to the congestion probability. This correction coefficient is determined by the relationship between the probability of power curtailment events and actual revenue loss in historical data. When the voltage over-limit risk is higher than a preset threshold, the portion of expected revenue from electricity sales to the grid is adjusted according to the discount rate corresponding to the over-limit risk. The adjusted operational benefit forecast is closer to the actual achievable revenue level, avoiding continuous overestimation due to ignoring grid constraints.

[0032] S5, adjust the operation strategy of energy assets according to the power grid operation limitation risk indicators; In this step, based on the grid operation constraint risk index generated in step S3, the operation strategy of energy assets is proactively adjusted to avoid risks before actual power curtailment or grid constraints occur. When the grid congestion probability index shows a high level, the battery energy storage system is preferentially switched to charging mode to store as much electrical energy as possible during periods of high photovoltaic output rather than transmitting it to the grid, thereby reducing the risk of curtailment. When the risk of voltage exceeding limits is high, the reactive power output of the photovoltaic inverter is adjusted or the voltage support function of the energy storage system is activated to proactively participate in the voltage regulation of the local grid, transforming it from a potential source of power curtailment risk into a grid supporter. For dispatch strategies involving ancillary services such as fast frequency regulation, the reserved capacity of battery energy storage systems that can participate in ancillary services is dynamically adjusted according to the frequency deviation tolerance index to ensure sufficient responsiveness during grid frequency fluctuations. Through this proactive strategy adjustment, economic benefits are maximized while actively adapting to grid operation constraints.

[0033] In the technical solution of the above embodiments, multi-source initial power information is acquired, including electricity market transaction information, grid operation parameters at the grid connection point, operation information of other energy facilities in the region, local meteorological information, and facility load information. A battery energy storage system is used to perform small-power disturbances on the grid and measure the grid's response information to directly obtain the instantaneous physical characteristics of the grid connection point. The initial power information and response information are comprehensively analyzed to generate quantified grid operation constraint risk indicators. Based on these risk indicators, the operational benefit forecasts and operational strategies for energy assets are adjusted. This method transforms facilities from passive receivers of grid commands to proactive observers of grid status. In complex environments where external grid information is incomplete and grid operation is highly uncertain, it achieves forward-looking perception and quantification of grid operation constraints and potential power curtailment risks, avoiding the problem of continuously overestimating operational benefits and ensuring the reliability and accuracy of financial forecasts for facility energy assets.

[0034] In some preferred embodiments, assume a large industrial facility has installed a 5 MW distributed photovoltaic (PV) power generation system on its roof, equipped with a 10 MWh battery energy storage system. During midday on a high-irradiance day, PV output reaches over 80% of its rated power, while the facility's own electricity load is at a low level. The facility's existing operational benefit forecasting system, based on historical electricity price data and PV output forecasts, anticipates high revenue from selling electricity to the grid during this period. However, the grid operator is about to issue a power curtailment order due to the risk of local grid overload caused by multiple PV power plants operating at high output simultaneously in the area. Step S1 collects real-time data showing an unexpected drop in electricity market transaction prices within a short period, the grid connection voltage consistently approaching the upper limit of the nominal voltage, and a collective downward trend in the declared output of other PV power plants in the area. Step S2 indicates that the grid connection voltage exhibits abnormal sensitivity to power injection during earlier probe tests that day. Step S3 combines this information to determine that the current grid congestion probability is high, generating a corresponding grid operation restriction risk indicator. Step S4 accordingly adjusts the expected electricity sales revenue forecast downwards. Step S5 proactively adjusts the operating strategy, switching the energy storage system to charging mode to absorb excess power generation, thus avoiding the loss of some power generation due to power curtailment and reducing the overestimation of expected returns caused by power curtailment.

[0035] In this embodiment, Figure 2 This is a flowchart illustrating step S2 of Example 1, where a small power disturbance is performed using a battery energy storage system, and the response information is measured. (Refer to...) Figure 2 In step S2, the step of using a battery energy storage system to perform a small power disturbance on the power grid and measuring the response information of the power grid includes: S21, continuously collects grid voltage and current waveform data at the grid connection point; In this step, high-precision power quality monitoring equipment, such as a multi-channel power quality analyzer conforming to relevant International Electrotechnical Commission (IEC) standards, is installed at the facility's grid connection point. This equipment can simultaneously acquire instantaneous waveform data of three-phase voltage and three-phase current, with a sampling frequency no lower than a preset kilohertz level to ensure the capture of high-frequency disturbances and transient events in the power grid. The acquired raw waveform data is transmitted in real time to the central data processing unit via the Industrial Ethernet protocol. During transmission, a high-precision GPS or Network Time Protocol (NAT) timestamp is added to each sampling point to ensure time synchronization accuracy during subsequent spectrum analysis and multi-source data fusion.

