Power distribution network scheduling method considering power regulation capability of photovoltaic power station cluster

Abstract

The invention relates to the technical field of smart power grids, and provides a power distribution network scheduling method considering photovoltaic power station cluster power regulation capability, comprising: collecting each group of monitoring data, the monitoring data comprising photovoltaic power station cluster operation data and power distribution network operation data to obtain a proportionality coefficient and associated feature similarity of an initial regulation margin, obtaining first and second association influence weights; acquiring a real adjustment margin, and acquiring a predicted value and a contradictory time period of the photovoltaic output value; and according to the operation data of the power distribution network, constructing a scheduling priority, and determining a distribution scheduling scheme and a hierarchical scheduling instruction in combination with the predicted value of the photovoltaic output value, the contradictory time period and the maximum reactive power. The accuracy and the stability of power distribution network dispatching can be improved.

Description

Technical Field

[0001] This application relates to the field of smart grid technology, specifically to a distribution network dispatching method that considers the power regulation capability of photovoltaic power plant clusters. Background Technology

[0002] Photovoltaic (PV) power output in power systems is characterized by randomness, intermittency, and fluctuation. To meet various grid dispatching needs, multiple distributed PV power stations are treated as a whole, and flexible power adjustment is achieved through coordinated control—this is known as PV power station cluster power regulation. Utilizing the power regulation capabilities of PV power station clusters to achieve distribution network dispatching can ensure grid security, improve renewable energy absorption, reduce the distribution network's dependence on traditional peak-shaving power sources, optimize operational efficiency, and achieve a balanced and stable energy allocation.

[0003] During the distribution network dispatching process, the lack of sufficient consideration of the power regulation capabilities of photovoltaic power plant clusters can easily lead to a disconnect between the distribution network dispatching plan and the actual operating status of the distribution network. Specifically: a sudden increase in photovoltaic output can easily cause overload of distribution network lines and voltage rise at nodes, but the dispatching system cannot fully utilize the reactive power regulation capabilities of the clusters to suppress voltage exceedances; a sudden drop in photovoltaic output can easily lead to a power deficit in the distribution network, and the distribution network dispatching system can only supplement energy through the upstream grid or energy storage devices. The dispatching system ignores the active power reserve regulation capabilities of the clusters, which increases power supply costs and reduces power supply reliability. Summary of the Invention

[0004] This application provides a distribution network dispatching method that considers the power regulation capability of photovoltaic power plant clusters. This addresses the problem that distribution network dispatching does not fully consider the power regulation capability of photovoltaic power plant clusters, leading to significant deviations between distribution network dispatching and actual conditions, making it difficult to accurately regulate the power system's state. The specific technical solution adopted is as follows:

[0005] One embodiment of this application provides a distribution network dispatching method that considers the power regulation capability of photovoltaic power plant clusters. The method includes the following steps:

[0006] Collect monitoring data for different natural days, and each set of monitoring data includes photovoltaic power plant cluster operation data and distribution network operation data;

[0007] Calculate the initial regulation margin and maximum reactive power of each photovoltaic inverter in the photovoltaic power plant cluster. Combine the photovoltaic power plant cluster operation data to perform correlation feature analysis, obtain the proportional coefficient of the initial regulation margin of the photovoltaic power plant cluster and the correlation feature similarity of each set of monitoring data. Cluster the correlation feature similarity and determine the target cluster. Based on the target cluster, obtain the first correlation influence weight of temperature on the initial regulation margin and the second correlation influence weight of irradiance on the initial regulation margin.

[0008] Based on the photovoltaic power station cluster operation data, the predicted values ​​of all total load, photovoltaic output, irradiance, and temperature for the next natural day corresponding to the monitoring data are obtained. Combining the relationship between temperature and irradiance and the proportional coefficient of the initial regulation margin, the predicted value of the proportional coefficient of the initial regulation margin of the photovoltaic power station cluster for the next natural day corresponding to the monitoring data is calculated. Combining the first correlation influence weight and the second correlation influence weight, the actual regulation margin of the monitoring data is obtained. Based on the actual regulation margin of the monitoring data and the predicted value of photovoltaic output, the adjustable range of photovoltaic output is determined, and the predicted value of photovoltaic output is corrected and the conflicting periods are identified.

[0009] Based on the distribution network operation data, a scheduling priority is constructed. Combined with the predicted value of photovoltaic power output, conflict periods, and maximum reactive power, the allocation scheduling scheme and hierarchical scheduling instructions are determined.

[0010] Furthermore, the method for determining the initial regulation margin and maximum reactive power of the photovoltaic inverter is as follows:

[0011] The difference between the maximum active power of all photovoltaic inverters in the photovoltaic power plant cluster and the active power of the photovoltaic inverter itself is recorded as the initial adjustment margin of the photovoltaic inverter.

[0012] The arithmetic square root of the difference between the squares of the apparent power and the active power of a photovoltaic inverter is denoted as the maximum reactive power of the photovoltaic inverter.

[0013] Furthermore, the method for determining the proportional coefficient of the initial adjustment margin of the photovoltaic power station cluster is as follows:

[0014] The ratio of the sum of the initial regulation margins of all photovoltaic inverters in the photovoltaic power plant cluster to the rated regulation margin of the photovoltaic power plant cluster is denoted as the proportional coefficient of the initial regulation margin of the photovoltaic power plant cluster.

[0015] Furthermore, the specific method for obtaining the similarity of the associated features of the monitored data is as follows:

[0016] The photovoltaic power plant cluster operation data includes the temperature and irradiance of the photovoltaic inverters. Based on the proportional coefficient of all initial regulation margins, the sum of the temperatures of all photovoltaic inverters, and the sum of the irradiance of all photovoltaic inverters within the same set of monitoring data of the photovoltaic power plant cluster, regulation margin characteristic sequences, temperature characteristic sequences, and irradiance characteristic sequences of the same set of monitoring data are constructed respectively.

