A time-series power grid planning method based on renewable energy consumption rate

Abstract

The present invention discloses a time-series power grid planning method based on the renewable energy consumption rate, and specifically relates to the technical field of power grid planning. By collecting and analyzing the renewable energy output fluctuation and load change data in different time periods in the power grid, a new energy and load fluctuation curve is constructed, and the severity of harmonic distortion is evaluated in combination with the real-time data of the harmonic monitoring equipment, the load influence of the harmonic on the reactive power regulation device and the reactive power margin fluctuation are identified, and the power quality risk is quantified based on the comprehensive analysis result, and the time period is divided into high risk, medium risk and low risk levels, and a power quality improvement plan is formulated. In the medium risk time period, by predicting the voltage support risk, the output of the reactive power regulation device and the activation frequency of the harmonic filter are dynamically adjusted to ensure the voltage stability of the power grid under the condition of harmonic fluctuation, thereby significantly improving the power quality and the safety and stability of renewable energy consumption.

Description

Technical Field

[0001] The present invention relates to the technical field of power grid planning, and in particular to a time-series power grid planning method based on new energy consumption rate. Background Art

[0002] Sequential grid planning based on the absorption rate of new energy refers to the consideration of the time dynamic characteristics of the absorption rate of new energy (such as wind energy, solar energy, etc.) in the planning of the power system to optimize the construction and operation layout of the power grid. The absorption rate of new energy refers to the actual utilization rate of new energy by the power grid, that is, the proportion of new energy power generation in the total power generation. This indicator reflects the adequacy of the utilization of new energy. Sequential grid planning is to gradually optimize the layout and operation strategy of the power grid in stages according to the time variation law of power load and new energy power generation, so as to adapt to the fluctuation of new energy output, improve the absorption rate, and reduce the phenomenon of wind and solar power abandonment.

[0003] Specifically, the planning method first needs to analyze the power load curve of the power system, the characteristics of renewable energy generation in different time periods, and their fluctuation range. By introducing a variety of timing optimization models, the output of renewable energy can be predicted in different time periods (such as day, season, and year), and a dynamic grid expansion plan can be formulated. However, due to the unstable generation frequency of renewable energy, short-term large fluctuations may affect the power quality, including voltage fluctuations, frequency imbalance, insufficient reactive power compensation, and other problems. The timing optimization model mainly focuses on the matching of load and output, and has limited consideration of the impact on the power quality of the power grid. Poor power quality can cause damage to power equipment at the user end, increase the loss costs of power users and power grid companies, and even cause frequent malfunctions of the power grid protection system. At the same time, if the power quality is poor for a long time, it may cause an imbalance in the distribution of reactive power in the power grid. Especially during periods of high renewable energy output, the reactive power demand of the power grid may not be fully compensated, resulting in insufficient voltage support. The timing optimization model usually does not make detailed plans for reactive power, which makes the power grid face the risk of reactive voltage out of control in certain periods. Summary of the invention

[0004] The purpose of the present invention is to provide a time-series power grid planning method based on the new energy consumption rate to solve the shortcomings of the background technology.

[0005] In order to achieve the above object, the present invention provides the following technical solution: a time-series power grid planning method based on new energy consumption rate, comprising the following steps:

[0006] S1: Collect the output fluctuation data of renewable energy in several time periods and the dynamic change data of load, and construct the fluctuation curves of load and renewable energy output in each time period respectively;

[0007] S2: Based on the fluctuation curve of renewable energy output in each time period, and the current and voltage content of each harmonic measured by the harmonic monitoring equipment installed at the grid access point, the severity of harmonic distortion caused by the fluctuation of renewable energy output is evaluated;

[0008] S3: If the severity of harmonic distortion is high, then based on the harmonic content data, analyze the consumption status of each harmonic on the reactive power of the power grid, determine the impact of harmonics on the load of the reactive power regulation device, and evaluate the abnormal fluctuation of the reactive power margin of the reactive power regulation device under harmonic conditions;

[0009] S4: Comprehensively analyze the severity of harmonic distortion caused by fluctuations in renewable energy output and abnormal fluctuations in reactive power margin of reactive power regulation devices under harmonic conditions, quantify the risk level of power quality in each time period based on the analysis results, and evaluate the voltage support risk faced by the power grid in different time periods;

[0010] S5: Divide the risk level of power quality in each time period into high-risk time period, medium-risk time period and low-risk time period, and formulate corresponding power quality improvement plans;

[0011] S6: Predict the severity of the voltage support risk faced by the power grid during the medium-risk period, and dynamically adjust the reactive output of the reactive power regulation device and the activation frequency of the harmonic filter based on the prediction results.

