A power distribution network operation characteristic influence analysis method and related device

Abstract

The application is suitable for the technical field of power distribution network, and provides a power distribution network operation characteristic influence analysis method and related device, which comprises the following steps: obtaining time series data, performing quasi-steady operation segment interception on the time series data based on an electrical state change rate threshold and an adaptive time length threshold, and obtaining at least one effective analysis segment; constructing a power distribution network source-load coupling sensitivity matrix, calculating the marginal contribution degree of each distributed power grid connection point to the feeder operation characteristic abnormal index, and locating the target distributed power source cluster causing the operation characteristic abnormality; performing time stamp alignment and spatial topology association on the first sampling frequency data corresponding to the target distributed power source cluster on the transformer area side and the second sampling frequency data on the feeder side, restoring the disturbance propagation path of the operation characteristic abnormality on the power distribution network feeder, and generating the operation characteristic influence analysis result of the power distribution network; and the accurate analysis of the influence of the power distribution network operation characteristic under the condition of high penetration rate of distributed power is realized.

Description

Technical Field

[0001] This application relates to the field of power distribution network technology, and in particular to a method and related apparatus for analyzing the impact of power distribution network operation characteristics. Background Technology

[0002] Currently, the analysis of the impact of distributed generation access on the operation characteristics of distribution networks mainly relies on two types of technical methods: one is the offline analysis method based on simulation calculation, which uses power system simulation software to build a distribution network topology model, sets the output curve and load curve of distributed generation, and obtains the voltage and power distribution of each node through power flow calculation, thereby analyzing the impact of different access capacities and access locations on operation characteristics; the other is the online analysis method based on monitoring data, which uses distribution automation terminals and smart meters deployed in the distribution network to collect electrical quantity data such as voltage, current, and power, and compares the collected data with the operating limits in fixed time segments to determine whether there are abnormal operating conditions such as voltage exceeding limits, line overload or overload at each node, and generates corresponding alarm information accordingly.

[0003] High-penetration distributed power generation output exhibits drastic, second-level fluctuations due to weather conditions. Existing analysis methods based on fixed time sections cannot effectively distinguish between the true quasi-steady-state process and brief transient disturbances during power grid operation. This leads to a large amount of photovoltaic power fluctuation data being mistakenly included as steady-state data in the analysis, affecting voltage sensitivity modeling. The input data for subsequent analyses such as load-bearing capacity assessment also contains significant noise interference, ultimately resulting in a large deviation between the operational characteristic analysis conclusions and the actual physical state of the distribution network, leading to insufficient accuracy in the operational characteristic analysis results. Summary of the Invention

[0004] This application provides a method and related apparatus for analyzing the impact of distribution network operation characteristics, which is used to solve the problem that there is a large deviation between the conclusions of operation characteristic analysis and the actual physical state of the distribution network, and the accuracy of the operation characteristic analysis results is insufficient.

[0005] The first aspect of this application provides a method for analyzing the impact of distribution network operation characteristics, including: Acquire time-series data of distribution network feeder output voltage, reverse active power, grid connection point voltage of each distributed power source, and output of each distributed power source; Based on a preset electrical state change rate threshold and an adaptive duration threshold adjusted by the output fluctuation rate of distributed power sources, the time series data is truncated into quasi-steady-state operation segments to obtain at least one effective analysis segment. For the effective analysis segment, a distribution network source-load coupling sensitivity matrix is ​​constructed. Based on the sensitivity matrix, the marginal contribution of each distributed power source grid connection point to the abnormal feeder operation characteristics index is calculated. Based on the marginal contribution, the target distributed power source cluster causing the abnormal operation characteristics is located. The first sampling frequency data on the transformer side and the second sampling frequency data on the feeder side corresponding to the target distributed power cluster are timestamped and spatially topologically associated to reconstruct the disturbance propagation path of the abnormal operating characteristics from the target distributed power cluster to other nodes on the distribution network feeder. Based on the aforementioned quasi-steady-state operation segment, target distributed power source cluster, and disturbance propagation path, the analysis results of the impact on the operation characteristics of the distribution network are generated.