[0036] S22, Perform spectrum analysis on the power grid voltage and current waveform data, identify and quantify low-amplitude harmonic or interharmonic components within a specific frequency range, and extract the amplitude, frequency and phase characteristics of the background signal; In this step, the voltage and current waveform data acquired in step S21 are subjected to spectral analysis. First, the raw waveform data undergoes DC component removal and windowing preprocessing using a Hanning or Hamming window to reduce spectral leakage. Then, the preprocessed data is processed using a Fast Fourier Transform or Wavelet Transform to convert the time-domain waveform to the time-frequency domain, obtaining a time-frequency distribution map reflecting the energy distribution of the signal at different times and frequencies. Harmonic and interharmonic components present in the power grid background are identified from the time-frequency distribution map, with a focus on low-amplitude harmonics near integer multiples of the power frequency and interharmonic components at non-integer multiples of the power frequency. For each identified harmonic or interharmonic component, its frequency, amplitude, and phase angle relative to the reference signal are quantized, and the extracted amplitude, frequency, and phase features are stored in the background signal feature database.

[0037] S23, Based on the amplitude, frequency and phase characteristics of the background signal, a composite power disturbance signal is generated. The composite power disturbance signal includes a preset micro power probe test signal and a cancellation signal with opposite phase and matched amplitude to the background signal. In this step, based on the background signal features extracted in step S22, a composite power disturbance signal is generated using digital signal synthesis technology. This composite signal consists of two superimposed components: the first component is a preset micro-power probe test signal with preset power amplitude, frequency, and duration. The power amplitude is typically set to 0.5% to 1% of the rated power of the battery energy storage system, the frequency is selected in a band that does not overlap with the background harmonic frequency to avoid interference, and the duration is approximately 5-10 seconds. The second component is a cancellation signal with opposite phase and matched amplitude to the background signal. This signal is generated in real time based on the amplitude and frequency information of the background harmonics extracted in step S22, and its phase is set to be 180° out of phase with the background harmonics. By superimposing these two components, the final composite power disturbance signal is generated, which is then amplified and output by the power conversion system of the battery energy storage system.

[0038] S24, the composite power disturbance signal is injected into the power grid, the voltage and current waveform data of the power grid at the grid connection point are collected, the voltage and frequency response caused by the micro power probe test signal is extracted, and the instantaneous carrying capacity of the power grid is evaluated. In this step, the composite power disturbance signal generated in step S23 is injected into the grid at the grid connection point in the form of a voltage source or a current source through the power conversion system of the battery energy storage system. During the injection process, voltage and current waveform data at the grid connection point are continuously collected at the same sampling rate and accuracy as in step S21. After the injection is completed, the collected response data is analyzed and processed: firstly, the influence of background harmonics is canceled out using the cancellation signal component, so that the voltage change and frequency change extracted from the response data mainly come from the contribution of the micro-power probe test signal. Specifically, in this step, the voltage sensitivity Sv and frequency sensitivity Sf of the grid in step S2 are calculated according to the following formula: Where ΔV represents the change in grid connection voltage before and after the probe test, Δf represents the change in grid frequency before and after the probe test, and ΔP represents the power amplitude of the probe test signal. Sv and Sf reflect the grid's voltage sensitivity and frequency sensitivity to external power injection, respectively. If the ratio of voltage change to power change exceeds a preset voltage sensitivity threshold, it indicates that the grid's carrying capacity at that connection point is weak and its response to external power injection is sensitive. This may mean that the grid is operating under high load or near its stability limit.

[0039] In the technical solution of the above embodiments, voltage and current waveform data of the power grid at the grid connection point are continuously collected. Spectral analysis is performed on the waveform data to identify and quantify the amplitude, frequency, and phase characteristics of background harmonics and interharmonics. Based on the background signal characteristics, a composite power disturbance signal containing probe test signals and background cancellation signals is generated. This composite signal is injected into the power grid, and response data is collected. After canceling background interference, the voltage and frequency responses caused by the probe test signals are extracted to evaluate the instantaneous carrying capacity of the power grid. This method utilizes cancellation signal technology to eliminate the interference of power grid background harmonics on the probe test results, enabling accurate extraction of power grid response characteristics caused by small power disturbances from a complex power grid environment, thus improving the accuracy and reliability of power grid carrying capacity assessment.