[0012] Preferably, in S2, the correlation between the output fluctuation of new energy and each harmonic is analyzed to generate a harmonic distortion correlation anomaly index, and the method for obtaining the harmonic distortion correlation anomaly index is:

[0013] Collect the new energy output fluctuation data and the current and voltage content data of each harmonic in different time periods to form a matrix , where n is the number of data samples, m is the number of features included, R is the real number domain, and the matrix X is standardized to obtain the standardized matrix Z. The covariance matrix C is calculated for the standardized data matrix Z, and the expression is: ; In the formula, T is the matrix transpose, and the covariance matrix C is decomposed by eigenvalue to obtain the eigenvalue and the corresponding eigenvector, which is expressed as: ; In the formula, λi is the i-th eigenvalue of the covariance matrix, which represents the variance of the i-th principal component, and vi is the i-th eigenvector corresponding to the covariance matrix. According to the cumulative variance contribution rate, the first k eigenvalues ​​and corresponding eigenvectors are selected to form the principal components, and the original data is reduced to k-dimensional space. The data matrix after dimensionality reduction is: ; In the formula, Y is the data matrix after dimension reduction, Vk is the matrix composed of the first k eigenvectors, and the clustering algorithm is used to cluster the data Y after dimension reduction to identify normal data and abnormal data patterns. After clustering analysis of Y, we get Cluster centers, the expression is: ; In the formula, Cj represents the cluster center of the jth class, is the sample set belonging to the jth class, is the number of samples in the jth category, and the distance from each sample Yi to the center Cj of its category is calculated, that is, the harmonic distortion correlation anomaly index is calculated, and the expression is: ; Where HA is the harmonic distortion correlation anomaly index.

[0014] Preferably, in S3, after analyzing the fluctuation of the reactive margin residual value of the reactive regulation device, a reactive margin residual value fluctuation index is generated, and the method for obtaining the reactive margin residual value fluctuation index is:

[0015] Get the time series data of the reactive power margin residual value , where t represents time, from the sequence Find all local maxima and minima in , construct the upper envelope and lower envelope respectively through the local maxima and local minima, calculate the average value of the upper envelope and the lower envelope, get the envelope mean m1, and subtract the envelope mean from the original signal: ; Check whether h1 satisfies the IMF condition. If so, h1 is the first IMF, denoted as , remove the extracted IMF from the original signal, and finally obtain several intrinsic mode functions and residual , so that: ; According to the fluctuation range of each IMF, the fluctuation index of reactive margin residual value is calculated, and the expression is; ; In the formula, is the value of the ith IMF at time t, is the mean of the ith IMF, It is the reactive power margin residual value fluctuation index.

[0016] Preferably, in S4, a comprehensive analysis is performed on the severity of harmonic distortion caused by the fluctuation of the output of renewable energy and the abnormal fluctuation of the reactive margin of the reactive power regulation device under harmonic conditions, and the risk level of the power quality in each time period is quantified according to the analysis results, specifically:

[0017] The harmonic distortion correlation anomaly index and the reactive margin residual value fluctuation index are converted into the first eigenvector, and the first eigenvector is used as the input of the machine learning model. The machine learning model predicts the risk value label of the power quality in each time period with each group of first eigenvectors as the prediction target, and minimizes the sum of the prediction errors of the risk value labels of all power quality in each time period as the training target. The machine learning model is trained until the sum of the prediction errors converges and the model training is stopped. The risk value of the power quality in each time period is determined according to the model output results, wherein the machine learning model is a polynomial regression model.

[0018] Preferably, in S5, the risk level of power quality in each time period is divided into a high-risk time period, a medium-risk time period and a low-risk time period, specifically:

[0019] The risk value of the power quality in each time period is compared with the gradient risk threshold, the gradient risk threshold includes a first risk threshold and a second risk threshold, and the first risk threshold is less than the second risk threshold, and the risk value of the power quality in each time period is compared with the first risk threshold and the second risk threshold;

[0020] If the risk value of power quality in the time period is greater than the second risk threshold, it means that the voltage support risk faced by the power grid in the time period is high. At this time, a first-level warning signal is generated and the corresponding time period is divided into a high-risk time period;

[0021] If the risk value of power quality in the time period is greater than or equal to the first risk threshold and less than or equal to the second risk threshold, it means that the voltage support risk faced by the power grid in the time period is general. At this time, a secondary warning signal is generated and the corresponding time period is divided into a medium risk time period.

[0022] If the risk value of power quality in the time period is less than the first risk threshold, it means that the voltage support risk faced by the power grid in the time period is low. At this time, a third-level warning signal is generated and the corresponding time period is divided into a low-risk time period.

[0023] Preferably, in S6, for a medium-risk time period, that is, the risk value of the power quality in the time period is greater than or equal to the first risk threshold and less than or equal to the second risk threshold, the risk values ​​in subsequent time periods that are greater than or equal to the first risk threshold and less than or equal to the second risk threshold are collected, a data set is established, and the mean and standard deviation of the data set are calculated. After analyzing the data set, the severity of the voltage support risk faced by the power grid in the medium-risk time period is predicted, and the reactive output of the reactive regulation device and the activation frequency of the harmonic filter are dynamically adjusted according to the prediction results.