[0006] Furthermore, the step of extracting quasi-steady-state operation segments from the time-series data based on a preset electrical state change rate threshold and an adaptive duration threshold adjusted by the output fluctuation rate of distributed power sources includes: A multi-dimensional electrical state vector characterizing the operating state of the distribution network at each sampling moment is constructed from the time-series data of the feeder outlet voltage, reverse active power, and voltage at each distributed power source grid connection point. Calculate the change amplitude of the multidimensional electrical state vector at adjacent sampling times, and use it as the electrical state vector change rate; When the rate of change of the electrical state vector is detected to be less than the preset threshold for the rate of change of the electrical state, the current sampling time is recorded as the candidate start time, and the rate of change of the electrical state vector at subsequent sampling times is continuously monitored. When the rate of change of the electrical state vector at each adjacent sampling time is less than the electrical state change rate threshold within a duration starting from the candidate start time, and the duration reaches the adaptive duration threshold adjusted by the output fluctuation rate of the distributed power source, the time series data from the candidate start time to the current time is extracted as a valid analysis segment.

[0007] Furthermore, the adaptive duration threshold is determined in the following way: Calculate the standard deviation of the total active power output of each of the distributed power generation grid connection points during the current analysis period, and use it as a quantitative value of the output fluctuation of the distributed power generation. Based on the quantized value, an adaptive duration threshold is determined through a preset negative correlation function relationship; wherein, the larger the quantized value, the smaller the determined adaptive duration threshold.

[0008] Furthermore, during continuous monitoring, if the rate of change of the electrical state vector at any sampling moment is detected to be greater than the electrical state change rate threshold before the duration reaches the adaptive duration threshold, the monitoring record starting from that candidate start moment is abandoned, and continuous monitoring is re-executed with the sampling moment where the rate of change exceeds the limit as the new monitoring start moment. When a sampling moment where the rate of change of the electrical state vector is subsequently detected to be less than the electrical state change rate threshold, the sampling moment where the rate of change exceeds the limit is taken as the new candidate start moment.

[0009] Furthermore, for the effective analysis segment, a distribution network source-load coupling sensitivity matrix is ​​constructed, and based on the sensitivity matrix, the marginal contribution of each distributed power source connection point to the abnormal indicators of feeder operation characteristics is calculated, including: Based on the distribution network topology and impedance parameters corresponding to the effective analysis segment, a distribution network source-load coupling sensitivity matrix is ​​constructed to characterize the relationship between the active power output changes at each distributed power source grid connection point and the voltage changes at each key node. In response to the presence of abnormal feeder operating characteristics indicators within the effective analysis segment, each of the distributed power generation grid connection points is divided into multiple participant subsets, and the change in the abnormal operating characteristics indicators caused by the combined action of each participant subset in the distribution network is determined. Based on the changes in indicators of each subset of participants, the marginal contribution of each distributed power grid connection point to the abnormal operating characteristic indicators is calculated using the allocation rules of cooperative game theory.

[0010] Furthermore, the step of locating the target distributed power cluster causing abnormal operating characteristics based on the marginal contribution includes: From all the distributed power grid connection points, select those distributed power grid connection points with positive marginal contribution values ​​and positive contribution to the abnormal operating characteristics. The marginal contribution of each selected distributed power grid connection point is sorted in descending order of value; The marginal contribution values ​​are accumulated sequentially according to the sorting order. When the accumulated value reaches or exceeds the preset cumulative contribution threshold, all distributed power grid connection points participating in the accumulation are identified as the target distributed power cluster.

[0011] Furthermore, the step of aligning the first sampling frequency data on the transformer substation side and the second sampling frequency data on the feeder side corresponding to the target distributed power cluster with timestamps and associating them with spatial topology includes: Acquire first sampling frequency data collected by the intelligent measurement terminal on the target distributed power cluster area side, and second sampling frequency data collected by the feeder terminal unit on the distribution network feeder, wherein the sampling frequency corresponding to the first sampling frequency data is higher than the sampling frequency corresponding to the second sampling frequency data. Using the sampling time of the second sampling frequency data as a reference, interpolation calculation is performed on the first sampling frequency data that falls within a preset time window to obtain equivalent data that is aligned with the sampling time of the second sampling frequency data; Based on the topology information of the power distribution network feeder, the aligned first sampling frequency data is associated and bound with the second sampling frequency data at the corresponding spatial location.