[0017] Based on the correlation coefficients between the adjustment margin feature sequence and the temperature feature sequence, and between the adjustment margin feature sequence and the irradiance feature sequence of the same set of monitoring data, the feature coordinates of the same set of monitoring data are established.

[0018] Any set of monitoring data is denoted as the target group monitoring data, and the mean of the Euclidean distances between the target group monitoring data and all other groups of monitoring data is denoted as the first feature difference of the target group monitoring data.

[0019] The maximum DTW distance between the target group monitoring data and the monitoring data of the same type from all other groups is denoted as the second characteristic distance of the target group monitoring data. The ratio of the second characteristic distance of the target group monitoring data to the maximum second characteristic distance of all other groups monitoring data is denoted as the second characteristic difference of the target group monitoring data.

[0020] The positive correlation between the first feature difference and the second feature difference of the target group monitoring data is denoted as the correlation feature similarity of the target group monitoring data.

[0021] Furthermore, the method for determining the target cluster is as follows:

[0022] Cluster the monitoring data of all groups by association feature similarity, obtain clusters, and select the cluster with the most association feature similarity as the target cluster.

[0023] Furthermore, the specific steps included in the first correlation influence weight of temperature on the initial conditioning margin and the second correlation influence weight of irradiance on the initial conditioning margin are as follows:

[0024] The absolute value of the mean of the correlation coefficients between temperature and initial adjustment margin of all different groups of monitoring data in the target cluster is denoted as the first correlation of the target cluster, and the absolute value of the mean of the correlation coefficients between irradiance and initial adjustment margin of all different groups of monitoring data in the target cluster is denoted as the second correlation of the target cluster.

[0025] The ratio of the first correlation of the target cluster to the sum of the first and second correlations is denoted as the weight of the first correlation influence of temperature on the initial adjustment margin.

[0026] The ratio of the second correlation of the target cluster to the sum of the first and second correlations is denoted as the second correlation influence weight of irradiance on the initial adjustment margin.

[0027] Furthermore, the specific calculation method for the predicted value of the proportional coefficient of the initial adjustment margin of the photovoltaic power station cluster for the next natural day following the natural day corresponding to the monitoring data is as follows:

[0028] Based on the proportional coefficients of temperature, irradiance, and the initial regulation margin of the photovoltaic power station cluster, the relationship between temperature and irradiance and the proportional coefficients of the initial regulation margin is obtained and denoted as the first curve. Substituting the predicted values ​​of all irradiance and temperature for the next natural day corresponding to the monitoring data into the first curve, the predicted value of the proportional coefficients of the initial regulation margin of the photovoltaic power station cluster for the next natural day corresponding to the monitoring data is obtained.

[0029] Furthermore, the method for obtaining the true adjustment margin of the monitoring data is as follows:

[0030] The weight of the first correlation between temperature determined by the monitoring data and the initial regulation margin is used as the weight of the proportional coefficient of the initial regulation margin of the photovoltaic power station cluster. The weight of the second correlation between irradiance determined by the monitoring data and the initial regulation margin is used as the weight of the predicted value of the proportional coefficient of the initial regulation margin of the photovoltaic power station cluster on the next natural day of the natural day corresponding to the monitoring data. The weighted sum is recorded as the proportional coefficient of the regulation margin of the monitoring data.

[0031] The product of the scaling factor of the monitoring data's regulation margin and the rated regulation margin of the photovoltaic power station cluster is denoted as the actual regulation margin of the monitoring data.

[0032] Furthermore, the specific methods for determining the adjustable range of photovoltaic output based on the actual adjustment margin of monitoring data and the predicted value of photovoltaic output, and for correcting the predicted value of photovoltaic output, include:

[0033] The natural days corresponding to the monitoring data The adjustable range of photovoltaic output value is ,in, Indicates the corresponding calendar day The predicted value of photovoltaic output based on monitoring data. Indicates the corresponding calendar day The true adjustment margin of the monitoring data;

[0034] Less than The monitoring data corresponds to the natural day The predicted value of the photovoltaic output is assigned as: ; will be greater than The monitoring data corresponds to the natural day The predicted value of the photovoltaic output is assigned as: .

[0035] Furthermore, the specific methods for determining the contradictory time periods include:

[0036] For the next natural day following the natural day corresponding to the monitoring data, the continuous period of time in which the photovoltaic output value continues to decrease while the total load continues to increase is recorded as the contradictory period of load peak - photovoltaic output low peak; the continuous period of time in which the photovoltaic output value continues to increase while the total load continues to decrease is recorded as the contradictory period of load trough - photovoltaic output peak.

[0037] The beneficial effects of this application are:

[0038] This application first calculates the initial regulation margin and maximum reactive power of each photovoltaic inverter within a photovoltaic power plant cluster based on the cluster's operational data. Then, it analyzes the correlation between active and reactive power output under varying irradiance and temperature conditions, obtaining the correlation feature similarity for each set of monitoring data. The greater the correlation feature similarity, the greater the difference in the impact of environmental changes on the photovoltaic power plant cluster's regulation capability. To further accurately analyze the correlation features of environmental changes on the photovoltaic power plant cluster's regulation capability, a correlation feature analysis is performed on all sets of monitoring data corresponding to the correlation feature similarity in the target cluster. This avoids the influence of random interference during monitoring on the correlation feature analysis, obtaining the first and second correlation influence weights. Based on the regulation capability of the photovoltaic power plant cluster under irradiance and temperature... This process involves correlation analysis to obtain the true regulation margin of monitoring data, accurately defining the boundary of the power regulation potential of photovoltaic power plant clusters, avoiding the problem of dispatch instructions exceeding actual capacity and causing execution failures, identifying conflicting periods, and planning power allocation strategies for the cluster grid in advance for conflicting periods on the next day after the identified day to avoid power surplus or shortage in the distribution network. Finally, based on the distribution network operation data, a dispatch priority is constructed, and combined with the predicted value of photovoltaic output, conflicting periods, and maximum reactive power, an allocation dispatch scheme and hierarchical dispatch instructions are determined to achieve distribution network dispatch based on the power regulation capacity of photovoltaic power plant clusters. This addresses the problem that distribution network dispatching does not fully consider the power regulation capacity of photovoltaic power plant clusters, resulting in significant deviations between distribution network dispatching and actual conditions, making it difficult to accurately regulate the state of the power system. Attached Figure Description