[0024] Preferably, if the mean of the risk values ​​in the data set is greater than or equal to the reference threshold of the mean of the risk values, and the standard deviation of the risk values ​​is less than the reference threshold of the standard deviation of the risk values, the risk value is high overall but has small fluctuations, indicating that the voltage support risk is high and stable, and the output of the reactive power regulation device is increased to keep the filter working at a low activation frequency;

[0025] If the risk value mean is greater than or equal to the reference threshold of the risk value mean, and the risk value standard deviation is greater than or equal to the reference threshold of the risk value standard deviation, the risk value is high and fluctuates greatly, indicating that the voltage support risk is high and unstable, and the reactive output of the reactive power regulation device is increased, and the activation frequency of the harmonic filter is increased;

[0026] If the risk value mean is less than the reference threshold of the risk value mean, and the risk value standard deviation is greater than or equal to the reference threshold of the risk value standard deviation, the overall risk value is low but the volatility is large, indicating that the voltage support risk is small, the existing reactive power regulation output is maintained, and the activation frequency of the filter is increased;

[0027] If the risk value mean is less than the reference threshold of the risk value mean, and the risk value standard deviation is less than the reference threshold of the risk value standard deviation, the risk value is low and the fluctuation is small, the existing output of the reactive power regulation device is maintained, and the filter maintains a low activation frequency.

[0028] In the above technical solution, the technical effects and advantages provided by the present invention are:

[0029] 1. The present invention collects new energy output and load fluctuation data, and constructs the harmonic distortion correlation anomaly index and reactive margin residual value fluctuation index in combination with harmonic monitoring data to comprehensively evaluate the power quality and voltage support risk of the power grid. This method can not only quantify the power quality risk value of each time period through a machine learning model, but also divide the high, medium and low risk time periods based on the risk level, and generate corresponding early warning signals to timely adjust the reactive power regulation device output and harmonic filter frequency of the power grid, thereby improving the dynamic response capability and stability of the power grid.

[0030] 2. The present invention effectively solves the problems of harmonic distortion and reactive power regulation load fluctuation caused by fluctuations in the output of new energy, and ensures the stability of the power quality and voltage support of the power grid when the output of new energy fluctuates frequently. Especially during the medium-risk period, the operating status of the reactive power regulation device and the filter is dynamically adjusted through data set analysis and risk prediction, which optimizes reactive power distribution and harmonic control, helps to improve the safety, economy and power quality management level of the power grid, and improves the absorption capacity of new energy and the reliability of power grid operation. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0032] Figure 1 The figure is a flow chart of the method of the present invention. DETAILED DESCRIPTION

[0033] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0034] For examples, see Figure 1 As shown, the time-series power grid planning method based on the new energy consumption rate described in this embodiment includes the following steps:

[0035] S1: Collect the output fluctuation data of renewable energy in several time periods and the dynamic change data of load, and construct the fluctuation curves of load and renewable energy output in each time period respectively;

[0036] S2: Based on the fluctuation curve of renewable energy output in each time period, and the current and voltage content of each harmonic measured by the harmonic monitoring equipment installed at the grid access point, the severity of harmonic distortion caused by the fluctuation of renewable energy output is evaluated;

[0037] S3: If the severity of harmonic distortion is high, then based on the harmonic content data, analyze the consumption status of each harmonic on the reactive power of the power grid, determine the impact of harmonics on the load of the reactive power regulation device, and evaluate the abnormal fluctuation of the reactive power margin of the reactive power regulation device under harmonic conditions;

[0038] S4: Comprehensively analyze the severity of harmonic distortion caused by fluctuations in renewable energy output and abnormal fluctuations in reactive power margin of reactive power regulation devices under harmonic conditions, quantify the risk level of power quality in each time period based on the analysis results, and evaluate the voltage support risk faced by the power grid in different time periods;

[0039] S5: Divide the risk level of power quality in each time period into high-risk time period, medium-risk time period and low-risk time period, and formulate corresponding power quality improvement plans;

[0040] S6: Predict the severity of the voltage support risk faced by the power grid during the medium-risk period, and dynamically adjust the reactive output of the reactive power regulation device and the activation frequency of the harmonic filter based on the prediction results.

[0041] Among them, in S1, the output fluctuation data of renewable energy in several time periods and the dynamic change data of load are collected, and the fluctuation curves of load and renewable energy output in each time period are constructed respectively, specifically:

[0042] Identify data sources, including renewable energy output data (such as wind power and photovoltaic output), load demand data, meteorological data (such as wind speed, light intensity, temperature), etc. Data sources usually include monitoring systems for renewable energy power generation equipment, grid load monitoring systems, and meteorological monitoring agencies. Determine the frequency and cycle of data sampling. Renewable energy output and load demand usually fluctuate rapidly, and it is recommended to use minutes or hours as the sampling interval. The time range of data collection should cover different time periods such as days, seasons, and years to ensure data integrity.

[0043] By connecting to the monitoring systems of new energy equipment such as wind power and photovoltaic power generation, real-time output data of new energy power generation in each time period is collected. The data includes the instantaneous output and total output of new energy, and records the output at different time points. Meteorological data of each time period, such as light intensity, wind speed, humidity, etc., are collected simultaneously to facilitate the subsequent interpretation and analysis of the fluctuations in new energy output. The original output data is denoised to remove outliers and fill in missing data to ensure the continuity and accuracy of the data. The preprocessed data can form a multidimensional time series data set, including output and meteorology.

[0044] The power grid load monitoring system is used to collect power load data in various time periods. The data should include total load, peak and valley load values, and hourly load change rates to ensure that the data can reflect load fluctuations in different time periods. Peak load, valley load, and their changing trends are extracted from the load data, and the load fluctuation cycle and typical characteristics are analyzed to lay the foundation for subsequent comparative analysis with new energy output. Data smoothing technology (such as the moving average method) is used to eliminate noise, make the load change curve clearer and more stable, and process abnormal load data to avoid deviations.

