A method for assessing the operating status of a high-proportion renewable energy power system

Abstract

This invention relates to the field of load adjustment technology, specifically to a method for assessing the operating status of a high-proportion renewable energy power system. This method utilizes random sampling consensus regression analysis to extract the dominant voltage-reactive power sensitivity, and generates a positive correlation indicator bit through sensitivity polarity. Simultaneously, it analyzes the differential fluctuation characteristics of the fitted deviation sequence to determine the impedance characteristic weights. The periodic active power sequence is then corrected using these impedance characteristic weights to obtain active power benefit values. Furthermore, a risk penalty term based on voltage urgency is introduced to obtain effective load shedding value and generate a load shedding control strategy. This invention, by constructing a decoupled identification model based on the voltage-reactive power micro-response mechanism, achieves accurate assessment of feeder source-load attributes, improves the robustness of identification, and enhances the safety and accuracy of power system emergency control.

Description

Technical Field

[0001] This invention relates to the field of load adjustment technology, and more specifically to a method for assessing the operating status of a high-proportion renewable energy power system. Background Technology

[0002] As the penetration rate of distributed photovoltaic (PV) power in distribution networks continues to increase, the source-load characteristics of distribution network feeders have undergone significant changes. Traditional low-frequency load shedding strategies mainly rely on the active power measurement values ​​at substation gates to determine the priority of load shedding. However, the net active power measured at the gate is actually the algebraic sum of the actual internal load and the output of distributed power sources.

[0003] During periods of high photovoltaic (PV) output, the system may contain a significant amount of load offset by PV power. Data at the control point may mask the true load demand, making it impossible to effectively distinguish between "true low load" and "high source-load offset." Cutting off internal active power sources can result in a much lower actual load reduction than expected, and may also cause localized voltage dips due to the simultaneous disconnection of inverters with reactive power support capabilities. This information masking effect creates a risk of erroneously disconnecting lines containing a large amount of active power sources in emergency control strategies, leading to frequency degradation, localized voltage dips, and other problems that severely impact the safety and accuracy of power system emergency control. Summary of the Invention

[0004] To address the technical challenges of effectively identifying the source-load attributes within feeders based solely on gateway measurement data under high-proportion renewable energy integration, and to resolve the technical issues of erroneous disconnection of high photovoltaic penetration lines and voltage risks caused by power masking effects, this invention aims to provide a method for assessing the operational status of high-proportion renewable energy power systems. The specific technical solution adopted is as follows: This invention provides a method for assessing the operational status of a high-proportion renewable energy power system, the method comprising: The electrical quantities collected at the power system feeder junctions are processed to obtain periodic voltage sequences, periodic reactive power sequences, and periodic active power sequences. The quiescent state is determined based on the noise level of the periodic voltage sequences. In the non-quiescent state, the periodic voltage sequences and periodic reactive power sequences are standardized to construct a standardized fluctuation scatter set. A random sampling consistency regression analysis against outliers is performed on the standardized fluctuation scatter set to extract the dominant voltage-reactive power sensitivity and generate a positive correlation indicator based on the sensitivity polarity. The fitting deviation sequence of the standardized fluctuation scatter set relative to the regression benchmark is calculated, and a response continuity index is generated based on the statistical characteristics of the fitting deviation sequence. The impedance characteristic weight is generated by integrating the positive correlation indicator, the response continuity index, and the fluctuation characteristics of the periodic active power sequence. The periodic active power sequence is corrected using impedance characteristic weights to obtain active power gain value, and a voltage risk penalty term based on the current voltage level is introduced to generate effective load reduction value; an evaluation control strategy is generated based on the effective load reduction value.

[0005] Furthermore, the method for determining the silent state includes: Calculate the standard deviation of the periodic voltage sequence within a preset time window. If the standard deviation is less than the preset noise floor threshold, it is determined to be in a silent state, and the impedance characteristic weight of the previous moment is reused. If the standard deviation is greater than or equal to the preset noise floor threshold, it is determined to be in a non-silent state.

[0006] Furthermore, the method for obtaining the positive correlation indicator bit includes: The standardized fluctuation scatter set is iteratively fitted using a random sampling consensus algorithm, and the linear model with the most inliers is selected as the best baseline straight line model. The slope of the optimal reference straight line model is extracted as the dominant voltage-reactive power sensitivity; when the dominant voltage-reactive power sensitivity is greater than the preset judgment threshold, the positive correlation indicator bit is generated as a valid value; otherwise, the positive correlation indicator bit is generated as an invalid value.