[0040] In some preferred embodiments, it is assumed that the battery energy storage system of the facility has a rated power of 10 MW, and there are strong 3rd and 5th background harmonics at the grid connection point, with harmonic amplitudes reaching 2% and 1% of the fundamental frequency amplitude, respectively. If a 50 kW probe test signal is directly injected without background cancellation technology, the voltage response change caused by the probe test signal may be overwhelmed by the natural fluctuations of the background harmonics, making it impossible to accurately assess the instantaneous carrying capacity of the grid. Step S22 identifies the frequency, amplitude, and phase characteristics of the 3rd and 5th harmonics through Fast Fourier Transform. Step S23, while generating the probe test signal, synthesizes a cancellation signal with opposite phase and matched amplitude to the 3rd and 5th harmonics, and superimposes the two to generate a composite power disturbance signal. Step S24, after injecting the composite signal into the grid through the battery energy storage system, the cancellation signal effectively suppresses the contribution of the background harmonics at the grid connection point, so that the extracted voltage and frequency responses only contain the contribution of the probe test signal. The analysis results show that the ratio of voltage change to power change is close to the preset warning threshold, indicating that the grid's carrying capacity is relatively weak during this period, and the injection of photovoltaic power into the grid should be carefully controlled.

[0041] In this embodiment, Figure 3 This is a flowchart of the spectrum analysis in step S22 of Example 1. (Refer to...) Figure 3 In step S22, the step of performing spectrum analysis on the grid voltage and current waveform data includes: S221, Use short-time Fourier transform or wavelet transform to process the power grid voltage and current waveform data to obtain a time-frequency distribution map; In this step, time-frequency analysis is performed on the preprocessed voltage and current waveform data, using either short-time Fourier transform (SFT) or wavelet transform. When using SFT, a suitable window function and window length are selected, balancing frequency and time domain resolution. A Hamming window is used to reduce sidelobe interference. The discrete Fourier transform is calculated segment by segment using a sliding window to obtain the signal's energy distribution at different times and frequencies. When using wavelet transform, a Morlet wavelet or complex Morlet wavelet is chosen as the mother wavelet function. Multi-scale decomposition yields the signal's frequency components at different time scales. Wavelet transform is advantageous for handling non-stationary signals and intermittent events. Regardless of the transform method used, the final output is a 2D time-frequency distribution plot, with the horizontal axis representing time and the vertical axis representing frequency. Color or grayscale depth indicates the signal energy intensity at that time and frequency. The time-frequency distribution plot is stored as a matrix data format for subsequent feature extraction steps.

[0042] S222, Identify the frequency components and amplitudes of background harmonics from the time-frequency distribution diagram; In this step, background harmonics are identified on the time-frequency distribution map generated in step S221. First, the energy of each frequency component is averaged over time or the median is taken along the time axis in the frequency dimension to obtain a frequency-amplitude spectrum reflecting the steady-state background of the power grid. Then, local peaks near integer multiples of the power frequency are searched on this frequency-amplitude spectrum; these peaks correspond to integer-order background harmonics in the power grid. Simultaneously, local peaks at non-integer multiples of the power frequency are searched; these peaks correspond to interharmonic components. For each identified frequency component, its amplitude sequence over time is extracted from the time-frequency distribution map, and the mean of this amplitude sequence is calculated as the amplitude value of that frequency component, expressed as a percentage relative to the fundamental frequency amplitude. For each identified background harmonic component, its frequency and amplitude values ​​are recorded.

[0043] S223, quantify the phase characteristics of the background harmonics and establish a background signal feature database; In this step, phase feature quantization is performed on each background harmonic and interharmonic component identified in step S222. For each harmonic frequency component, the complex value of each time sampling point at that frequency is obtained on the time-frequency distribution map. The argument of this complex value is the instantaneous phase angle of that frequency component at that moment. Statistical analysis is performed on the instantaneous phase angle sequence within a certain time window, and its mean is calculated as the characteristic phase angle of that frequency component. Its standard deviation is calculated as a measure of phase stability. The frequency value, amplitude value, and characteristic phase angle of each identified harmonic or interharmonic component are recorded as a single record, along with information such as the acquisition time and acquisition location, and stored in the background signal feature database. This database is implemented using a time series database engine, supporting queries by time range and filtering by frequency range, providing data support for subsequent cancellation signal generation and historical trend analysis.

[0044] In the technical solution of the above embodiments, a time-frequency distribution map is obtained by performing time-frequency analysis on the grid voltage and current waveform data using short-time Fourier transform or wavelet transform. The frequency components and amplitudes of background harmonics are identified from the time-frequency distribution map, and the phase characteristics of the harmonics are quantified to establish a background signal feature database. This method provides accurate background harmonic characteristic parameters for subsequent cancellation signal generation, enabling the cancellation component in the composite power disturbance signal to accurately cancel the phase with the background harmonics, thus minimizing the impact of background interference on the probe testing accuracy.