[0039] To more clearly illustrate the technical solutions in the embodiments of this application 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 this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0040] Figure 1A schematic flowchart of a distribution network dispatching method considering the power regulation capability of a photovoltaic power plant cluster is provided as an embodiment of this application;

[0041] Figure 2 This is a flowchart illustrating the initial adjustment margin acquisition process provided in one embodiment of this application. Detailed Implementation

[0042] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0043] Please see Figure 1 The diagram illustrates a distribution network dispatching method considering the power regulation capability of photovoltaic power plant clusters, according to an embodiment of this application. The method includes the following steps:

[0044] Step S001: Collect monitoring data for different natural days. Each set of monitoring data includes photovoltaic power station cluster operation data and distribution network operation data.

[0045] The system collects operational data from photovoltaic power plant clusters and distribution networks. The photovoltaic power plant cluster operational data includes the active power, reactive power, power factor, maximum regulation capacity, irradiance, and temperature of each photovoltaic inverter within the cluster, the total load and photovoltaic output of the cluster, and the voltage, current, and frequency at the cluster's grid connection point. The maximum regulation capacity includes the upper limit of active power reserve capacity and reactive power compensation. The distribution network operational data includes the active power, reactive power, and line loss rate of each line in the distribution network, as well as the voltage at each node and the real-time power of adjustable loads.

[0046] Specifically, irradiance sensors, temperature sensors, power sensors, voltage sensors, and current sensors are installed on the photovoltaic power plant cluster side to collect photovoltaic power plant cluster operation data; feeder terminal units, distribution terminal units, and smart meters are installed on the distribution network side to collect distribution network operation data.

[0047] The photovoltaic power plant cluster operation data and distribution network operation data are transmitted to edge computing nodes. The edge computing nodes preprocess the photovoltaic power plant cluster operation data and distribution network operation data to reduce the impact of noise interference on data quality. The edge computing nodes transmit the preprocessed photovoltaic power plant cluster operation data and distribution network operation data to the regional dispatch gateway via wireless transmission network or fiber optic network. The regional dispatch gateway encrypts the transmission using SSL or TLS encryption protocols and then aggregates and transmits the data received by the regional dispatch gateway to the distribution network dispatch center.

[0048] Data cleaning and standardization are performed on the photovoltaic power station cluster operation data and distribution network operation data transmitted to the distribution network dispatch center. Specifically, 3D processing is used. Outliers are removed according to the criteria, and missing data is filled in by linear interpolation. In this embodiment, the maximum-minimum normalization algorithm is selected to normalize the data of the same type. The photovoltaic power station cluster operation data and the distribution network operation data are classified and stored in the corresponding data table according to the data type, and timestamps and equipment numbers are added to facilitate the distribution network dispatch center to trace the data source.

[0049] All photovoltaic power station cluster operation data and distribution network operation data collected within a natural day are taken as a group of monitoring data, and a first preset threshold group of monitoring data is collected. In this embodiment, the value of the first preset threshold is 100.

[0050] At this point, monitoring data from different groups has been obtained.

[0051] Step S002: Calculate the initial regulation margin and maximum reactive power of each photovoltaic inverter in the photovoltaic power station cluster. Combine the photovoltaic power station cluster operation data to perform correlation feature analysis, obtain the proportional coefficient of the initial regulation margin of the photovoltaic power station cluster and the correlation feature similarity of each set of monitoring data. Cluster the correlation feature similarity and determine the target cluster. Based on the target cluster, obtain the first correlation influence weight of temperature on the initial regulation margin and the second correlation influence weight of irradiance on the initial regulation margin.

[0052] Based on the correlation between the active and reactive power regulation capabilities of photovoltaic power plant clusters in various regions and meteorological factors, as well as the correlation between the status changes of various nodes and lines in the distribution network and the regulation capabilities of photovoltaic power plant clusters, a comprehensive analysis of the distribution network dispatch area and photovoltaic regulation efficiency is conducted to accurately analyze the impact of the power regulation capabilities of photovoltaic power plant clusters on distribution network dispatch.

[0053] First, the regulation capability of the photovoltaic power plant cluster is analyzed based on the operation data of the photovoltaic power plant cluster.

[0054] The difference between the maximum active power of all photovoltaic inverters in the photovoltaic power plant cluster and the active power of the photovoltaic inverter itself is denoted as the initial regulation margin of the photovoltaic inverter. The arithmetic square root of the difference between the squares of the apparent power and the active power of the photovoltaic inverter is denoted as the maximum reactive power of the photovoltaic inverter.

[0055] The flowchart for obtaining the initial adjustment margin is as follows: Figure 2As shown; a positive initial regulation margin of the photovoltaic inverter indicates that the photovoltaic power station cluster can increase active power output to achieve the dispatch and regulation of the distribution network; a negative initial regulation margin of the photovoltaic inverter indicates that the photovoltaic power station cluster can reduce active power output to suppress power excess; the apparent power of the photovoltaic inverter is the rated parameter of the photovoltaic inverter itself, which can be read from the photovoltaic inverter's instruction manual; the maximum reactive power of the photovoltaic inverter can determine the reactive power compensation range provided by the photovoltaic power station cluster.

[0056] Furthermore, the correlation between active and reactive power output under changes in irradiance and temperature is analyzed separately.