[0007] Furthermore, the method for obtaining the fitting deviation sequence includes: The fitted ordinate values ​​of each point in the standardized fluctuation scatter plot are calculated using the optimal baseline linear model. The difference between the actual ordinate value and the fitted ordinate value of each point is then calculated to obtain the fitting deviation sequence.

[0008] Furthermore, the method for obtaining the response continuity index includes: Calculate the first-order difference between adjacent data points in the fitted deviation sequence of the standardized volatility scatter plot, and average the sum of squares of all difference values ​​to obtain the difference volatility. Then, combine the difference volatility with a preset sensitivity coefficient and perform an inverse proportional mapping to obtain the response continuity index.

[0009] Furthermore, the method for obtaining the impedance characteristic weights includes: The ratio of the standard deviation to the mean of the periodic active power sequence within a preset time window is calculated as the active power volatility. The environmental suppression coefficient is obtained through the active power volatility, and the environmental suppression coefficient is negatively correlated with the active power volatility. By combining the positive correlation indicator, the response continuity index, and the environmental suppression coefficient, the impedance characteristic weights are obtained.

[0010] Furthermore, the method for obtaining the effective load reduction value includes: The active power gain value is obtained by multiplying the value of the periodic active power sequence at the current moment by the impedance characteristic weight. A voltage urgency function is constructed based on the degree of drop of the current voltage level relative to the low voltage safety threshold; a voltage risk penalty term is obtained by combining the periodic active power sequence, impedance characteristic weights, and the voltage urgency function. The difference between the active power gain and the voltage risk penalty is used as the effective load reduction value.

[0011] Furthermore, the method for obtaining the voltage risk penalty term includes: Obtain the current feeder voltage per unit value; construct a voltage urgency function: when the voltage per unit value is greater than or equal to the set low voltage safety threshold, the function value is zero; when the voltage per unit value is less than the set low voltage safety threshold, calculate the deviation between the set low voltage safety threshold and the voltage per unit value, and use the product of the deviation value and the preset penalty gain coefficient as the function value; The proportion of non-impedance components is obtained based on impedance characteristic weights; the impedance characteristic weights are negatively correlated with the proportion of non-impedance components. By combining the current periodic active power sequence value, the non-impedance ratio, and the voltage urgency function value, a voltage risk penalty term is obtained.

[0012] Furthermore, the effective load reduction value generation assessment and control strategy includes: The feeder operating status is evaluated and classified based on the effective load reduction value; if the effective load reduction value is negative or zero, the feeder is marked as blocked protection; if the effective load reduction value is positive, the feeders are sorted from largest to smallest effective load reduction value to obtain the cut-off sequence; according to the evaluation and classification results, the cut-off command is generated in the order of the cut-off sequence.

[0013] Furthermore, the method for obtaining the standardized fluctuation scatter set includes: In the non-quiet state, the standard deviations of the periodic voltage sequence and the periodic reactive power sequence are calculated respectively; the corresponding standard deviations are used to normalize the periodic voltage sequence and the periodic reactive power sequence to generate a standardized scatter set of dimensionless coordinate points.

[0014] The present invention has the following beneficial effects: This invention utilizes random sampling consensus regression analysis to extract dominant voltage-reactive power sensitivity, effectively identifying positively damped physical loads that follow Ohm's law. By analyzing the differential fluctuation characteristics of the fitted deviation sequence, it captures high-frequency step noise in the digital control of power electronic equipment, further distinguishing between physical inertial components and analog control power supplies. Simultaneously, it introduces a risk penalty term based on voltage urgency to generate effective load shedding value, transforming the decision-making basis for disconnection operations into a quantitative indicator that comprehensively considers frequency gains and voltage risks. This effectively enhances the ability to prevent local voltage collapse while restoring the active power balance of the entire network. Furthermore, by constructing a decoupled identification model based on the voltage-reactive power micro-response mechanism, this invention achieves accurate assessment of feeder source-load attributes, improves the robustness of identification, and enhances the safety and accuracy of emergency control in the power system. Attached Figure Description

[0015] To more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 The flowchart illustrates a method for assessing the operational status of a high-proportion renewable energy power system, as provided in one embodiment of the present invention. Detailed Implementation

[0017] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of a method for assessing the operating status of a high-proportion renewable energy power system proposed according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0018] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0019] The following description, in conjunction with the accompanying drawings, details the specific scheme of the method for assessing the operating status of a high-proportion renewable energy power system provided by the present invention.