[0045] In this embodiment, Figure 4 This is a flowchart illustrating the comprehensive analysis and generation of power grid operation limitation risk indicators in step S3 of Example 1. (Refer to...) Figure 4 In step S3, the step of comprehensively analyzing the initial power information and the response information to generate power grid operation limitation risk indicators includes: S31, input the electricity market transaction information, power grid operation parameters, other energy facility operation information, local meteorological information and facility load information from the initial power information into the pre-established comprehensive analysis model; In this step, the various initial power information collected through multi-source data channels in step S1 is formatted and then input into the comprehensive analysis model. Power market transaction information is input in time series format, including the trend of real-time and day-ahead electricity prices over a past period, price volatility, and clearing prices in the ancillary services market. Grid operation parameters are input in statistical characteristic format, including the mean, maximum, and minimum voltage at the grid connection point and the percentage deviation from the nominal value, the mean grid frequency, and the duration and magnitude of frequency offset. Other energy facility operation information is input in normalized trend indicators, including the relative change rate of output from other photovoltaic power plants in the region and the electricity consumption trend of large industrial loads. Local meteorological information is input in physical quantity format, including forecast and measured values ​​of solar radiation intensity, ambient temperature, wind speed, and cloud cover. Facility load information is input in power curve format, including the current load level, load change trend, and adjustable load capacity. These multi-source heterogeneous data are uniformly timestamped and interpolated for missing values ​​before being input into the model, and then normalized or standardized according to the format required by the model.

[0046] S32, Input the response information into the comprehensive analysis model; In this step, the power grid response information obtained through probe testing in step S2 is input into the comprehensive analysis model. The response information is input in the form of key indicators, including the ratio of voltage change to disturbance power change, the ratio of frequency change to disturbance power change, and the trends of these ratios relative to historical baseline values. When the most recent probe test has been conducted for more than a preset period, the model will lower the confidence weight of the response information and rely more on other information sources; when the probe test was recently conducted and the response indicators are stable, the model assigns a higher confidence weight to the response information, because direct physical measurements are generally more valuable than indirect inferences. The response information input also includes metadata such as the probe test execution time, test duration, and test power amplitude, for the model to reference when performing time-series analysis.

[0047] S33, the comprehensive analysis model outputs power grid operation limitation risk indicators, wherein the power grid operation limitation risk indicators include power grid congestion probability, voltage over-limit risk and frequency deviation tolerance; In this step, the comprehensive analysis model fuses and performs inference analysis on the input multi-source information, outputting multi-dimensional power grid operation limitation risk indicators. Power grid congestion probability. Calculate using the following formula: in, This is a congestion probability component assessed based on abnormal signals from electricity market trading prices. This is a congestion probability component assessed based on grid connection point voltage sensitivity and current voltage level. This is a congestion probability component inferred based on operational information of other energy facilities within the region. This is a baseline congestion probability based on the frequency of congestion events during the same historical period. Let be the weighting coefficients of each component, satisfying... Voltage exceedance risk indicators are expressed as risk levels or probabilities, reflecting the likelihood that the voltage at the grid connection point will exceed the normal operating range specified by the grid operator. This indicator is calculated based on a comprehensive assessment of the current voltage level, voltage sensitivity, and predicted photovoltaic output trends. Frequency deviation tolerance. Calculate using the following formula: in, This represents the upper limit of frequency sensitivity safety, determined based on historical frequency fluctuation data. The three indicators are output in a structured data format, along with their respective calculated confidence scores, main contributing factors, and suggested decision-making time windows.

[0048] In the technical solution of the above embodiments, multi-source initial power information is input into a comprehensive analysis model in a structured format, and grid response information obtained from probe tests is input into the model in the form of key indicators. After the model fuses and processes various types of information, it outputs a multi-dimensional grid operation limitation risk indicator that includes grid congestion probability, voltage over-limit risk, and frequency deviation tolerance. This method achieves an organic combination of indirect inference and direct measurement, enabling the assessment of grid operation risk to not only cover market signals and regional behavior patterns at the macroeconomic level, but also accurately reflect the instantaneous physical state of the facility's local grid connection point, resulting in a more comprehensive and reliable assessment result.