[0057] The ratio of the sum of the initial regulation margins of all photovoltaic inverters in the photovoltaic power plant cluster to the rated regulation margin of the photovoltaic power plant cluster is denoted as the proportional coefficient of the initial regulation margin of the photovoltaic power plant cluster. The proportional coefficients of all initial regulation margins within the same set of monitoring data of the photovoltaic power plant cluster are arranged in ascending order according to their corresponding time periods to obtain the regulation margin characteristic sequence of the same set of monitoring data. Similarly, the sum of the temperatures of all photovoltaic inverters within the same set of monitoring data of the photovoltaic power plant cluster is arranged in ascending order according to their corresponding time periods to obtain the temperature characteristic sequence of the same set of monitoring data. Finally, the sum of the irradiance of all photovoltaic inverters within the same set of monitoring data of the photovoltaic power plant cluster is arranged in ascending order according to their corresponding time periods to obtain the irradiance characteristic sequence of the same set of monitoring data.

[0058] The rated regulation margin of a photovoltaic power station cluster is the difference between the rated maximum active power output of the photovoltaic power station cluster and the active power output of the photovoltaic power station cluster under standard test conditions. The standard test conditions are an ideal environment with an irradiance of 1000W / ㎡, a module temperature of 25℃, and an atmospheric quality of AM1.5. The rated regulation margin of the photovoltaic power station cluster is obtained by experimental measurement by those skilled in the art.

[0059] Correlation feature analysis was performed on the adjustment margin feature sequences, temperature feature sequences, and irradiance feature sequences of different sets of monitoring data measured under different environmental conditions.

[0060] The correlation coefficient between the adjustment margin feature sequence and the temperature feature sequence of the same set of monitoring data is used as the x-axis value, and the correlation coefficient between the adjustment margin feature sequence and the irradiance feature sequence of the same set of monitoring data is used as the y-axis value to obtain the feature coordinates of the same set of monitoring data. Any set of monitoring data is denoted as the target set of monitoring data. The mean of the Euclidean distances between the target set of monitoring data and all other sets of monitoring data is denoted as the first feature difference of the target set of monitoring data. The maximum DTW distance between the target set of monitoring data and all other sets of monitoring data of the same type is denoted as the second feature distance of the target set of monitoring data. The ratio of the second feature distance of the target set of monitoring data to the maximum second feature distance of all other sets of monitoring data is denoted as the second feature difference of the target set of monitoring data. The positive correlation result between the first feature difference and the second feature difference of the target set of monitoring data is denoted as the association feature similarity of the target set of monitoring data.

[0061] This embodiment uses the Pearson correlation coefficient to calculate the correlation coefficient between the sequences. The calculation of the correlation coefficient between sequences is a well-known technique and will not be described in detail here.

[0062] It is understood that a positive correlation processing is applied to the first and second feature differences of the target group monitoring data, ensuring that the first and second feature differences of the target group monitoring data are positively correlated with the similarity of the associated features of the target group monitoring data. It is understood that the positive correlation in this application refers to the relationship between the independent and dependent variables, where the independent variables are the first and second feature differences of the target group monitoring data, and the dependent variable is the similarity of the associated features of the target group monitoring data. A positive correlation means that the dependent variable increases (decreases) as the independent variable increases (decreases), and can be an additive or multiplicative relationship.

[0063] Preferably, as an embodiment of this application, the product of the first feature difference and the second feature difference of the target group monitoring data is denoted as the correlation feature similarity of the target group monitoring data.

[0064] The same method can be used to obtain the correlation feature similarity of any set of monitoring data.

[0065] The greater the similarity of the correlation characteristics among the monitoring data of each group, the greater the difference in the impact of environmental changes on the regulation capability of photovoltaic power plant clusters.

[0066] Cluster the correlation feature similarity of the monitoring data of all groups to obtain clusters, and select the cluster with the most correlation feature similarity as the target cluster.

[0067] In this embodiment, the AP clustering algorithm is used to cluster the similarity of associated features. In practical applications, in addition to achieving the purpose of clustering, implementers may use other existing methods such as DBSCAN clustering algorithm, Mean Shift clustering algorithm, OPTICS clustering algorithm, Gaussian Mixture Models clustering algorithm, and Spectral Clustering for clustering. This application does not impose any special restrictions.

[0068] To further and more accurately analyze the correlation characteristics between environmental changes and the regulation capacity of photovoltaic power plant clusters, we conducted a correlation characteristic analysis on the monitoring data of all groups corresponding to the similarity of correlation characteristics in the target cluster, avoiding the impact of random interference differences during the monitoring process on the correlation characteristic analysis.

[0069] The absolute value of the mean correlation coefficient between temperature and initial regulation margin of all different groups of monitoring data in the target cluster is denoted as the first correlation of the target cluster. The absolute value of the mean correlation coefficient between irradiance and initial regulation margin of all different groups of monitoring data in the target cluster is denoted as the second correlation of the target cluster. The ratio of the first correlation of the target cluster to the sum of the first and second correlations is denoted as the first association influence weight of temperature on initial regulation margin. The ratio of the second correlation of the target cluster to the sum of the first and second correlations is denoted as the second association influence weight of irradiance on initial regulation margin.

[0070] This embodiment uses the Pearson correlation coefficient to calculate the correlation coefficient between temperature and initial conditioning margin, as well as the correlation coefficient between irradiance and initial conditioning margin. The calculation of the Pearson correlation coefficient is a well-known technique and will not be described in detail here.

[0071] Thus, the first correlation weight of temperature on the initial conditioning margin and the second correlation weight of irradiance on the initial conditioning margin are obtained.