[0020] Please see Figure 1 The diagram illustrates a flowchart of a method for assessing the operational status of a high-proportion renewable energy power system according to an embodiment of the present invention. The method includes the following steps: S1: Process the electrical quantities collected at the power system feeder junctions to obtain periodic voltage sequences, periodic reactive power sequences, and periodic active power sequences; determine the quiescent state based on the noise level of the periodic voltage sequences; and standardize the periodic voltage sequences and periodic reactive power sequences in the non-quiescent state to construct a standardized fluctuation scatter set.

[0021] To effectively separate the micro-dynamic information reflecting the control characteristics of internal feeder equipment from the mixed power signals collected at substation gateways, a time-synchronous observation mechanism based on the power frequency cycle is first established. Since the original high-frequency sampling data is massive and contains a large amount of grid background noise unrelated to equipment regulation characteristics, such as rated voltage DC bias and slow daily load drift, directly using the raw data for analysis would lead to the masking of features.

[0022] In one specific embodiment of the present invention, voltage and current electrical quantity data are acquired at a sampling frequency of 1200Hz using voltage transformers and current transformers installed at the feeder terminals of substations. To ensure the time-series continuity and stability of the data for subsequent analysis, the global master clock for system operation is defined as the power frequency cycle index k, with k starting from index 1, i.e., k=1,2,3,… Each index k corresponds to a complete power frequency cycle, and the power frequency cycle is 0.02s.

[0023] The sampled data points are aligned using a discrete integral algorithm to obtain periodic voltage, periodic reactive, and periodic active sequences. The periodic voltage sequence is obtained by calculating the root mean square (RMS) of the instantaneous voltage values ​​within a single cycle. It is then filtered using a bandpass filter with a passband frequency of 0.2Hz to 5Hz to remove DC components and slow trends that mask the dynamic characteristics of the equipment, thus representing the effective voltage level at the feeder port. The periodic reactive sequence is calculated using known techniques such as the time-domain phase-shifting method or Hilbert transform, and similarly filtered using a bandpass filter. This sequence represents the total reactive power response of the feeder to voltage excitation. The periodic active sequence is obtained by calculating the average product of voltage and current samples within a single power frequency cycle, without bandpass filtering. It represents the direction of active energy flow at the feeder port and the power fluctuation characteristics caused by environmental factors.

[0024] Furthermore, in this embodiment of the invention, a preset time window is established to determine the steady-state operation period. The window size can be set to 150 power frequency cycles. When k is greater than or equal to the window size, it indicates that the system has entered the steady-state operation period and the analysis process can be triggered. Since the fluctuation of the periodic voltage sequence is minimal when the power grid is running smoothly, it does not elicit an effective response from the equipment. The quiescent state is a state without effective equipment response. Therefore, a preliminary quiescent state analysis is first performed using a noise threshold.

[0025] In this embodiment of the invention, the standard deviation of the periodic voltage sequence within a preset time window is first calculated to reflect the fluctuation intensity of the current grid voltage. When the standard deviation is less than a preset noise floor threshold, it indicates that the current grid has not elicited an effective equipment response, and is therefore determined to be in a silent state. At this time, in order to maintain the continuity of the system output, the impedance characteristic weights of the previous moment can be reused, and subsequent scatter set construction and feature identification are no longer performed.

[0026] When the standard deviation is greater than or equal to the preset noise floor threshold, it is determined to be a non-quiet state, indicating that there is effective excitation in the power grid, and further analysis is required. In this embodiment of the invention, the preset noise floor threshold can be set to... The per-unit value is not restricted here.

[0027] In the non-quiet state, a scatter set is constructed through standardization. In this embodiment of the invention, in the non-quiet state, the standard deviations of the periodic voltage sequence and the periodic reactive power sequence are calculated respectively. The periodic voltage sequence and the periodic reactive power sequence are then normalized using the corresponding standard deviations to generate a standardized fluctuation scatter set composed of dimensionless coordinate points. Specifically, as an example, the normalization expression is: ； In the formula, Represented as the first in a periodic voltage sequence One normalized value; Represented as the first in a periodic voltage sequence A number; It is expressed as the standard deviation of a periodic voltage sequence; Represented as the th in a periodic reactive power sequence One normalized value; Represented as the th in a periodic reactive power sequence A number; This is expressed as the standard deviation of the periodic reactive power sequence. It should be noted that the standard deviation of a determined periodic voltage sequence under non-quiet conditions cannot be zero, and reactive power fluctuations usually accompany voltage fluctuations; therefore, the standard deviation... and Since none of them are zero, there is no case where the formula is meaningless.