[0045] The data is divided into different time periods (such as day, week, month, season, year) and integrated into the load and renewable energy output data of the corresponding time period. Fluctuation curves are constructed with different time periods to ensure that the load and output fluctuation characteristics of different time scales can be reflected in the curve.

[0046] Draw fluctuation curves: Draw fluctuation curves of renewable energy output in different time periods, such as daily output fluctuation curves, seasonal fluctuation curves, etc. This curve should be able to show the changing trend and fluctuation range of renewable energy output in different time periods such as one day and one month. Draw the change curve of grid load to show the dynamic change characteristics of load over time to identify peak and valley load periods. Draw load curves of different time periods separately for subsequent analysis.

[0047] The new energy output curve and the load change curve are superimposed to analyze the fluctuation patterns of the two in different time periods. It is possible to identify whether there is a timing misalignment between the new energy output peak period and the load trough period, providing data support for subsequent power grid optimization and dispatching strategies.

[0048] According to the fluctuation curve, the fluctuation characteristics of each time period are analyzed, including the peak value, valley value, fluctuation range of renewable energy power generation, and the periodic change characteristics of load. The load and output fluctuation curves and related data of different time periods are stored in the database or data management system to facilitate subsequent modeling, prediction and analysis. A regular data update mechanism is established to cope with the new characteristics of renewable energy output and load demand over time, and to ensure the real-time and accuracy of the fluctuation curve.

[0049] S2: Based on the fluctuation curve of renewable energy output in each time period, and the current and voltage content of each harmonic measured by the harmonic monitoring equipment installed at the grid access point, the severity of harmonic distortion caused by the fluctuation of renewable energy output is evaluated.

[0050] According to the output fluctuation curve of renewable energy, analyze the frequency and amplitude of output changes. Focus on time periods with large fluctuation amplitudes or high frequencies (such as wind power output when wind speed changes dramatically, or photovoltaic output when light intensity changes frequently). These time periods with large fluctuations often cause higher harmonics. Use harmonic monitoring equipment installed at the grid access point to measure the harmonic content generated when renewable energy power generation output fluctuates, including the current and voltage values ​​of low-order harmonics such as the 3rd, 5th, and 7th order and other harmonics. Divide by time period and record the total harmonic distortion (THD) of the harmonics in each time period.

[0051] After analyzing the correlation between the output fluctuation of new energy and each harmonic, the harmonic distortion correlation anomaly index is generated. The method for obtaining the harmonic distortion correlation anomaly index is as follows:

[0052] Collect the new energy output fluctuation data and the current and voltage content data of each harmonic in different time periods to form a matrix , where n is the number of data samples (i.e., the number of time periods), m is the number of features included, such as output fluctuations, current and voltage values ​​of each harmonic, etc., and R is the real number domain. After standardizing the matrix X, the standardized matrix Z is obtained. The covariance matrix C is calculated for the standardized data matrix Z, and the expression is: ; In the formula, T is the matrix transpose, and the covariance matrix C is decomposed by eigenvalue to obtain the eigenvalue and the corresponding eigenvector, which is expressed as: ; In the formula, λi is the i-th eigenvalue of the covariance matrix, which represents the variance of the i-th principal component, and vi is the i-th eigenvector corresponding to the covariance matrix. According to the cumulative variance contribution rate (generally the first 95%), the first k eigenvalues ​​and corresponding eigenvectors are selected to form the principal components, and the original data is reduced to the k-dimensional space. The data matrix after dimensionality reduction is: ; In the formula, Y is the data matrix after dimension reduction, Vk is the matrix composed of the first k eigenvectors, and clustering algorithms (such as K-means or DBSCAN) are used to cluster the dimension-reduced data Y to identify normal data and abnormal data patterns. Taking K-means as an example, after clustering analysis of Y, we get Cluster centers, the expression is: ; In the formula, Cj represents the cluster center of the jth class, is the sample set belonging to the jth class, is the number of samples in the jth category. The distance from each sample Yi to the center Cj of its category is calculated to indicate its degree of deviation from the normal mode, that is, to calculate the harmonic distortion correlation anomaly index, which is expressed as: ; Where HA is the harmonic distortion correlation anomaly index.

[0053] The larger the harmonic distortion correlation anomaly index is, the more serious the harmonic distortion anomaly caused by the fluctuation of renewable energy output is, that is, the severity of harmonic distortion is high. A larger harmonic distortion correlation anomaly index indicates that the frequent fluctuation of renewable energy output is strongly correlated with each harmonic, resulting in obvious harmonic distortion, which may have exceeded the normal tolerance range of the power grid. In this case, the harmonic content in the power grid increases, resulting in increased voltage fluctuations, decreased power quality, and may even affect the normal operation of sensitive equipment in the power grid. A continuously large harmonic distortion correlation anomaly index usually indicates that compensation measures need to be taken during this time period, such as the dispatch of reactive compensation devices or the activation of filtering equipment, to reduce the impact of harmonics on the power grid.