[0072] Step S003: Based on the photovoltaic power station cluster operation data, obtain the predicted values ​​of all total load, photovoltaic output, irradiance, and temperature for the next natural day following the natural day corresponding to the monitoring data. Combining the relationship between temperature and irradiance and the proportional coefficient of the initial adjustment margin, calculate the predicted value of the proportional coefficient of the initial adjustment margin of the photovoltaic power station cluster for the next natural day following the natural day corresponding to the monitoring data. Combining the first correlation influence weight and the second correlation influence weight, obtain the actual adjustment margin of the monitoring data. Based on the actual adjustment margin of the monitoring data and the predicted value of the photovoltaic output, determine the adjustable range of the photovoltaic output, and correct the predicted value of the photovoltaic output and determine the conflicting periods.

[0073] An LSTM (Long Short-Term Memory) network model is trained using the total load and photovoltaic (PV) output values ​​of a photovoltaic (PV) power plant cluster from its operational data. By inputting all total load and PV output values ​​from a set of monitoring data into the LSTM network model, predicted values ​​for all total load and PV output values ​​for the next day can be obtained. Similarly, an LSTM network model is trained using the irradiance and temperature data of the PV power plant cluster from its operational data. By inputting all irradiance and temperature data from a set of monitoring data into the LSTM network model, predicted values ​​for all irradiance and temperature values ​​for the next day can be obtained.

[0074] The loss function of the LSTM (Long Short-Term Memory) network model is the cross-entropy loss function, and the optimizer is the Adam optimizer. The training process of the LSTM network model is a well-known technique and will not be described in detail here.

[0075] Using temperature and irradiance as independent variables and the proportional coefficient of the initial regulation margin of the photovoltaic power station cluster as the dependent variable, the least squares method is used for curve fitting. Based on the temperature, irradiance and the corresponding proportional coefficient of the initial regulation margin in all monitoring data of the target cluster, the relationship between temperature and irradiance and the proportional coefficient of the initial regulation margin is obtained and recorded as the first curve. The predicted values ​​of all irradiance and temperature for the next natural day corresponding to the monitoring data are substituted into the first curve to obtain the predicted value of the proportional coefficient of the initial regulation margin of the photovoltaic power station cluster for the next natural day corresponding to the monitoring data.

[0076] The least squares method for curve fitting is a well-known technique and will not be elaborated further.

[0077] The weighted average of the temperature (determined from monitoring data) and the initial regulation margin is used as the weight of the proportional coefficient of the initial regulation margin of the photovoltaic power plant cluster. The weighted average of the irradiance (determined from monitoring data) and the initial regulation margin is used as the weight of the predicted value of the proportional coefficient of the initial regulation margin of the photovoltaic power plant cluster for the next natural day following the day corresponding to the monitoring data. The proportional coefficient of the regulation margin of the monitoring data is then multiplied by the rated regulation margin of the photovoltaic power plant cluster, and this product is recorded as the actual regulation margin of the monitoring data.

[0078] Based on the actual adjustment margin of the monitoring data and the predicted value of the photovoltaic (PV) output, the adjustable range of the PV output is determined. Then, the predicted PV output for the corresponding calendar day is corrected according to this adjustable range, reducing the error in judging the power regulation capability of the PV power plant cluster due to environmental changes. Specifically: the calendar day corresponding to the monitoring data... The adjustable range of photovoltaic output value is ,in, Indicates the corresponding calendar day The predicted value of photovoltaic output based on monitoring data. Indicates the corresponding calendar day The true adjustment margin of the monitoring data; when the monitoring data corresponds to a natural day Predicted value of photovoltaic power output Less than At that time, the natural day corresponding to the monitoring data will be... The predicted value of the photovoltaic output is assigned as: When the monitoring data corresponds to the natural day Predicted value of photovoltaic power output Greater than At that time, the natural day corresponding to the monitoring data will be... The predicted value of the photovoltaic output is assigned as: .

[0079] It is important to understand that the true adjustment margin of the monitoring data is obtained by correlating and correcting the adjustment capability of the photovoltaic power plant cluster with irradiance and temperature. This can accurately define the boundary of the power adjustment potential of the photovoltaic power plant cluster and avoid the problem of scheduling instructions exceeding the actual capacity, which leads to execution failure.

[0080] For the next natural day following the natural day corresponding to the monitoring data, the continuous period of time in which the photovoltaic output value continues to decrease while the total load continues to increase is recorded as the contradictory period of load peak - photovoltaic output low peak, and the continuous period of time in which the photovoltaic output value continues to increase while the total load continues to decrease is recorded as the contradictory period of load trough - photovoltaic output peak.

[0081] Understandably, during conflicting periods on the next day following the identified day, it is necessary to plan the power allocation strategy for the cluster power grid in advance to avoid power surplus or shortage in the distribution network.

[0082] Thus, the contradictory time period of the next natural day corresponding to the monitored data was identified.

[0083] Step S004: Based on the distribution network operation data, construct the scheduling priority, and combine the predicted value of photovoltaic power output, conflict periods and maximum reactive power to determine the allocation scheduling scheme and hierarchical scheduling instructions.

[0084] Based on the power of each line and the voltage of each node in the distribution network operation data, the Newton-Raphson method is used to solve the power flow equations, calculate the power loss, voltage deviation and line load rate of each line, and then analyze the power flow distribution of the photovoltaic power station cluster grid connection point and surrounding nodes.

[0085] The specific rules are as follows: If the voltage deviation at the grid connection point exceeds ±5%, or the load rate of the line exceeds 80%, the area where the node is located will be marked as a "area to be dispatched and intervened"; Based on the sensitivity of the impact of photovoltaic power output changes on power flow, the contribution of photovoltaic power plant cluster power regulation to power flow optimization will be determined, the node voltage change corresponding to a 1MW change in photovoltaic power output will be calculated, and the ratio of the voltage change to the rated voltage will be used as the contribution of photovoltaic power plant cluster power regulation to power flow optimization.

[0086] Understandably, the greater the contribution, the more significant the impact of the power regulation of the photovoltaic power plant cluster on the scheduling and allocation. Areas with higher contribution should prioritize the use of photovoltaic power plant cluster regulation; while areas with low contribution need to link energy storage, capacitors and other equipment to improve scheduling efficiency.