[0028] Therefore, the standardized volatility scatter set , Represented as a standardized scatter plot, Represented as the first coordinate points, This represents the total number of data points in the standardized fluctuation scatter set. The standardized fluctuation scatter set can characterize the voltage-reactive power micro-dynamic response relationship of the internal equipment of the feeder, eliminating dimensional differences and enabling subsequent anti-outlier regression analysis and source-load attribute identification.

[0029] S2: Perform random sampling consistency regression analysis against outliers on the standardized scatter set to extract the dominant voltage-reactive power sensitivity and generate a positive correlation indicator based on the sensitivity polarity; calculate the fitting deviation sequence of the standardized scatter set relative to the regression benchmark and generate a response continuity index based on the statistical characteristics of the fitting deviation sequence; integrate the positive correlation indicator, the response continuity index and the fluctuation characteristics of the periodic active power sequence to generate impedance characteristic weights.

[0030] In scenarios with a high proportion of renewable energy integration, the power backfeeding or local consumption of distributed generation sources can mask the true load characteristics, making a single power criterion inaccurate. Considering that the first-principles difference between physical impedance loads and power electronic power supplies lies in their "response mechanism," meaning that from a macroscopic distribution network perspective, the dominant response trend of physical load groups statistically exhibits a positive correlation, manifesting as positively damped voltage-reactive power sensitivity, and possessing microscopic trajectory inertia due to the energy storage effect of inductors and capacitors, while power electronic devices follow manually set control strategies, often exhibiting negative damping and dead-zone characteristics, and possessing microscopic trajectory step characteristics due to digital regulation. Although individual devices may exhibit nonlinearity, under the macroscopic aggregation effect at the feeder junction, physical loads and regulating power supplies show statistically significant differences in damping characteristics. Therefore, based on the constructed microscopic observation set, by decoupling the geometric slope polarity and temporal trajectory smoothness—two orthogonal features—mechanistic-level identification of feeder source-load attributes can be achieved without communication.

[0031] First, a robust outlier-resistant linear regression analysis is performed on the standardized fluctuation scatter set to extract voltage-reactive power sensitivity, i.e., the sensitivity of reactive power changes caused by voltage fluctuations. Since the control curves of distributed power sources, such as photovoltaic inverters, often contain nonlinear elements such as dead zones or saturation zones, nonlinear outliers will form in the standardized fluctuation scatter set. If the commonly used least squares method is used for regression, the nonlinear points will severely lower the fitting slope, leading to feature distortion. Therefore, the Random Sample Consensus (RANSAC) algorithm, which is robust to outliers, is selected to extract the linear model representing the main response characteristics from the mixed data.

[0032] In this embodiment of the invention, a random sampling consensus algorithm is used to iteratively fit the standardized fluctuating scatter set. In one specific implementation, the number of iterations is set to 100, and the inlier determination threshold is 0.05. In each iteration, two points in the scatter set are randomly selected to construct a temporary linear model. The Euclidean distance from all other points to this line is calculated, and points whose Euclidean distance is less than the inlier determination value are considered inliers. Finally, the linear model with the most inliers is selected as the optimal reference straight line model. This model can reflect the voltage-reactive power response trend of the dominant equipment inside the feeder to the greatest extent and eliminate nonlinear interference such as inverter control dead zone.

[0033] At this point, the slope of the optimal reference straight line model is extracted as the dominant voltage-reactive power sensitivity. The slope characterizes the average rate and direction of reactive power fluctuation with voltage fluctuation in the current analysis time sequence. It can reflect the direction and intensity of the reactive power response of the dominant equipment inside the feeder to voltage changes, serving as the core geometric feature to distinguish physical loads from power electronic equipment.

[0034] The extracted sensitivity parameters are used to verify whether the feeder conforms to the electrical characteristics of physical impedance. According to the steady-state mechanism of power systems, traditional impedance-type loads, such as motors and lighting loads, have positive damping characteristics, meaning that when the voltage increases, the inductive reactive power absorbed by them naturally increases, exhibiting a positive slope. Conversely, distributed generation is usually configured with negative damping characteristics to maintain voltage stability, that is, reducing reactive power output or increasing inductive absorption when the voltage rises. However, under generator conventions, this usually manifests as a reduction in reactive power injection, or a zero slope in the voltage dead zone.