[0012] Furthermore, the restoration of the disturbance propagation path from the target distributed power cluster to other nodes on the distribution network feeder, as described above, includes: From the equivalent data after timestamp alignment and spatial topology association, extract the voltage drop start time of each grid-connected point in the target distributed power cluster during the abnormal occurrence period; Starting from each grid connection point in the target distributed power cluster, the order of the voltage drop start times of adjacent monitoring nodes is compared step by step along the topology information towards the feeder outlet to determine the disturbance propagation direction. Based on the disturbance propagation direction between adjacent nodes, a directed topology graph for disturbance propagation is constructed, starting from the target distributed power cluster and ending at the feeder outlet node of the distribution network feeder, as the disturbance propagation path.

[0013] Furthermore, the analysis results of the impact on the operating characteristics of the distribution network, which are generated by integrating the quasi-steady-state operation segment, the target distributed power generation cluster, and the disturbance propagation path, include: Steady-state characteristic parameters are extracted from the effective analysis segments in quasi-steady-state operation; the identifiers of each power source and the marginal contribution of each power source are extracted from the target distributed power source cluster; and the disturbance node sequence and time series features are extracted from the disturbance propagation path. The extracted steady-state characteristic parameters, the identifiers of each power source and the marginal contribution of that power source, the perturbation node sequence and time series characteristics are assembled according to preset association rules to generate a multi-dimensional impact analysis report that includes steady-state analysis, anomaly tracing and propagation process.

[0014] The second aspect of this application provides a device for analyzing the impact of distribution network operation characteristics, comprising: The time-series data acquisition unit is used to acquire time-series data of distribution network feeder output voltage, reverse active power, grid connection point voltage of each distributed power source, and output of each distributed power source. The effective analysis segment determination unit is used to extract quasi-steady-state operation segments from the time series data based on a preset electrical state change rate threshold and an adaptive duration threshold adjusted by the output fluctuation rate of the distributed power source, so as to obtain at least one effective analysis segment. The target distributed power generation cluster location unit is used to construct a distribution network source-load coupling sensitivity matrix for the effective analysis segment, calculate the marginal contribution of each distributed power generation grid connection point to the abnormal feeder operation characteristics index based on the sensitivity matrix, and locate the target distributed power generation cluster that causes the abnormal operation characteristics based on the marginal contribution. The disturbance propagation path restoration unit is used to perform time stamp alignment and spatial topology association on the first sampling frequency data of the transformer side and the second sampling frequency data of the feeder side corresponding to the target distributed power cluster, and restore the disturbance propagation path of the abnormal operation characteristics from the target distributed power cluster to other nodes on the distribution network feeder. The unit for generating operational characteristic impact analysis results is used to integrate the quasi-steady-state operation segment, the target distributed power generation cluster, and the disturbance propagation path to generate operational characteristic impact analysis results for the distribution network.

[0015] As can be seen from the above technical solutions, this application has the following advantages: This application provides a multi-dimensional data foundation for operational characteristic analysis by acquiring time-series data on the distribution network feeder outlet voltage, reverse active power, voltage at each distributed power source grid connection point, and output of each distributed power source. Based on a preset electrical state change rate threshold and an adaptive duration threshold adjusted by the output fluctuation rate of distributed power sources, quasi-steady-state operation segments are extracted from the time-series data, effectively eliminating transient disturbance data caused by severe fluctuations in photovoltaic power, ensuring that subsequent analysis is based on a real electrical quasi-steady-state process, and significantly improving the accuracy of operational characteristic analysis. By constructing a distribution network source-load coupling sensitivity matrix and calculating the marginal contribution of each distributed power source grid connection point, it is possible to accurately locate the source-load coupling sensitivity matrix in complex active distribution networks with multi-machine coupling and interactive effects. The target distributed power generation cluster, where abnormal operating characteristics play a dominant role, solves the problem of being unable to isolate single-point disturbance sources from group effects. By aligning the timestamps and spatial topology of high-frequency sampling data from the transformer substation side and low-frequency sampling data from the feeder side, and reconstructing the disturbance propagation path based on temporal causality, the project achieves visualized tracing of the propagation process of abnormal operating characteristics from the disturbance source to other nodes, providing dispatchers with a complete panoramic view of the anomaly evolution. Finally, by integrating quasi-steady-state operating segments, the target distributed power generation cluster, and the disturbance propagation path, a multi-dimensional impact analysis result is generated. This enables accurate analysis of the impact of high-penetration distributed power generation access on distribution network operating characteristics, providing reliable technical support for distribution network dispatching, operation, and control strategy formulation. Attached Figure Description