[0087] The PV curve method is used to analyze the voltage stability margin of the distribution network. Specifically, the voltage stability limit of the photovoltaic power station cluster grid connection point is determined, active power is gradually injected into the photovoltaic power station cluster grid connection point, and the node voltage change is recorded. When the voltage drops, the corresponding power value is recorded as the stability limit. If the photovoltaic output value is greater than 90% of the stability limit, it is determined that "high voltage stability risk" has occurred.

[0088] Based on the reactive power regulation capability of photovoltaic power plant clusters, the ability of reactive power compensation of photovoltaic power plant clusters to support voltage stability is analyzed. Specifically, the voltage change when the maximum reactive power of the photovoltaic inverter changes by 1 Mvar is calculated. If the voltage change is less than 0.5 kV, a scheduling plan needs to be formulated in conjunction with other voltage regulating equipment.

[0089] Establish scheduling priorities, which include three levels: Level 1, Level 2, and Level 3.

[0090] Specifically, photovoltaic power plant clusters with a contribution rate greater than 1% and a line load rate greater than 80% are designated as first-priority clusters. A contribution rate greater than 1% indicates that the photovoltaic power plant cluster has sufficient adjustment margin. First-priority clusters are areas with voltage exceeding limits or line overload, which need to be addressed first through power regulation of the photovoltaic power plant cluster, and the clusters are prioritized to reduce active power output.

[0091] Photovoltaic power plant clusters with photovoltaic output values ​​less than 90% of the voltage stability limit are designated as secondary priority areas. These secondary priority areas represent high-risk areas for voltage stability. If a photovoltaic power plant cluster lacks sufficient power regulation capacity to meet voltage boost requirements, the cluster will be scheduled to increase reactive power output, or other equipment will be coordinated for scheduling. Specifically, the voltage boost is calculated when the maximum reactive power change of the photovoltaic inverter is 1 Mvar. If the voltage boost is greater than or equal to 0.5 kV, the secondary priority area can be adjusted through the photovoltaic power plant cluster. If the voltage boost is less than 0.5 kV, other voltage regulation equipment needs to be coordinated to adjust the secondary priority area.

[0092] Photovoltaic power plant clusters with voltage deviations at each node of the distribution network within the allowable range of ±5%, line load rates less than or equal to 80%, and photovoltaic output values ​​less than 85% of the voltage stability limit are designated as Level 3 priority. Level 3 priority refers to the economic optimization area with no safety operation risks. Under the premise of safe operation of the distribution network, photovoltaic power plant clusters are dispatched to increase active power output, reduce traditional power generation, and lower power supply costs.

[0093] Tasks are prioritized and scheduled according to a specific plan:

[0094] 1. For first-priority scenarios, if a voltage exceeding the limit occurs, the reactive power regulation capability of the photovoltaic power station cluster is invoked first. The required reactive power compensation to resolve the voltage exceedance is calculated based on the voltage change when the maximum reactive power of the photovoltaic inverter changes by 1 Mvar. If the reactive power compensation is less than or equal to the maximum reactive power output, a photovoltaic reactive power compensation command is sent to adjust the reactive power output to compensate for the corresponding amount. If the reactive power compensation exceeds the photovoltaic regulation capability, the parallel capacitor bank is linked for reactive power compensation to reduce the voltage to the allowable range. For first-priority scenarios, if a voltage exceeding the limit occurs, the future output trend is first determined based on the photovoltaic output forecast. If the predicted output is increasing, compensation can be implemented based on the natural upward trend of photovoltaic output. If the predicted output is stable or decreasing, the photovoltaic power station cluster is dispatched to increase photovoltaic output while simultaneously performing corresponding reactive power compensation to stabilize and improve the voltage. For the first priority level, if the line load rate is greater than 80%, and if it is in a contradictory period of peak load and low photovoltaic output, the photovoltaic system will be dispatched to increase active power output, coordinate energy storage discharge, and reduce interruptible loads to reduce the line load rate to below 75%. If it is in a contradictory period of low load and peak photovoltaic output, a photovoltaic output reduction command will be sent to reduce the line load rate to below 75%.

[0095] Voltage exceeding limits refers to the actual operating voltage at a distribution network node exceeding the specified allowable deviation range. This is specifically divided into two categories: exceeding the limit by a factor of two and exceeding the limit by a factor of two by a factor of two by a factor of two by a factor of two by a factor of two. Exceeding the limit by a factor of two means the actual voltage is higher than the upper limit of the rated voltage (e.g., the actual voltage of a 10kV distribution network is greater than 10.5kV). Exceeding the limit by a factor of two means the actual voltage is lower than the lower limit of the rated voltage (e.g., the actual voltage of a 10kV distribution network is less than 9.5kV). Voltage stability corresponds to a certain range. If the voltage exceeds the upper limit, it indicates exceeding the limit by a factor of two by a factor of two; if the voltage is lower than the lower limit, it indicates exceeding the limit by a factor of two by a factor of two by a factor of two by a factor of two by a factor of three by a factor of two by a factor of two. Specifically, the determination of exceeding limits is based on specific distribution network parameters and is a known technique. For example, the stability range for medium-voltage distribution networks is ±5%, the single-phase range for 220V in low-voltage distribution networks is -10% ± 7%, and the three-phase stability range for 380V is ± 7%.

[0096] 2. For secondary priority, based on the voltage stability limit of the photovoltaic power station cluster grid connection point, calculate the safe output threshold, which is 90% of the voltage stability limit. Combine the predicted value of photovoltaic output for scheduling. Specifically: if the predicted value of photovoltaic output is greater than or equal to the safe threshold, reserve a 20% reactive power adjustment margin for the photovoltaic power station cluster in advance. At the same time, calculate the reactive power compensation required to reduce the voltage to the target value. If the reactive power adjustment margin can meet the requirements, issue a pre-adjustment command in advance. If the voltage change is less than 0.5kV, coordinate voltage regulation with the parallel capacitor bank and the on-load tap-changing transformer.