[0035] In this embodiment of the invention, the preset judgment threshold is 0.1 to tolerate measurement errors, which can be adjusted by the implementer. When the dominant voltage-reactive power sensitivity is greater than the preset judgment threshold, a positive correlation indicator bit is generated as a valid value (1), indicating that the dominant characteristics of the corresponding feeder within the preset time window conform to the "voltage-reactive power" positive correlation response law of the physical impedance load. Otherwise, an invalid positive correlation indicator bit is generated as an invalid value (0), indicating that the corresponding feeder exhibits power control characteristics or dead-zone characteristics.

[0036] The electromagnetic energy storage elements within a physical load prevent abrupt changes in its state, resulting in a smooth and continuous response trajectory over time. In contrast, digital control equipment is affected by maximum power point tracking (MPPT) step size, control dead zone determination, and discretization adjustments, leading to high-frequency jitter or step noise in its response trajectory at the microscopic time scale. Therefore, time-difference operations are introduced to analyze the continuity of the response by examining the statistical characteristics of the fitted deviation sequence.

[0037] In this embodiment of the invention, the fitted ordinate values ​​of each point in the standardized fluctuation scatter plot are calculated using the optimal baseline straight line model. The difference between the actual ordinate value and the fitted ordinate value of each point is calculated to obtain a fitting deviation sequence. This sequence reflects the degree to which each scatter plot deviates from the dominant response trend and characterizes the microscopic differences in the device response.

[0038] Furthermore, the first-order difference between adjacent data points in the fitted deviation sequence of the standardized fluctuation scatter plot is calculated, and the sum of the squares of all difference values ​​is averaged as the difference fluctuation. Specifically, the mean value is calculated after squaring each difference value to capture the abrupt change energy between adjacent sampling points, reflecting the severity of the abrupt change in the response trajectory. The step noise of the digital control device will significantly increase this value.

[0039] Finally, the differential volatility is combined with a preset sensitivity coefficient and then inversely mapped to obtain the response continuity index. In this embodiment of the invention, the preset sensitivity coefficient can be set to 20, the purpose of which is to amplify the influence of differential volatility on the response continuity index and improve the sensitivity to digital control noise. The implementer can adjust this as needed. As an example, the expression for the response continuity index is: In the formula, This is represented as a response continuity indicator. This is represented as the preset sensitivity coefficient. This is expressed as differential volatility. When the response trajectory is smooth and continuous, the smaller the differential volatility, the larger the response continuity index, indicating the presence of physical inertia. However, when digital control noise is present, the differential volatility increases, and the response continuity index decreases.

[0040] To further improve the reliability of identification, the fluctuation characteristics of periodic active power sequences are introduced to assist in the judgment. The active power of physical loads usually remains relatively stable within a short time window, while photovoltaic power generation is affected by environmental factors such as cloud cover and airflow disturbances, and its active power output has significant random fluctuations.

[0041] In this embodiment of the invention, the ratio of the standard deviation to the mean of the periodic active power sequence within a preset time window is calculated as the active power volatility, reflecting the environmental randomness of periodic active power. The greater the volatility, the more likely it is to include distributed power sources such as photovoltaic power generation. An environmental suppression coefficient is obtained through the active power volatility; that is, the greater the active power volatility, the smaller the environmental suppression coefficient, thereby reducing the interference of highly volatile power sources on impedance characteristic weights. The environmental suppression coefficient is negatively correlated with the active power volatility. In this embodiment of the invention, the active power volatility can be obtained by inverse proportional mapping, which will not be elaborated here.

[0042] Finally, by combining the positive correlation indicator, the response continuity index, and the environmental suppression coefficient, the impedance characteristic weight is obtained. In a specific embodiment of the present invention, the product of the positive correlation indicator, the response continuity index, and the environmental suppression coefficient is normalized and used as the impedance characteristic weight. This weight represents the degree to which the operating characteristics of the corresponding feeder at the current moment conform to the actual physical impedance load. The larger the impedance characteristic weight, the closer the operating characteristics of the corresponding feeder are to the actual physical impedance load, and the more significant the benefits of subsequent load reduction operations.

[0043] It should be noted that normalization is a well-known technique in the art. In this case, the normalization process can be performed using the standardization normalization method. In other embodiments of the present invention, the hyperbolic tangent function, etc., can also be used, which will not be elaborated here.