[0016] Figure 1This is a flowchart illustrating a method for analyzing the impact of distribution network operation characteristics in this application; Figure 2 This is a flowchart illustrating the process of extracting valid analysis segments in this application; Figure 3 This is a schematic diagram illustrating the process of calculating the marginal contribution and locating the target distributed power cluster in this application. Figure 4 This is a schematic diagram illustrating the process of aligning the timestamps of the first and second sampling frequency data with spatial topology in this application; Figure 5 This is a schematic diagram of the process for reconstructing the disturbance propagation path in this application. Detailed Implementation

[0017] The terms “first,” “second,” “third,” “fourth,” etc. (if present) in the specification and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a particular order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms “comprising” and “corresponding to,” and any variations thereof, are intended to cover a non-exclusive inclusion, e.g., a process; method; system; product; or device that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to these processes; methods; products; or devices.

[0018] Example 1 The implementation method in this embodiment can be implemented in a system, on a server, or on a terminal; no specific limitation is made. The method in this application will be described below from the perspective of system implementation. Please refer to... Figure 1 The method for analyzing the impact of distribution network operation characteristics provided in this application includes the following steps: S1. Obtain the output voltage of the distribution network feeder; reverse active power; voltage at the grid connection point of each distributed power source and time-series data of the output of each distributed power source; Specifically, the voltage timing data of the feeder outlet is acquired through a voltage transformer installed on the outgoing side of the feeder circuit breaker in the substation's distribution automation system. This voltage transformer collects the effective value of the three-phase voltage at the feeder outlet at a fixed sampling period and converts it into per-unit values ​​for storage and transmission according to actual operational requirements. The reference value for the per-unit value is the rated voltage of the feeder. The reverse active power timing data of the feeder is acquired through a power direction measurement unit installed at the same feeder outlet. This measurement unit uses a combination of current transformers and voltage transformers to measure the magnitude and direction of the active power flowing through the feeder outlet. When the power flow direction is from the feeder to the substation busbar, it is determined to be reverse active power. This reverse active power value is then normalized, with the normalization reference value being the rated transmission capacity of the feeder. This eliminates dimensional differences caused by different feeder capacity levels, making the data between different feeders comparable. The grid-connected voltage time-series data is collected in real-time by grid-connected monitoring terminals installed at each distributed power source connection point, which then convert the effective voltage value at the grid-connected point into per-unit value form. The base value of the per-unit value is the rated voltage of that grid-connected point. The distributed power source output time-series data is collected synchronously by the terminal from the active power output values ​​of each distributed power source. Both types of data are collected from the same source and both carry a unified satellite synchronous clock time tag. The collected data are aggregated to the distribution network operation data analysis platform through the distribution communication network. All of the above time-series data carry a unified satellite synchronous clock time tag, and each data sequence maintains synchronous sampling alignment in the time dimension, together forming the multi-dimensional data foundation for subsequent quasi-steady-state operation segment extraction and analysis.

[0019] S2. Based on the preset electrical state change rate threshold and the adaptive duration threshold adjusted by the output fluctuation rate of distributed power sources, the time series data is truncated into quasi-steady-state operation segments to obtain at least one effective analysis segment; In distribution networks with high penetration of distributed power sources, the output of renewable energy sources such as distributed photovoltaics is affected by meteorological factors such as cloud movement and local shading, resulting in drastic, second-level fluctuations in active power output. This causes frequent switching between steady-state and transient electrical quantities such as voltage and power flow in the distribution network. Quasi-steady-state operation refers to a distribution network where the variation amplitude of various electrical state quantities remains within a small range over a continuous period of time, and the system as a whole is in a relatively stable operating state, which is different from transient disturbances caused by brief fluctuations in renewable energy output or instantaneous load switching. If time-series data containing transient disturbances is directly used for subsequent voltage sensitivity modeling, load-bearing capacity assessment, and attribution analysis, the accuracy of the analysis results will be severely reduced. This approach employs a dual-threshold judgment logic, combining an electrical state change rate threshold and an adaptive duration threshold, to scan and filter time-series data at each sampling moment. Transient disturbance data segments caused by brief fluctuations in renewable energy output are eliminated, retaining only quasi-steady-state operation segments where the electrical state change amplitude remains consistently within a small range and the duration meets the adaptive duration requirement. This provides a clean and reliable data foundation for subsequent sensitivity matrix construction and disturbance propagation path reconstruction. Please refer to [link to relevant documentation]. Figure 2 Specifically, it includes the following: S21. Construct a multi-dimensional electrical state vector characterizing the operating state of the distribution network at each sampling time using time-series data of feeder output voltage, reverse active power, and voltage at each distributed power source grid connection point; For any sampling time, the per-unit value of the feeder outlet voltage, the normalized value of the feeder reverse active power, and the per-unit value of the voltage at each distributed power source grid connection point are combined into a multi-dimensional vector. This vector comprehensively reflects the electrical state characteristics of the distribution network at that time in two dimensions: voltage level and power flow distribution, and each component has been dimensionless.