[0097] 3. For the third-priority system, during periods of conflict between low load and peak photovoltaic output, the system schedules photovoltaic power plant clusters to maximize active power output to reduce the amount of electricity purchased from the upper-level grid. Simultaneously, surplus output is used to charge energy storage devices, and the stored energy is then discharged during subsequent peak load periods. For the third-priority system, during periods of conflict between peak load and low photovoltaic output, the system maintains basic active power output and reserves sufficient active power regulation margin to cope with sudden load increases.

[0098] The above-mentioned allocation and scheduling scheme is simulated on the distribution network digital twin platform. The photovoltaic power plant cluster power dispatch and adjustment schemes determined under different priorities are input into the platform. The platform simulates the power flow changes, voltage fluctuations, and line load rate changes of the distribution network within the next hour, and the simulation results are obtained. If the voltage of all nodes drops to the allowable range and the line load rate is below 80% in the simulation results, the allocation and scheduling scheme is deemed feasible. If voltage overruns or voltage exceeding limits still exist in the simulation results, the photovoltaic power plant cluster power dispatch and adjustment scheme is adjusted, and the simulation is repeated until a feasible result is obtained.

[0099] After determining whether the allocation and scheduling scheme is feasible, hierarchical scheduling instructions are executed according to the scheme. Specifically:

[0100] 1. For the first-priority area, power regulation commands are sent to the photovoltaic power station cluster control system to specify the active or reactive power regulation amount and regulation duration. At the same time, commands are sent to the switching equipment of the distribution network. If the photovoltaic power station cluster regulation cannot completely solve the line overload problem, some non-critical load lines can be disconnected through the switching equipment for coordinated control.

[0101] 2. For secondary priority areas, if the reactive power regulation capacity of the photovoltaic power station cluster is insufficient, the dispatch center will compensate for reactive power by linking parallel capacitor banks and coordinate voltage regulation with the photovoltaic power station cluster.

[0102] 3. For the third-priority areas, dispatch instructions are sent through the distribution network dispatch center to increase the photovoltaic output of the photovoltaic power station clusters and adjust it to the maximum adjustable margin at the current time, while reducing the amount of electricity purchased from the upper-level grid.

[0103] After executing the hierarchical dispatching instructions, the dispatching results are monitored and adjusted in real time. Specifically, the distribution network dispatching center compares and analyzes the real-time received monitoring data, comparing the deviation between the target value of the dispatching instructions and the actual executed value. If the deviation is less than or equal to 5%, the current instructions are maintained; if the deviation is greater than 5%, the adjustment amount is recalculated and a correction instruction is sent. Simultaneously, during the execution of the hierarchical dispatching instructions, a simulation analysis of the dispatching scheme is performed every hour, and the dispatching strategy is dynamically optimized based on the real-time operating status of the distribution network to ensure the safe and stable operation of the distribution network.

[0104] The deviation value is defined as the absolute value of the difference between the target value and the actual executed value, and the ratio of the absolute value to the target value is used as the deviation value.

[0105] Thus, based on the power regulation capability of the photovoltaic power plant cluster, the power distribution network can be dispatched.

[0106] The above description is only a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the principles of this application should be included within the protection scope of this application.

Claims

1. A distribution network dispatching method considering the power regulation capability of photovoltaic power plant clusters, characterized in that, The method includes the following steps: Collect monitoring data for different natural days, and each set of monitoring data includes photovoltaic power plant cluster operation data and distribution network operation data; Calculate the initial regulation margin and maximum reactive power of each photovoltaic inverter in the photovoltaic power plant cluster. Combine the photovoltaic power plant cluster operation data to perform correlation feature analysis, obtain the proportional coefficient of the initial regulation margin of the photovoltaic power plant cluster and the correlation feature similarity of each set of monitoring data. Cluster the correlation feature similarity and determine the target cluster. Based on the target cluster, obtain the first correlation influence weight of temperature on the initial regulation margin and the second correlation influence weight of irradiance on the initial regulation margin. Based on the photovoltaic power station cluster operation data, the predicted values ​​of all total load, photovoltaic output, irradiance, and temperature for the next natural day corresponding to the monitoring data are obtained. Combining the relationship between temperature and irradiance and the proportional coefficient of the initial regulation margin, the predicted value of the proportional coefficient of the initial regulation margin of the photovoltaic power station cluster for the next natural day corresponding to the monitoring data is calculated. Combining the first correlation influence weight and the second correlation influence weight, the actual regulation margin of the monitoring data is obtained. Based on the actual regulation margin of the monitoring data and the predicted value of photovoltaic output, the adjustable range of photovoltaic output is determined, and the predicted value of photovoltaic output is corrected and the conflicting periods are identified. Based on the distribution network operation data, a scheduling priority is constructed. Combined with the predicted value of photovoltaic power output, conflict periods, and maximum reactive power, the allocation scheduling scheme and hierarchical scheduling instructions are determined.

2. The distribution network dispatching method considering the power regulation capability of photovoltaic power plant clusters according to claim 1, characterized in that, The method for determining the initial regulation margin and maximum reactive power of the photovoltaic inverter is as follows: The difference between the maximum active power of all photovoltaic inverters in the photovoltaic power plant cluster and the active power of the photovoltaic inverter itself is recorded as the initial adjustment margin of the photovoltaic inverter. The arithmetic square root of the difference between the squares of the apparent power and the active power of a photovoltaic inverter is denoted as the maximum reactive power of the photovoltaic inverter.