[0044] This completes the mechanistic identification of the feeder load attributes, providing core confidence parameters for subsequent calculation of effective load reduction value.

[0045] S3: Correct the periodic active power sequence using impedance characteristic weights to obtain active power gain value, and introduce a voltage risk penalty term based on the current voltage level to generate effective load reduction value; generate an evaluation control strategy based on the effective load reduction value.

[0046] Operational assessments can be used to ultimately guide emergency control. In scenarios of "high source-load offsetting," the apparent power at the cutoff point can be deceptive, and direct disconnection may lead to the erroneous operation of "power supply cut-off." Based on the principle of risk assessment, the net value of a disconnection operation should equal the frequency improvement benefit from removing the effective load, minus the voltage support loss cost caused by removing the power supply component. Therefore, in order to transform the abstract source-load attributes into specific power operation values, it is necessary to correct the active power gain through impedance characteristic weighting and introduce voltage risk penalties to achieve accurate quantification of the load shedding value.

[0047] In this embodiment of the invention, the current value of the periodic active power sequence is multiplied by the impedance characteristic weight to obtain the active power gain value. This quantifies the actual impact of disconnecting the corresponding feeder on the system. The impedance characteristic weight quantifies the probability that the feeder is a real physical load. Multiplying the two effectively removes the offsetting effect of distributed power generation output, truly reflecting the actual reduction in effective load after disconnecting the corresponding feeder, and quantifying the positive contribution of the disconnection operation to system frequency recovery. A larger impedance characteristic weight indicates that the feeder is more likely to be a purely physical load, and the active power gain value is closer to the original value of the periodic active power sequence. A smaller impedance characteristic weight indicates that the feeder is more likely to contain a large amount of power source components, and the active power gain value is smaller. Therefore, a larger active power gain value indicates a more significant contribution of disconnecting the line to restoring the system's active power balance.

[0048] Furthermore, a voltage urgency function is constructed based on the degree of voltage drop relative to the low-voltage safety threshold, and the penalty intensity is dynamically adjusted according to the current voltage level. A voltage risk penalty term is obtained by combining the periodic active power sequence, impedance characteristic weights, and the voltage urgency function, representing the potential negative consequences of a cutoff operation.

[0049] In this embodiment of the invention, the per-unit value of the feeder voltage at the current moment is obtained. The per-unit voltage value is the ratio of the current periodic voltage sequence value to the system rated voltage, representing the degree to which the current voltage deviates from the rated level. For example, when the rated voltage is 10kV and the current voltage is 9.5kV, the per-unit voltage value is 0.95. A voltage urgency function is constructed: when the per-unit voltage value is greater than or equal to the set low-voltage safety threshold, it indicates that the system voltage is within the normal range, the risk of power disconnection is low, and the penalty term can be set to zero, so the function value is zero. In this embodiment of the invention, the low-voltage safety threshold is set to 0.90pu, which can be adjusted by the implementer and will not be elaborated here.

[0050] When the per-unit voltage value is less than the set low-voltage safety threshold, the risk of disconnecting inverters with potential reactive power support capabilities increases linearly and sharply as the voltage decreases. Therefore, the deviation between the set low-voltage safety threshold and the per-unit voltage value is calculated, and the product of the deviation and a preset penalty gain coefficient is used as a function value. The larger the deviation, the larger the voltage urgency function value, indicating a stronger penalty. In this embodiment of the invention, the preset penalty gain coefficient can be set to 5, representing the sensitivity of the disconnection operation to voltage stability. The implementer can adjust it as needed, and there are no restrictions here.

[0051] Furthermore, based on the impedance characteristic weight, the non-impedance proportion is obtained, quantifying the proportion of non-physical load components such as distributed power sources in the feeder. The larger the impedance characteristic weight, the fewer the non-impedance components. The impedance characteristic weight and the non-impedance proportion are negatively correlated. In this embodiment of the invention, the difference between the value 1 and the impedance characteristic weight is used as the non-impedance proportion.

[0052] Finally, by combining the current periodic active power sequence value, the non-impedance ratio, and the voltage urgency function value, a voltage risk penalty term is obtained. In one specific embodiment of the invention, the product of the current periodic active power sequence value, the non-impedance ratio, and the voltage urgency function value is used as the voltage risk penalty term. A larger periodic active power sequence value indicates a higher corresponding load contribution; a higher non-impedance ratio indicates a greater power source component; and a higher voltage urgency function indicates a greater deviation of the voltage from the safety threshold, resulting in a greater voltage risk during the disconnection operation. Therefore, a higher penalty term value indicates a greater likelihood of voltage support loss due to disconnecting lines with high power source components.