[0054] On the contrary, the smaller the harmonic distortion correlation anomaly index is, the less impact the fluctuation of renewable energy output has on harmonic distortion, that is, the severity of harmonic distortion is low, and the voltage, frequency and reactive power levels in the power grid system are relatively stable. This indicates that during this period, the change in renewable energy output was small, or the output fluctuation did not significantly increase the harmonic distortion. A smaller harmonic distortion correlation anomaly index usually means that the power quality remains at a good level and the voltage fluctuation of the power grid is controlled within a safe range. At this time, no additional reactive compensation or filtering measures are required, which can reduce the operating cost of the power grid and ensure the economy and stability of the power grid.

[0055] S3: If the severity of harmonic distortion is high, then based on the harmonic content data, analyze the consumption status of each harmonic on the reactive power of the power grid, determine the impact of harmonics on the load of the reactive power regulation device, and evaluate the abnormal fluctuation of the reactive power margin of the reactive power regulation device under harmonic conditions.

[0056] Harmonic monitoring equipment is used to collect current and voltage data of harmonics of different frequencies (such as 3rd, 5th, 7th harmonics, etc.), and the reactive power of each harmonic is calculated based on these data. The calculation formula for reactive power is: ; Where Qh is the reactive power of the h-th harmonic, Vh is the effective value of the voltage and current of the h-th harmonic, and φh is the phase difference between the current and voltage of the h-th harmonic.

[0057] Summarize the total reactive power consumption of each harmonic, calculate the total reactive power consumption under harmonic conditions, and determine its impact on the overall reactive power demand of the power grid. Add the reactive power consumption of all harmonics to get the total reactive power consumption under harmonic distortion conditions. , the expression is: ; The total reactive power consumption value Qtotal is obtained, indicating the total impact of harmonics on the demand for reactive power regulation equipment, providing a reference for subsequent reactive power regulation load analysis. Combining the rated capacity Qrated of the reactive power regulation device and the current harmonic reactive power consumption Qtotal, the working load rate of the reactive power regulation device is calculated, and the expression is: ; When the load factor approaches or exceeds the safety threshold (usually 80-90%), it means that the load on the reactive power regulation device is high and may face overload risk.

[0058] Calculate the reactive margin Qmargin of the reactive power regulation device, the expression is: , analyze the residual value of reactive power margin and record it in each time period. If the reactive power margin is too low or fluctuates frequently, it may cause the risk of insufficient voltage support.

[0059] After analyzing the fluctuation of the reactive margin residual value of the reactive regulation device, a reactive margin residual value fluctuation index is generated. The method for obtaining the reactive margin residual value fluctuation index is:

[0060] Get the time series data of the reactive power margin residual value , where t represents time, from the sequence Find all local maxima and minima in , construct the upper envelope and lower envelope respectively through the local maxima and local minima, calculate the average value of the upper envelope and the lower envelope, get the envelope mean m1, and subtract the envelope mean from the original signal: ; Check whether h1 satisfies the intrinsic mode function IMF condition (that is, the number of positive and negative zero crossing points and extreme value points of the signal in the entire time range is almost equal, and the local average value is zero). If it is satisfied, h1 is the first IMF, denoted as If the IMF condition is not met, repeat the above steps to iterate h1 until an IMF that meets the condition is extracted. Remove the extracted IMF from the original signal and repeat the above steps until the remaining part becomes a monotonic function or contains only trend components, and finally obtain several intrinsic mode functions. and residual , so that: ; According to the fluctuation range of each IMF, the fluctuation index of reactive margin residual value is calculated, and the expression is; ; In the formula, is the value of the ith IMF at time t, is the mean of the ith IMF, It is the reactive power margin residual value fluctuation index.

[0061] The larger the reactive margin residual value fluctuation index is, the more drastic the load fluctuation of the reactive power regulation device is under harmonic conditions. A larger fluctuation index means that the reactive power regulation device needs to frequently adjust its output to cope with the reactive power demand fluctuation caused by harmonics, which increases the load pressure and operational instability of the equipment. In this case, the reactive power margin of the regulation device is low or in a state of rapid consumption, the voltage support capacity of the power grid is weakened, and there may even be risks such as overload or overheating of the regulation device, resulting in a decrease in power quality, which in turn affects the normal operation of the user-side equipment.

[0062] On the contrary, when the reactive margin residual value fluctuation index is small, it means that the impact of harmonics on the reactive regulation device is relatively limited, the equipment load is relatively stable, and the fluctuation amplitude is small. This shows that the regulation demand of the reactive regulation device is low, its reactive margin is sufficient, the equipment can be maintained in a relatively ideal working state, and the voltage support capacity of the power grid is stable. Under such conditions, the impact of harmonics on the load of the reactive regulation device is slight, the power quality is high, the overall operation of the power grid is more reliable, and the stability of the power grid under harmonic conditions is guaranteed.

[0063] S4: Conduct a comprehensive analysis of the severity of harmonic distortion caused by fluctuations in renewable energy output and the abnormal fluctuations in reactive power margin of reactive power regulation devices under harmonic conditions. Based on the analysis results, quantify the risk level of power quality in each time period and assess the voltage support risk faced by the power grid in different time periods.