3. The distribution network dispatching method considering the power regulation capability of photovoltaic power plant clusters according to claim 1, characterized in that, The method for determining the proportional coefficient of the initial adjustment margin of the photovoltaic power station cluster is as follows: The ratio of the sum of the initial regulation margins of all photovoltaic inverters in the photovoltaic power plant cluster to the rated regulation margin of the photovoltaic power plant cluster is denoted as the proportional coefficient of the initial regulation margin of the photovoltaic power plant cluster.

4. A distribution network dispatching method considering the power regulation capability of photovoltaic power plant clusters according to claim 1, characterized in that, The specific method for obtaining the similarity of the associated features of the monitored data is as follows: The photovoltaic power plant cluster operation data includes the temperature and irradiance of the photovoltaic inverters. Based on the proportional coefficient of all initial regulation margins, the sum of the temperatures of all photovoltaic inverters, and the sum of the irradiance of all photovoltaic inverters within the same set of monitoring data of the photovoltaic power plant cluster, regulation margin characteristic sequences, temperature characteristic sequences, and irradiance characteristic sequences of the same set of monitoring data are constructed respectively. Based on the correlation coefficients between the adjustment margin feature sequence and the temperature feature sequence, and between the adjustment margin feature sequence and the irradiance feature sequence of the same set of monitoring data, the feature coordinates of the same set of monitoring data are established. Any set of monitoring data is denoted as the target group monitoring data, and the mean of the Euclidean distances between the target group monitoring data and all other groups of monitoring data is denoted as the first feature difference of the target group monitoring data. The maximum DTW distance between the target group monitoring data and the monitoring data of the same type from all other groups is denoted as the second characteristic distance of the target group monitoring data. The ratio of the second characteristic distance of the target group monitoring data to the maximum second characteristic distance of all other groups monitoring data is denoted as the second characteristic difference of the target group monitoring data. The positive correlation between the first feature difference and the second feature difference of the target group monitoring data is denoted as the correlation feature similarity of the target group monitoring data.

5. A distribution network dispatching method considering the power regulation capability of photovoltaic power plant clusters according to claim 1, characterized in that, The method for determining the target cluster is as follows: Cluster the monitoring data of all groups by association feature similarity, obtain clusters, and select the cluster with the most association feature similarity as the target cluster.

6. A distribution network dispatching method considering the power regulation capability of photovoltaic power plant clusters according to claim 1, characterized in that, The specific steps included in the weighting of the first correlation between temperature and the initial conditioning margin, and the weighting of the second correlation between irradiance and the initial conditioning margin, are as follows: The absolute value of the mean of the correlation coefficients between temperature and initial adjustment margin of all different groups of monitoring data in the target cluster is denoted as the first correlation of the target cluster, and the absolute value of the mean of the correlation coefficients between irradiance and initial adjustment margin of all different groups of monitoring data in the target cluster is denoted as the second correlation of the target cluster. The ratio of the first correlation of the target cluster to the sum of the first and second correlations is denoted as the weight of the first correlation influence of temperature on the initial adjustment margin. The ratio of the second correlation of the target cluster to the sum of the first and second correlations is denoted as the second correlation influence weight of irradiance on the initial adjustment margin.

7. A distribution network dispatching method considering the power regulation capability of photovoltaic power plant clusters according to claim 1, characterized in that, The specific calculation method for the predicted value of the proportional coefficient of the initial adjustment margin of the photovoltaic power station cluster for the next natural day of the day corresponding to the monitoring data is as follows: Based on the proportional coefficients of temperature, irradiance, and the initial regulation margin of the photovoltaic power station cluster, the relationship between temperature and irradiance and the proportional coefficients of the initial regulation margin is obtained and denoted as the first curve. Substituting the predicted values ​​of all irradiance and temperature for the next natural day corresponding to the monitoring data into the first curve, the predicted value of the proportional coefficients of the initial regulation margin of the photovoltaic power station cluster for the next natural day corresponding to the monitoring data is obtained.

8. A distribution network dispatching method considering the power regulation capability of photovoltaic power plant clusters according to claim 1, characterized in that, The method for obtaining the true adjustment margin of the monitoring data is as follows: The weight of the first correlation between temperature determined by the monitoring data and the initial regulation margin is used as the weight of the proportional coefficient of the initial regulation margin of the photovoltaic power station cluster. The weight of the second correlation between irradiance determined by the monitoring data and the initial regulation margin is used as the weight of the predicted value of the proportional coefficient of the initial regulation margin of the photovoltaic power station cluster on the next natural day of the natural day corresponding to the monitoring data. The weighted sum is recorded as the proportional coefficient of the regulation margin of the monitoring data. The product of the scaling factor of the monitoring data's regulation margin and the rated regulation margin of the photovoltaic power station cluster is denoted as the actual regulation margin of the monitoring data.

9. A distribution network dispatching method considering the power regulation capability of photovoltaic power plant clusters according to claim 1, characterized in that, The specific methods for determining the adjustable range of photovoltaic output based on the actual adjustment margin of monitoring data and the predicted value of photovoltaic output, and for correcting the predicted value of photovoltaic output, include: The natural days corresponding to the monitoring data The adjustable range of photovoltaic output value is ,in, Indicates the corresponding calendar day The predicted value of photovoltaic output based on monitoring data. Indicates the corresponding calendar day The true adjustment margin of the monitoring data; Less than The monitoring data corresponds to the natural day The predicted value of the photovoltaic output is assigned as: ; will be greater than The monitoring data corresponds to the natural day The predicted value of the photovoltaic output is assigned as: .

10. A distribution network dispatching method considering the power regulation capability of photovoltaic power plant clusters according to claim 1, characterized in that, The specific methods for determining the contradictory time periods are as follows: For the next natural day following the natural day corresponding to the monitoring data, the continuous period of time in which the photovoltaic output value continues to decrease while the total load continues to increase is recorded as the contradictory period of load peak - photovoltaic output low peak; the continuous period of time in which the photovoltaic output value continues to increase while the total load continues to decrease is recorded as the contradictory period of load trough - photovoltaic output peak.

Images