[0053] The difference between the active power gain and the voltage risk penalty is used as the effective load reduction value, which comprehensively reflects the net system benefits that can be obtained by disconnecting the corresponding feeder. An evaluation control strategy can be generated based on this effective load reduction value.

[0054] In this embodiment of the invention, the operating status of feeders is evaluated and classified based on their effective load reduction value. By assessing the positive and negative characteristics of the effective load reduction value, feeders are divided into two categories, reflecting their suitability for participating in emergency load reduction. The classification results directly correspond to the control logic, accurately reflecting the differences in the impact of different feeders on the system's emergency control.

[0055] If the effective load reduction value is negative or zero, it indicates that the risk of disconnecting the corresponding line outweighs the benefit. Such lines are usually high photovoltaic penetration lines or lines in low voltage areas. The feeder should be marked as locked and protected, and disconnection of the corresponding feeder should be prohibited to avoid the risk of accidental power disconnection or exacerbating voltage drop in scenarios with high source-load offsetting.

[0056] If the effective load shedding value is positive, it means that the benefits of cutting the corresponding line outweigh the risks. The larger the value, the higher the overall contribution of cutting the line to system stability. Such lines are usually traditional load lines with low photovoltaic penetration. The feeders are then sorted from largest to smallest effective load shedding value to obtain a cut-off sequence. When the low-frequency load shedding device operates, the feeders at the top of the sequence are cut off first to restore frequency at the lowest cost.

[0057] Based on the assessment and classification results, disconnection commands are generated in the order of the disconnection sequence. For example, if the effective load reduction values ​​of three feeders in a substation are 1.2MW, 0.8MW, and -0.3MW respectively, then the feeder with an effective load reduction value of 1.2MW is listed as the first feeder in the disconnection sequence, the feeder with 0.8MW as the second feeder, and the feeder with -0.3MW as a blocked protection state. When the system triggers low-frequency load reduction, the disconnection command for the first feeder is executed first. If the frequency does not recover to a safe level, the disconnection command for the second feeder is then executed. Feeders in the blocked protection state are never disconnected, achieving more precise and safer emergency control.

[0058] In summary, this invention utilizes random sampling consensus regression analysis to extract dominant voltage-reactive power sensitivity, effectively identifying positively damped physical loads that follow Ohm's law. Furthermore, by analyzing the differential fluctuation characteristics of the fitted deviation sequence, it captures high-frequency step noise in the digital control of power electronic equipment, further distinguishing between physical inertial components and analog control power supplies. Simultaneously, it introduces a risk penalty term based on voltage urgency to generate effective load shedding value, transforming the decision-making basis for disconnection operations into a quantitative indicator that comprehensively considers frequency gains and voltage risks. This effectively enhances the ability to prevent local voltage collapse while restoring the active power balance of the entire network. By constructing a decoupled identification model based on the voltage-reactive power micro-response mechanism, this invention achieves accurate assessment of feeder source-load attributes, improves the robustness of identification, and enhances the safety and accuracy of emergency control in the power system.

[0059] It should be noted that the order of the above embodiments of the present invention is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0060] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

Claims

1. A method for assessing the operational status of a high-proportion renewable energy power system, characterized in that, The method includes: The electrical quantities collected at the power system feeder junctions are processed to obtain periodic voltage sequences, periodic reactive power sequences, and periodic active power sequences. The quiescent state is determined based on the noise level of the periodic voltage sequences. In the non-quiescent state, the periodic voltage sequences and periodic reactive power sequences are standardized to construct a standardized fluctuation scatter set. A random sampling consistency regression analysis against outliers is performed on the standardized fluctuation scatter set to extract the dominant voltage-reactive power sensitivity and generate a positive correlation indicator based on the sensitivity polarity. The fitting deviation sequence of the standardized fluctuation scatter set relative to the regression benchmark is calculated, and a response continuity index is generated based on the statistical characteristics of the fitting deviation sequence. The impedance characteristic weight is generated by integrating the positive correlation indicator, the response continuity index, and the fluctuation characteristics of the periodic active power sequence. The periodic active power sequence is corrected using impedance characteristic weights to obtain active power gain value, and a voltage risk penalty term based on the current voltage level is introduced to generate effective load shedding value; an evaluation control strategy is generated based on the effective load shedding value. The method for obtaining the effective load reduction value includes: The active power gain value is obtained by multiplying the value of the periodic active power sequence at the current moment by the impedance characteristic weight. A voltage urgency function is constructed based on the degree of drop of the current voltage level relative to the low voltage safety threshold; a voltage risk penalty term is obtained by combining the periodic active power sequence, impedance characteristic weights, and the voltage urgency function. The difference between the active power gain and the voltage risk penalty is used as the effective load reduction value. The method for obtaining the voltage risk penalty item includes: Obtain the current feeder voltage per unit value; construct a voltage urgency function: when the voltage per unit value is greater than or equal to the set low voltage safety threshold, the function value is zero; when the voltage per unit value is less than the set low voltage safety threshold, calculate the deviation between the set low voltage safety threshold and the voltage per unit value, and use the product of the deviation value and the preset penalty gain coefficient as the function value; The proportion of non-impedance components is obtained based on impedance characteristic weights; the impedance characteristic weights are negatively correlated with the proportion of non-impedance components. By combining the current periodic active power sequence value, the non-impedance ratio, and the voltage urgency function value, a voltage risk penalty term is obtained.