[0064] The harmonic distortion correlation anomaly index and the reactive margin residual value fluctuation index are converted into the first eigenvector, and the first eigenvector is used as the input of the machine learning model. The machine learning model predicts the risk value label of the power quality in each time period with each group of first eigenvectors as the prediction target, and minimizes the sum of the prediction errors of the risk value labels of all power quality in each time period as the training target. The machine learning model is trained until the sum of the prediction errors converges and the model training is stopped. The risk value of the power quality in each time period is determined according to the model output results, wherein the machine learning model is a polynomial regression model.

[0065] The method for obtaining the risk value of power quality in each time period is as follows: from the first eigenvector training data of the trained machine learning model, the corresponding function expression is obtained: ; In the formula, is the output function of the model, HA is the harmonic distortion correlation anomaly index, is the reactive power margin residual value fluctuation index, is the risk value of power quality in each time period.

[0066] S5: Divide the risk level of power quality in each time period into high-risk time period, medium-risk time period and low-risk time period, and formulate corresponding power quality improvement plans.

[0067] The risk value of the power quality in each time period is compared with the gradient risk threshold, the gradient risk threshold includes a first risk threshold and a second risk threshold, and the first risk threshold is less than the second risk threshold, and the risk value of the power quality in each time period is compared with the first risk threshold and the second risk threshold;

[0068] If the risk value of power quality in a time period is greater than the second risk threshold, it means that the voltage support risk faced by the power grid in this time period is higher. At this time, a first-level warning signal is generated and the corresponding time period is divided into a high-risk time period;

[0069] If the risk value of power quality in the time period is greater than or equal to the first risk threshold and less than or equal to the second risk threshold, it means that the voltage support risk faced by the power grid in the time period is general. At this time, a secondary warning signal is generated and the corresponding time period is divided into a medium risk time period;

[0070] If the risk value of power quality during the time period is less than the first risk threshold, it means that the voltage support risk faced by the power grid during the time period is lower. At this time, a third-level warning signal is generated and the corresponding time period is divided into a low-risk time period.

[0071] It should be noted that the importance of the first-level warning signal is greater than that of the second-level warning signal, and the importance of the second-level warning signal is greater than that of the third-level warning signal. Relevant personnel can formulate corresponding power quality improvement plans according to different warning signal levels, specifically:

[0072] High-risk time periods: Adopt enhanced power quality improvement solutions, such as increasing reactive power compensation capacity or using filtering equipment to reduce the impact of harmonics and reactive load fluctuations. Enable backup reactive power regulation devices or dynamically adjust energy storage systems to ensure stable voltage support.

[0073] Medium-risk period: Implement conventional power quality management measures, including enabling some reactive power compensation devices and adjusting the charging and discharging strategies of energy storage devices to smooth out power fluctuations. Monitor reactive power margin and harmonic levels through the intelligent dispatching system, and respond quickly when fluctuations intensify.

[0074] Low-risk time period: Maintain the operating status of existing regulating devices, appropriately reduce the adjustment frequency, ensure economical and efficient operation of equipment, and reduce operating costs.

[0075] S6: Predict the severity of the voltage support risk faced by the power grid during the medium-risk period, and dynamically adjust the reactive output of the reactive power regulation device and the activation frequency of the harmonic filter based on the prediction results.

[0076] For medium-risk time periods, that is, the risk value of power quality in the time period is greater than or equal to the first risk threshold and less than or equal to the second risk threshold, the risk values ​​greater than or equal to the first risk threshold and less than or equal to the second risk threshold in subsequent time periods are collected, a data set is established, and the mean and standard deviation of the data set are calculated. After analysis, the severity of the voltage support risk faced by the power grid in the medium-risk time period is predicted, and the reactive output of the reactive regulation device and the activation frequency of the harmonic filter are dynamically adjusted according to the prediction results.

[0077] If the mean risk value in the data set is greater than or equal to the reference threshold of the mean risk value, and the standard deviation of the risk value is less than the reference threshold of the standard deviation of the risk value, the overall risk value is high but the fluctuation is small, indicating that the voltage support risk is high and stable. By appropriately increasing the output of the reactive power regulation device and keeping the filter working at a lower activation frequency, the stability of the voltage support can be ensured;

[0078] If the risk value mean is greater than or equal to the reference threshold of the risk value mean, and the risk value standard deviation is greater than or equal to the reference threshold of the risk value standard deviation, the risk value is high and fluctuates greatly, indicating that the voltage support risk is high and unstable, and it is necessary to significantly increase the reactive output of the reactive regulation device and increase the activation frequency of the harmonic filter to ensure that the power grid can quickly respond to unstable voltage support needs;

[0079] If the risk value mean is less than the reference threshold of the risk value mean, and the risk value standard deviation is greater than or equal to the reference threshold of the risk value standard deviation, the risk value is generally low but has high volatility, indicating that the voltage support risk is small, but there is a certain degree of volatility uncertainty. The existing reactive power regulation output can be maintained, and the activation frequency of the filter can be appropriately increased to cope with unstable fluctuations.

[0080] If the risk value mean is less than the reference threshold of the risk value mean, and the risk value standard deviation is less than the reference threshold of the risk value standard deviation, the risk value is low and the fluctuation is small, indicating that the voltage support risk is small and stable. The existing output of the reactive power regulation device can be maintained, and the filter can maintain a low activation frequency to ensure economy and stability.