[0049] In this embodiment, Figure 5 This is a flowchart of adjusting the operational benefit forecast results in step S4 of Example 1. (Refer to...) Figure 5 In step S4, the step of adjusting the operational benefit forecast results of energy assets based on the power grid operation limitation risk index includes: S41, Obtain initial operational benefit forecasts for energy assets; In this step, initial operational benefit projections for the energy assets during the current operating cycle are obtained from the facility's energy management system. These initial projections are based on forecasted photovoltaic power generation. Load forecast Electricity market forecast prices And the planned charge and discharge power of the battery energy storage system at each time step t The initial operating benefit forecast obtained from comprehensive calculations is as follows: Here, Δt represents the time step, typically with a forecast window of 24 to 48 hours, outputting the expected return value for each time step. Input parameters for the initial forecast include weather forecast data, photovoltaic module performance parameters, the state of charge and health of the battery energy storage system, contract electricity price, and forecast market electricity price. This forecast reflects the expected returns that facility energy assets can realize in the electricity market and ancillary services market under the assumption of ideal grid conditions and no operational constraints.

[0050] S42, the initial operating benefit prediction result is corrected according to the power grid operation restriction risk index, wherein when the power grid operation restriction risk index shows that the probability of power grid congestion is higher than a preset threshold, the expected benefit prediction value is reduced; In this step, the grid operation limitation risk index generated in step S3 is used as an external correction factor to adjust the initial operating benefit forecast results for risk. For each forecast time step, the grid operation limitation risk index corresponding to that step is first checked. When the grid congestion probability is higher than the preset risk tolerance threshold, the electricity sales revenue portion of the expected revenue forecast is multiplied by a congestion discount factor less than one. This discount factor is determined by the relationship function between the congestion probability and the actual revenue loss rate in historical power curtailment events. When the voltage over-limit risk is higher than the preset threshold, the ancillary service revenue portion of the expected revenue that depends on grid voltage stability is multiplied by the corresponding voltage risk discount factor. When the frequency deviation tolerance is lower than the preset safety level, the expected revenue from participating in frequency regulation ancillary services is adjusted accordingly. The correction process adopts a refined adjustment strategy based on time periods and categories. Different correction coefficients are applied to different revenue sources according to their sensitivity to various grid operation risks. The corrected operating benefit forecast is shown below. Calculate using the following formula: in, This is the initial projected operating benefit value. This is a congestion discount factor corresponding to the probability of power grid congestion. This is the voltage risk discount factor corresponding to the voltage over-limit risk. This is the frequency regulation capability factor corresponding to the frequency deviation tolerance (its value ranges from 0 to 1; the lower the tolerance, the closer the factor is to 0). When the grid congestion probability is below a preset threshold... The value is 0 when the risk of voltage exceeding the limit is lower than the preset threshold. The value is 0. The revised operating benefit forecast is closer to the actual achievable revenue level, which helps management to make more reliable financial budgets and investment decisions.

[0051] S43, output the revised operational benefit forecast results; In this step, the revised operational benefit forecast results from step S42 are output in a structured data format. The output includes the revised total expected revenue, breakdown forecasts for each revenue source, revenue distribution curves for each time step, the applied correction coefficients and their corresponding risk indicator values, and a comparison of revenue differences before and after the correction. The output data is presented in two forms: visual charts and data reports. Visual charts include a comparison curve of revenue forecasts before and after correction, a heatmap of the time-varying characteristics of grid operation constraint risk indicators, and a mapping diagram between risk indicators and correction coefficients. Data reports include detailed correction explanations for each time period and annotations of key risk points. The output results are displayed through the user interface of the facility's energy management system and can also be exported to a standard data format for subsequent analysis.

[0052] In the technical solution of the above embodiments, by obtaining the initial operational benefit prediction results of energy assets, the initial prediction results are refined and categorized according to the various dimensions of the grid operation constraint risk indicators. When the grid congestion probability is higher than a preset threshold, the predicted value of electricity sales revenue is reduced; when the voltage exceedance risk and frequency deviation tolerance are abnormal, the predicted value of ancillary service revenue is adjusted accordingly; and the revised prediction results, including a comparison before and after the correction and detailed correction explanations, are output. This method systematically integrates the external uncertainties of grid operation into the benefit prediction process in the form of quantified risk indicators, so that the prediction results not only reflect the intrinsic performance of energy assets but also reflect the actual constraints of the external grid environment, providing a more reliable data foundation for the financial management of facilities.