2. The method for assessing the operating status of a high-proportion renewable energy power system according to claim 1, characterized in that, The method for determining the silent state includes: Calculate the standard deviation of the periodic voltage sequence within a preset time window. If the standard deviation is less than the preset noise floor threshold, it is determined to be in a silent state, and the impedance characteristic weight of the previous moment is reused. If the standard deviation is greater than or equal to the preset noise floor threshold, it is determined to be in a non-silent state.

3. The method for assessing the operating status of a high-proportion renewable energy power system according to claim 1, characterized in that, The method for obtaining the positive correlation indicator bit includes: The standardized fluctuation scatter set is iteratively fitted using a random sampling consensus algorithm, and the linear model with the most inliers is selected as the best baseline straight line model. The slope of the optimal reference straight line model is extracted as the dominant voltage-reactive power sensitivity; when the dominant voltage-reactive power sensitivity is greater than the preset judgment threshold, the positive correlation indicator bit is generated as a valid value; otherwise, the positive correlation indicator bit is generated as an invalid value.

4. The method for assessing the operating status of a high-proportion renewable energy power system according to claim 3, characterized in that, The method for obtaining the fitting deviation sequence includes: The fitted ordinate values ​​of each point in the standardized fluctuation scatter plot are calculated using the optimal baseline linear model. The difference between the actual ordinate value and the fitted ordinate value of each point is then calculated to obtain the fitting deviation sequence.

5. The method for assessing the operating status of a high-proportion renewable energy power system according to claim 1, characterized in that, The method for obtaining the response continuity index includes: Calculate the first-order difference between adjacent data points in the fitted deviation sequence of the standardized volatility scatter plot, and average the sum of squares of all difference values ​​to obtain the difference volatility. Then, combine the difference volatility with a preset sensitivity coefficient and perform an inverse proportional mapping to obtain the response continuity index.

6. The method for assessing the operating status of a high-proportion renewable energy power system according to claim 1, characterized in that, The method for obtaining the impedance characteristic weight includes: The ratio of the standard deviation to the mean of the periodic active power sequence within a preset time window is calculated as the active power volatility. The environmental suppression coefficient is obtained through the active power volatility, and the environmental suppression coefficient is negatively correlated with the active power volatility. By combining the positive correlation indicator, the response continuity index, and the environmental suppression coefficient, the impedance characteristic weights are obtained.

7. The method for assessing the operating status of a high-proportion renewable energy power system according to claim 1, characterized in that, The control strategy based on the effective load reduction value generation assessment includes: The feeder operating status is evaluated and classified based on the effective load reduction value; if the effective load reduction value is negative or zero, the feeder is marked as blocked protection; if the effective load reduction value is positive, the feeders are sorted from largest to smallest effective load reduction value to obtain the cut-off sequence; according to the evaluation and classification results, the cut-off command is generated in the order of the cut-off sequence.

8. The method for assessing the operating status of a high-proportion renewable energy power system according to claim 1, characterized in that, The method for obtaining the standardized fluctuation scatter set includes: In the non-quiet state, the standard deviations of the periodic voltage sequence and the periodic reactive power sequence are calculated respectively; the corresponding standard deviations are used to normalize the periodic voltage sequence and the periodic reactive power sequence to generate a standardized scatter set of dimensionless coordinate points.

Images