[0081] In this embodiment, first, by collecting the output fluctuation and load change data of new energy in multiple time periods, the fluctuation curve of each time period is constructed; based on these curves and the measurement data of the harmonic monitoring equipment, the severity of the harmonic distortion caused by the output fluctuation of new energy is evaluated. If the harmonic distortion is serious, the influence of each harmonic on the reactive power consumption of the power grid is further analyzed, and the fluctuation of the reactive margin of the reactive power regulation device under harmonic conditions is evaluated. Next, the harmonic distortion and abnormal fluctuation of reactive margin caused by the output fluctuation of new energy are comprehensively considered to quantify the risk level of power quality in each time period, and the time periods are divided into high-risk, medium-risk and low-risk time periods, so as to formulate corresponding power quality improvement measures. In the medium-risk time period, the voltage support risk of the power grid is predicted, and the output of the reactive power regulation device and the activation frequency of the harmonic filter are dynamically adjusted according to the prediction results to ensure the voltage stability of the power grid under harmonic conditions.

[0082] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0083] The above embodiments can be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented by software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website site, computer, server or data center to another website site, computer, server or data center by wired or wireless (e.g., infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that contains one or more available media sets. The available medium can be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state hard disk.

[0084] It should be understood that the term "and / or" in this article is only a description of the association relationship of associated objects, indicating that there may be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone, where A and B can be singular or plural. In addition, the character " / " in this article generally indicates that the objects associated with each other are in an "or" relationship, but it may also indicate an "and / or" relationship, which can be understood by referring to the context. A person of ordinary skill in the art can appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed in this article 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 to exceed the scope of this application.

[0085] The above description is only a specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application.

Claims

1. A time-series power grid planning method based on new energy consumption rate, characterized by: The steps include: S1: Collect the output fluctuation data of renewable energy in several time periods and the dynamic change data of load, and construct the fluctuation curves of load and renewable energy output in each time period respectively; S2: Based on the fluctuation curve of renewable energy output in each time period, and the current and voltage content of each harmonic measured by the harmonic monitoring equipment installed at the grid access point, the severity of harmonic distortion caused by the fluctuation of renewable energy output is evaluated; S3: If the severity of harmonic distortion is high, then based on the harmonic content data, analyze the consumption status of each harmonic on the reactive power of the power grid, determine the impact of harmonics on the load of the reactive power regulation device, and evaluate the abnormal fluctuation of the reactive power margin of the reactive power regulation device under harmonic conditions; S4: Comprehensively analyze the severity of harmonic distortion caused by fluctuations in renewable energy output and abnormal fluctuations in reactive power margin of reactive power regulation devices under harmonic conditions, quantify the risk level of power quality in each time period based on the analysis results, and evaluate the voltage support risk faced by the power grid in different time periods; S5: Divide the risk level of power quality in each time period into high-risk time period, medium-risk time period and low-risk time period, and formulate corresponding power quality improvement plans, which include: During high-risk periods: adopt enhanced power quality improvement plans, including increasing reactive power compensation capacity or using filtering equipment to reduce the impact of harmonics and reactive load fluctuations; Medium-risk period: Implement conventional power quality management measures, including enabling some reactive power compensation devices and adjusting the charging and discharging strategies of energy storage devices; Low-risk time period: maintain the operation status of existing regulating devices and reduce the frequency of regulation; S6: Predict the severity of the voltage support risk faced by the power grid during the medium-risk period, and dynamically adjust the reactive output of the reactive power regulation device and the activation frequency of the harmonic filter based on the prediction results.

2. A time-series power grid planning method based on new energy consumption rate according to claim 1, characterized in that: In S2, the correlation between the output fluctuation of new energy and each harmonic is analyzed to generate the harmonic distortion correlation anomaly index. The method for obtaining the harmonic distortion correlation anomaly index is: Collect the new energy output fluctuation data and the current and voltage content data of each harmonic in different time periods to form a matrix X∈R n×m , where n is the number of data samples, m is the number of features included, R is the real number domain, and the matrix X is standardized to obtain the standardized matrix Z. The covariance matrix C is calculated for the standardized data matrix Z, and the expression is: In the formula, T is the matrix transpose, and the covariance matrix C is decomposed by eigenvalue to obtain the eigenvalue and the corresponding eigenvector, which is expressed as: Cvi = λivi; in the formula, λi is the i-th eigenvalue of the covariance matrix, which represents the variance of the i-th principal component, and vi is the i-th eigenvector corresponding to the covariance matrix. According to the cumulative variance contribution rate, the first k eigenvalues ​​and the corresponding eigenvectors are selected to form the principal component, and the original data is reduced to the k-dimensional space. The data matrix after dimensionality reduction is: Y = Z*Vk; in the formula, Y is the data matrix after dimensionality reduction, and Vk is the matrix composed of the first k eigenvectors. The clustering algorithm is used to cluster the data Y after dimensionality reduction to identify normal data and abnormal data patterns. After clustering analysis of Y, C1, C2, ..., Ck cluster centers are obtained, which are expressed as: Where Cj represents the cluster center of the jth class, Yj is the sample set belonging to the jth class, |Yj| is the number of samples in the jth class, and the distance from each sample Yi to the center Cj of its class is calculated, that is, the harmonic distortion correlation anomaly index is calculated, and the expression is: HA = ||Yi-Cj||; where HA is the harmonic distortion correlation anomaly index.