[0020] S22. Calculate the change amplitude of the multidimensional electrical state vector at adjacent sampling times, and use it as the rate of change of the electrical state vector; Specifically, for two adjacent sampling times and Obtain the corresponding multidimensional electrical state vectors respectively. and Using the Euclidean norm as a metric, the magnitude of change between the two is calculated using the following formula: in: and Indicates the sampling time and The per-unit value of the feeder outlet voltage is the ratio of the actual effective value of the voltage measured at the feeder outlet at that moment to the reference value of the feeder rated voltage. and Indicates the sampling time and The normalized value of the reverse active power of the feeder is the ratio of the actual value of the reverse active power measured at the feeder outlet at that moment to the reference value of the rated transmission capacity of the feeder. When the power flow direction is from the feeder to the substation bus, this value is positive. and Indicates the sampling time and The per-unit value of the distributed power source grid connection point voltage is taken as the ratio of the actual effective voltage value measured at the distributed power source grid connection point at that moment to the rated voltage reference value of that grid connection point. This variation range... The rate of change of the electrical state vector is used to quantify the overall change in the electrical state of the distribution network within adjacent sampling intervals.

[0021] S23. When the rate of change of the electrical state vector is detected to be less than the preset threshold for the rate of change of the electrical state vector, the current sampling time is recorded as the candidate start time, and the rate of change of the electrical state vector at subsequent sampling times is continuously monitored. Specifically, the preset electrical state change rate threshold is an empirically determined parameter pre-set based on the distribution network voltage level and power quality standards. It serves as a critical value to define whether the electrical state of the distribution network has undergone a significant change. Taking a 10kV distribution network feeder as an example, the threshold value ranges from 0.005 to 0.015, preferably 0.01. During the quasi-steady-state operation of the distribution network, the electrical state change amplitude caused by natural load fluctuations and minor changes in distributed power output between adjacent sampling times typically does not exceed 0.01. However, when transient events such as a sudden drop in photovoltaic power or load surges occur, the electrical state change amplitude will increase sharply and significantly exceed this threshold. For example, when a 10kV feeder is operating stably on a clear day, the rate of change of the electrical state vector at adjacent sampling times remains between 0.003 and 0.008, below the threshold of 0.01. However, when clouds rapidly obstruct the photovoltaic panels, causing the output of distributed power sources to drop by 40% within seconds, this rate of change will instantly jump to above 0.05, significantly exceeding the threshold. This accurately indicates that the system has entered a transient disturbance process. The specific value of this electrical state change rate threshold can be adaptively adjusted according to the actual operating procedures of the distribution network and the monitoring accuracy.

[0022] S24. When the rate of change of the electrical state vector at each adjacent sampling time is less than the threshold of the rate of change of electrical state vector within the duration from the candidate start time, and the duration reaches the adaptive duration threshold adjusted by the output fluctuation rate of the distributed power source, the time series data from the candidate start time to the current time is extracted as an effective analysis segment.