[0053] In some preferred embodiments, it is assumed that the initial operational benefit forecast for the facility on a high-irradiance day indicates that selling electricity to the grid via photovoltaic power generation is expected to generate high revenue between 10:00 AM and 2:00 PM. However, the grid operation limitation risk index generated in step S3 shows that the grid congestion probability during this period is over 70%, the voltage exceedance risk is at a moderate level, and the frequency deviation tolerance is within the normal range. After obtaining the initial forecast result in step S41, step S42 adjusts the electricity sales revenue and some ancillary service revenue according to the congestion probability and voltage exceedance risk: a larger congestion discount factor is applied to the electricity sales revenue due to the high congestion probability, a moderate voltage risk discount factor is applied to the voltage stability-related portion of the ancillary service revenue, while the predicted revenue of the frequency regulation service remains basically unchanged. Step S43 outputs the revised operational benefit forecast result, in which the expected revenue of the electricity sales portion is significantly lower than the initial forecast value, while the expected revenue of the frequency regulation service remains basically unchanged. Based on the revised results, facility energy managers shifted their dispatch strategy focus from electricity sales to participating in frequency regulation services, thus avoiding revenue losses caused by power curtailment in actual operation.

[0054] In this embodiment, Figure 6 This is a flowchart of the comprehensive analysis model establishment in step S41 of Example 1. (Refer to...) Figure 6 In step S41, the comprehensive analysis model is established in the following way: S411, collect historical power information samples and historical response information samples; In this step, historical sample data for model training is collected from the historical database of the facility's energy management system. Historical electricity information samples include electricity market transaction price sequences over a relatively long period, such as more than one year; time series of grid operating parameters at the grid connection point; operating data of other energy facilities in the region; local meteorological measurement data; and facility load records. Historical response information samples include records of the results of each historical probe test, including the ratio of voltage change to power change, the ratio of frequency change to power change, and corresponding test condition metadata. Simultaneously, records of actual grid operation restriction events are collected as labels for the model output, including the time, duration, power curtailment magnitude, and cause description of power curtailment orders issued by the grid operator. The sample data is divided into training and validation sets according to a preset ratio, ensuring that the training set contains samples of various typical grid operating conditions.

[0055] S412, Perform feature extraction and normalization processing on the historical power information samples and historical response information samples; In this step, feature engineering is performed on the collected historical sample data. For electricity market transaction price series, statistical features such as price change rate, price fluctuation standard deviation, and deviation of price from historical mean are extracted. For grid operation parameters, features such as cumulative time of voltage deviation from nominal value, frequency and amplitude of frequency offset events, and quantiles of voltage and frequency are extracted. For other energy facility operation data, regional output change trends and synchronicity indicators are extracted. For meteorological data, solar radiation intensity, temperature gradient, and marker features of extreme weather events are extracted. For load data, time-series load pattern features and correlation features between load and photovoltaic output are extracted. For response information samples, voltage sensitivity and frequency sensitivity indicators are directly extracted from the records. All extracted feature vectors are normalized using min-max normalization or standardization methods to unify features of different dimensions to the same scale, thereby eliminating the impact of feature magnitude differences on the machine learning algorithm training process.

[0056] S413 uses machine learning algorithms to train the processed samples and establishes a mapping relationship between input features and power grid operation limitation risk indicators; In this step, a suitable machine learning algorithm for multi-input multi-output regression prediction tasks, such as gradient boosting trees, random forests, or deep neural networks, is selected to train the model on the feature vectors processed in step S412 and the corresponding power grid operation restriction event labels. The model's input is the normalized multi-source feature vector, and the output consists of three risk indicators: power grid congestion probability, voltage over-limit risk, and frequency deviation tolerance. During training, cross-validation is used to evaluate the model's generalization ability, and the model's hyperparameters are adjusted using grid search or Bayesian optimization to obtain optimal performance. After training, the model's prediction accuracy is evaluated on the validation set, and the root mean square error and coefficient of determination for the three output indicators are calculated. If the model performance meets the preset accuracy requirements, the model parameters are saved and deployed to the comprehensive analysis module for online inference; if the model performance does not meet the requirements, the amount of training sample data is increased or the feature engineering and model structure are adjusted before retraining. After the model is deployed online, its online prediction performance is continuously monitored, and the model is updated periodically or triggered based on newly collected operational data and actual event feedback.

[0057] In the technical solution of the above embodiments, a training dataset is constructed by collecting historical power information samples and historical response information samples. Feature extraction and normalization are performed on the sample data to obtain a unified feature vector. Machine learning algorithms are then used to train the processed samples to establish a mapping relationship between input features and power grid operation limitation risk indicators. Based on long-term accumulated historical operation data and actual power grid limitation event records, this method establishes a comprehensive analytical model that accurately reflects complex nonlinear mapping relationships through a data-driven approach, providing a reliable computing engine for the real-time generation of power grid operation limitation risk indicators.