3. A time-series power grid planning method based on new energy consumption rate according to claim 2, characterized in that: In S3, the reactive margin residual value fluctuation situation of the reactive regulation device is analyzed to generate a reactive margin residual value fluctuation index, and the method for obtaining the reactive margin residual value fluctuation index is: Get the time series data Q of the reactive power margin residual value t = {Q1,Q2,…,Q n }, where t represents time, from the sequence Q t Find all local maxima and minima in , construct the upper envelope and lower envelope respectively through the local maxima and local minima, calculate the average value of the upper envelope and the lower envelope, get the envelope mean m1, subtract the envelope mean from the original signal: h1 = Q t -m1; check whether h1 satisfies the IMF condition. If so, h1 is the first IMF, denoted as C1. The extracted IMF is removed from the original signal, and finally several IMFs C1, C2, ..., C are obtained. k and the residual r k , so that: According to the fluctuation range of each IMF, the fluctuation index of reactive margin residual value is calculated, and the expression is: In the formula, C i,t is the value of the ith IMF at time t, μ Ci is the mean value of the ith IMF, and DF is the reactive margin residual value fluctuation index.

4. A time-series power grid planning method based on new energy consumption rate according to claim 3, characterized in that: In S4, a comprehensive analysis is conducted on the severity of harmonic distortion caused by fluctuations in renewable energy output and the abnormal fluctuations in reactive margin of reactive power regulation devices under harmonic conditions. Based on the analysis results, the risk level of power quality in each time period is quantified, specifically: The harmonic distortion correlation anomaly index and the reactive margin residual value fluctuation index are converted into the first eigenvector, and the first eigenvector is used as the input of the machine learning model. The machine learning model predicts the risk value label of the power quality in each time period with each group of first eigenvectors as the prediction target, and minimizes the sum of the prediction errors of the risk value labels of all power quality in each time period as the training target. The machine learning model is trained until the sum of the prediction errors converges and the model training is stopped. The risk value of the power quality in each time period is determined according to the model output results, wherein the machine learning model is a polynomial regression model.

5. A time-series power grid planning method based on new energy consumption rate according to claim 4, characterized in that: In S5, the risk level of power quality in each time period is divided into high-risk time period, medium-risk time period and low-risk time period, specifically: The risk value of the power quality in each time period is compared with the gradient risk threshold, the gradient risk threshold includes a first risk threshold and a second risk threshold, and the first risk threshold is less than the second risk threshold, and the risk value of the power quality in each time period is compared with the first risk threshold and the second risk threshold; If the risk value of power quality in the time period is greater than the second risk threshold, it means that the voltage support risk faced by the power grid in the time period is high. At this time, a first-level warning signal is generated and the corresponding time period is divided into a high-risk time period; If the risk value of power quality in the time period is greater than or equal to the first risk threshold and less than or equal to the second risk threshold, it means that the voltage support risk faced by the power grid in the time period is general. At this time, a secondary warning signal is generated and the corresponding time period is divided into a medium risk time period. If the risk value of power quality in the time period is less than the first risk threshold, it means that the voltage support risk faced by the power grid in the time period is low. At this time, a third-level warning signal is generated and the corresponding time period is divided into a low-risk time period.

6. The method for sequential power grid planning based on new energy consumption rate according to claim 1 is characterized in that: In S6, for the medium-risk time period, that is, the risk value of the power quality in the time period is greater than or equal to the first risk threshold and less than or equal to the second risk threshold, the risk values ​​in the subsequent time period that are greater than or equal to the first risk threshold and less than or equal to the second risk threshold are collected, a data set is established, and the mean and standard deviation of the data set are calculated. After analyzing them, the severity of the voltage support risk faced by the power grid in the medium-risk time period is predicted, and the reactive output of the reactive regulation device and the activation frequency of the harmonic filter are dynamically adjusted according to the prediction results.

7. A time-series power grid planning method based on new energy consumption rate according to claim 6, characterized in that: If the mean risk value in the data set is greater than or equal to the reference threshold of the mean risk value, and the standard deviation of the risk value is less than the reference threshold of the standard deviation of the risk value, the overall risk value is high but the fluctuation is small, indicating that the voltage support risk is high and stable, and the output of the reactive power regulation device is increased to keep the filter working at a low activation frequency; If the risk value mean is greater than or equal to the reference threshold of the risk value mean, and the risk value standard deviation is greater than or equal to the reference threshold of the risk value standard deviation, the risk value is high and fluctuates greatly, indicating that the voltage support risk is high and unstable, and the reactive output of the reactive power regulation device is increased, and the activation frequency of the harmonic filter is increased; If the risk value mean is less than the reference threshold of the risk value mean, and the risk value standard deviation is greater than or equal to the reference threshold of the risk value standard deviation, the overall risk value is low but the volatility is large, indicating that the voltage support risk is small, the existing reactive power regulation output is maintained, and the activation frequency of the filter is increased; If the risk value mean is less than the reference threshold of the risk value mean, and the risk value standard deviation is less than the reference threshold of the risk value standard deviation, the risk value is low and the fluctuation is small, the existing output of the reactive power regulation device is maintained, and the filter maintains a low activation frequency.

Images