[0023] In this embodiment, the adaptive duration threshold is used to determine whether a quasi-steady-state operation process has reached a critical time threshold for an effective duration. This threshold is a variable parameter that is dynamically adjusted according to the power output fluctuation of the distributed power source. The determination of the adaptive duration threshold specifically includes the following: 1. Calculate the standard deviation of the total active power output of each distributed power source grid connection point during the current analysis period, as a quantitative value of the output fluctuation of distributed power sources; 2. Based on the quantized value, determine the adaptive duration threshold through a preset negative correlation function relationship; where the larger the quantized value, the smaller the determined adaptive duration threshold.

[0024] Specifically, the fluctuation in the output of distributed power sources is calculated using the standard deviation of the time-series data of the total active power output of each distributed power source within the current analysis period. As a metric, its calculation expression is: in: This represents the number of sampling points within the current analysis period. For the first The total active power output of each distributed power source at each sampling time. This represents the average total active power output of all distributed power sources during the current analysis period. Standard deviation The larger the value, the more drastic and frequent the fluctuations in the output of the distributed power source within the current time period. Adaptive duration judgment value. With the degree of fluctuation in the output of distributed power sources They are negatively correlated, which can be expressed as an exponential decay function: in: The base duration is a benchmark value, ranging from 30 seconds to 180 seconds, with 60 seconds being the preferred value; The attenuation coefficient ranges from 0.5 to 2.0, with a preferred value of 1.0. This function smoothly reflects the continuous influence of fluctuation on the duration determination value: when the weather is sunny and the photovoltaic output is stable, The value is small. Approaching A longer steady-state determination time is used to ensure that the extracted segments have sufficient steady-state representativeness; when the weather is cloudy and the photovoltaic output fluctuates drastically, The value is relatively large. The value decreases rapidly in an exponential manner, which correspondingly shortens the minimum time required for steady-state determination, in order to avoid the incorrect rejection of a large number of true quasi-steady-state segments due to overly strict determination criteria.

[0025] When extracting valid analysis segments, starting from the candidate start time... Starting from this point, the rate of change of the electrical state vector at each subsequent sampling time is continuously monitored. When a certain sampling time is detected... At that time, if from to During this period, the rate of change of the electrical state vector in each adjacent sampling interval satisfies a threshold value less than or equal to the rate of change of the electrical state. The conditions, and the duration of that time interval. The adaptive duration threshold determined under the current operating conditions has been reached or exceeded. Then the time interval All time-series data within a given period are extracted into a single valid analysis segment. This valid analysis segment is a subset of continuous time-series data that has been determined to be in an electrical quasi-steady-state operating state. It eliminates disturbances caused by transient fluctuations in distributed generation output and instantaneous load switching, thus accurately reflecting the voltage distribution, power flow direction, and line load characteristics of the distribution network under quasi-steady-state operating conditions. For example, during a cloudy morning, if the monitoring system detects that the rate of change of the electrical state vector is consistently below 0.01 starting at 09:15:00, this moment is recorded as a candidate start time. For the next 20 seconds, the rate of change remains between 0.008 and 0.010, not exceeding the threshold. When the duration reaches the adaptive duration threshold of 22 seconds under this cloudy condition, the system extracts the time-series data from 09:15:00 to 09:15:22 into a valid analysis segment. This segment accurately records the quasi-steady-state operating characteristics of the distribution network during brief periods of clear weather between clouds.

[0026] In this embodiment, during continuous monitoring, if the rate of change of the electrical state vector at any sampling moment is detected to be greater than the electrical state change rate threshold before the duration reaches the adaptive duration threshold, the monitoring record starting from the candidate start time is abandoned, and continuous monitoring is re-executed with the sampling moment where the rate of change exceeds the limit as the new monitoring start time. When a sampling moment where the rate of change of the electrical state vector is less than the electrical state change rate threshold is subsequently detected, the sampling moment where the rate of change exceeds the limit is taken as the new candidate start time.