[0058] In this embodiment, the operation strategy of energy assets is adjusted according to the method of any of the above embodiments. Step S5, which involves adjusting the operation strategy of energy assets based on the grid operation limitation risk indicator, includes: obtaining the current operation strategy of the energy assets; adjusting the power output plan and energy storage system charging / discharging strategy in the current operation strategy according to the grid operation limitation risk indicator; and prioritizing the adjustment of the energy storage system's charging / discharging strategy to support the grid voltage when the grid operation limitation risk indicator shows a voltage over-limit risk higher than a preset threshold. This strategy adjustment proactively adjusts the facility's operation mode when grid operation limitation risks occur, avoiding revenue losses caused by passively accepting power curtailment orders.

[0059] In this embodiment, the method according to any of the above embodiments further includes continuous monitoring and triggered reassessment of power grid operation limitation risk indicators. The changing trends of power grid operation limitation risk indicators are monitored in real time. When an indicator undergoes a significant change exceeding a preset range—for example, a rapid increase in the probability of power grid congestion exceeding a preset percentage point or a shift in voltage over-limit risk from a low-risk to a high-risk state—a reassessment of the operational benefit forecast results and operational strategies is automatically triggered. This continuous monitoring and triggered response mechanism ensures that facility benefit forecasts and operational strategies can promptly reflect the dynamic changes in the power grid's operational status, improving the system's adaptability to uncertainties in the power grid.

[0060] In this embodiment, according to the method of any of the above embodiments, the step of obtaining initial power information further includes preprocessing the obtained initial power information. Preprocessing includes aligning the collected data from different data sources to timestamps according to GPS or Network Time Protocol timestamps, converting different data formats from different data sources into a structured standard format such as JSON or CSV, and performing integrity checks on data fields to identify and mark missing values, outliers, or duplicate records. This preprocessing step ensures the consistency of multi-source data input into the subsequent comprehensive analysis process in terms of time, format, and quality dimensions, providing a high-quality data foundation for the reliable operation of the comprehensive analysis model.

[0061] In summary, the distributed photovoltaic power generation operation benefit prediction method provided in Embodiment 1 of this invention actively collects initial power information from multiple sources and utilizes a battery energy storage system to perform small power disturbances to directly obtain grid response information. It combines indirect inference with direct measurement to comprehensively generate grid operation limitation risk indicators. Based on these risk indicators, the method adjusts the operation benefit prediction results to avoid continuously overestimating returns and proactively adjusts the operation strategy of energy assets to adapt to grid constraints. By identifying background harmonic characteristics through spectrum analysis and generating composite power disturbance signals using phase cancellation technology, the reliability of probe testing in complex grid environments is improved. By establishing a machine learning-based comprehensive analysis model that integrates multi-source information, it achieves multi-dimensional quantitative assessment of grid congestion probability, voltage over-limit risk, and frequency deviation tolerance, providing comprehensive data support for facility energy management decisions.

[0062] Example 2 Embodiment 2 of the present invention provides a distributed photovoltaic power generation operation benefit prediction system. Figure 7 This is a block diagram of the distributed photovoltaic power generation operation benefit prediction system in Example 2. Figure 7 As shown, the system includes: The information acquisition module 01 is used to acquire initial power information, which includes power market transaction information, power grid operation parameters at the grid connection point, operation information of other energy facilities in the region, local meteorological information, and facility load information.

[0063] The disturbance measurement module 02 is used to perform small-power disturbances on the power grid using a battery energy storage system and measure the response information of the power grid.

[0064] Risk analysis module 03 is used to comprehensively analyze the initial power information and the response information to generate power grid operation limitation risk indicators.

[0065] The prediction and adjustment module 04 is used to adjust the prediction results of the operating benefits of energy assets based on the power grid operation limitation risk indicators.

[0066] The strategy adjustment module 05 is used to adjust the operation strategy of energy assets based on the power grid operation limitation risk indicators.

[0067] The distributed photovoltaic power generation operation benefit prediction system described in the above embodiment acquires initial power information from multiple sources, utilizes a battery energy storage system to perform minor power disturbances and measures grid response information, comprehensively analyzes and generates grid operation limitation risk indicators, and adjusts the operation benefit prediction results and operating strategies based on these risk indicators. This system transforms facilities from passive receivers of grid commands to proactive observers of grid conditions. In environments with incomplete external grid information, it achieves forward-looking perception and quantification of grid operation risks and potential power curtailment events, ensuring the reliability and accuracy of financial forecasts for facility energy assets.

[0068] This invention also provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the steps of the method provided in Embodiment 1 above.

[0069] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this patent application should be determined by the appended claims.

Claims

1. A method for predicting the operational benefits of distributed photovoltaic power generation, characterized in that, Includes the following steps: Acquire initial power information, which includes power market transaction information, grid operation parameters at the grid connection point, operation information of other energy facilities in the region, local meteorological information, and facility load information; A small power disturbance is applied to the power grid using a battery energy storage system, and the power grid's response information is measured. The initial power information and the response information are analyzed comprehensively to generate power grid operation limitation risk indicators; Adjust the operational benefit forecast results of energy assets based on the aforementioned power grid operation constraint risk indicators; Adjust the operating strategy of energy assets based on the aforementioned grid operation limitation risk indicators.

2. The method according to claim 1, characterized in that, The step of using a battery energy storage system to perform a small power disturbance on the power grid and measuring the power grid's response information includes: Continuously collect grid voltage and current waveform data at the grid connection point; Spectral analysis is performed on the power grid voltage and current waveform data to identify and quantify low-amplitude harmonic or interharmonic components within a specific frequency range, and to extract the amplitude, frequency, and phase characteristics of the background signal. Based on the amplitude, frequency, and phase characteristics of the background signal, a composite power disturbance signal is generated. The composite power disturbance signal includes a preset micro-power probe test signal and a cancellation signal that is opposite in phase and matches the amplitude of the background signal. The composite power disturbance signal is injected into the power grid, and the voltage and current waveform data of the power grid at the grid connection point are collected. The voltage and frequency responses caused by the micro-power probe test signal are extracted, and the instantaneous carrying capacity of the power grid is evaluated.

3. The method according to claim 2, characterized in that, The steps for performing spectrum analysis on the grid voltage and current waveform data at the grid connection point include: The voltage and current waveform data of the power grid are processed using short-time Fourier transform or wavelet transform to obtain a time-frequency distribution map; Identify the frequency components and amplitudes of background harmonics from the time-frequency distribution diagram; The phase characteristics of the background harmonics are quantified to establish a background signal feature database.

4. The method according to claim 1, characterized in that, The step of comprehensively analyzing the initial power information and the response information to generate power grid operation limitation risk indicators includes: The electricity market transaction information, power grid operation parameters, other energy facility operation information, local meteorological information, and facility load information in the initial power information are input into a pre-established comprehensive analysis model; The response information is input into the comprehensive analysis model; The comprehensive analysis model outputs power grid operation limitation risk indicators, which include power grid congestion probability, voltage over-limit risk, and frequency deviation tolerance.

5. The method according to claim 4, characterized in that, The comprehensive analysis model was established in the following way: Collect historical power information samples and historical response information samples; Feature extraction and normalization are performed on the historical power information samples and historical response information samples; Machine learning algorithms are used to train the processed samples to establish a mapping relationship between input features and power grid operation limitation risk indicators.

6. The method according to claim 1, characterized in that, The step of adjusting the operational benefit forecast results of energy assets based on the power grid operation limitation risk indicators includes: Obtain initial operational benefit forecasts for energy assets; The initial operating benefit prediction result is corrected based on the power grid operation limitation risk index. Specifically, when the power grid operation limitation risk index shows that the probability of power grid congestion is higher than a preset threshold, the expected benefit prediction value is reduced. Output the revised operational benefit forecast results.

7. The method according to claim 1, characterized in that, The step of adjusting the operating strategy of energy assets based on the power grid operation limitation risk indicators includes: Acquire the current operating strategy of energy assets; Based on the aforementioned grid operation limitation risk indicators, adjust the power output plan and energy storage system charging and discharging strategy in the current operation strategy; When the grid operation limitation risk indicator shows that the voltage over-limit risk is higher than the preset threshold, the charging and discharging strategy of the energy storage system is adjusted first to support the grid voltage.

8. The method according to claim 1, characterized in that, Also includes: Real-time monitoring of changes in power grid operation limitation risk indicators; When the aforementioned power grid operation constraint risk indicators change significantly, a reassessment of the operational benefit forecast results and operational strategies is triggered.

9. The method according to claim 1, characterized in that, The step of obtaining initial power information also includes: The acquired initial power information is preprocessed, including timestamp alignment, data format conversion, and data integrity checks.

10. A distributed photovoltaic power generation operation benefit prediction system, characterized in that, The system includes: The information acquisition module is used to acquire initial power information, which includes power market transaction information, power grid operation parameters at the grid connection point, operation information of other energy facilities in the region, local meteorological information, and facility load information. The disturbance measurement module is used to perform small-power disturbances on the power grid using a battery energy storage system and measure the power grid's response information. The risk analysis module is used to comprehensively analyze the initial power information and the response information to generate power grid operation limitation risk indicators; The forecasting and adjustment module is used to adjust the forecast results of the operating benefits of energy assets based on the power grid operation limitation risk indicators. The strategy adjustment module is used to adjust the operation strategy of energy assets based on the power grid operation limitation risk indicators.

Images