[0027] The continuous monitoring process here refers to the process from the candidate's initial time. After being recorded, the rate of change is calculated and compared with the threshold for each subsequent sampling time; and a series of continuous operations are performed to accumulate the timing. In actual distribution network operation, the output of distributed power sources may experience brief fluctuations of only a few seconds due to momentary cloud cover. During this period, although the rate of change of electrical state briefly falls below the threshold, a truly stable quasi-steady-state operating state is not formed. If this is not excluded, such segments will be misjudged as valid analysis segments. Specifically, when the self The cumulative duration has not yet reached the adaptive duration threshold. At a certain sampling time The rate of change of the electrical state vector at that moment was detected. Greater than the threshold This indicates that from to During this time interval, the distribution network has not yet formed a stable quasi-steady-state operation; the brief periods of stability that occur are merely accidental phenomena during transient disturbances. Therefore, the candidate start time is determined. Invalid. Discard all accumulated monitoring records starting from this moment, including recorded electrical state change amplitude sequences and accumulated time spans, as well as other intermediate monitoring data, to ensure that invalid candidate segments are not incorrectly included in subsequent analysis processes; simultaneously, at the sampling moment where the rate of change exceeds the limit... Starting from a new monitoring point, continuous monitoring is restarted. When a sampling moment is subsequently detected where the rate of change of the electrical state vector is less than the threshold, this sampling moment is recorded as a new candidate starting moment, and the continuous monitoring and judgment process restarts. By promptly discarding invalid candidate moments and resetting the monitoring process, it is ensured that every valid analysis segment ultimately obtained has a true electrical quasi-steady-state physical meaning.

[0028] S3. For the effective analysis segment, construct the power grid source-load coupling sensitivity matrix, calculate the marginal contribution of each distributed power source grid connection point to the abnormal index of feeder operation characteristics based on the sensitivity matrix, and locate the target distributed power source cluster that causes the abnormal operation characteristics based on the marginal contribution. The power distribution network source-load coupling sensitivity matrix is ​​a coefficient matrix used to describe the quantitative response relationship between changes in active power output at each distributed generation point of connection and voltage changes at key nodes on the feeder in the power distribution network. This matrix reflects the sensitivity of the impact of changes in source-end power injection on load-end voltage distribution; furthermore, it identifies the target distributed generation cluster that plays a dominant role in anomalies based on the marginal contribution. (See also...) Figure 3 Specifically, it includes the following: S31. Based on the distribution network topology and impedance parameters corresponding to the effective analysis segment, construct a distribution network source-load coupling sensitivity matrix that characterizes the relationship between the change in active power output at each distributed power source grid connection point and the voltage change at each key node. Specifically, distribution network topology refers to the electrical connections between nodes on a distribution network feeder, while impedance parameters refer to the resistance and reactance values ​​of each line segment. Distribution network topology and impedance parameters can be used together to construct the node admittance matrix of the distribution network, and the sensitivity relationship between node voltage and injected power can be derived through the Jacobian matrix of the power flow equations. Assume that the distribution network feeder corresponding to the effective analysis segment contains a total of... There are 1 distributed power generation grid connection points, denoted as follows: and includes There are several key nodes that require operational characteristic monitoring. These key nodes include the feeder outlet node, as well as the nodes at each level of sectionalizing switches and the line end node distributed along the feeder route, denoted as follows: For any valid analysis segment, the average active power output of each distributed power source connected to the grid and the average voltage of each key node within that segment are extracted as the steady-state operating point of that segment. Based on the node admittance matrix and power flow calculation equations at this steady-state operating point, an AC power flow sensitivity analysis method is used to establish the node voltage-injected power sensitivity relationship. Let the th... Distributed power grid connection point With the Key Nodes The voltage-active power sensitivity coefficient between them is Its meaning is: when distributed power sources When the active power output changes by a unit, it causes a critical node. The change in voltage amplitude. Then the source-load coupling sensitivity matrix of this distribution network. It can be represented as follows OK Column matrix form: Where: sensitivity coefficient The calculation can be obtained by inverting the Jacobian matrix of the power flow equations of the distribution network. A small increase in active power output is applied to the grid connection point of a distributed power source. Calculate the first Voltage changes at key nodes Then the sensitivity coefficient This sensitivity matrix characterizes the independent influence of each distributed generation source on the voltage of each key node. Due to the radial topology of the distribution network, the voltage sensitivity of distributed generation sources located at the end of the feeder to upstream nodes is usually greater than that of distributed generation sources located at the beginning of the feeder. This characteristic will be reflected in the subsequent calculation of marginal contribution.

[0029] S32. In response to the presence of abnormal feeder operating characteristics indicators within the effective analysis segment, each distributed power source grid connection point is divided into multiple participant subsets, and the change in the abnormal operating characteristics indicators caused by the combined action of each participant subset in the distribution network is determined. Specifically, abnormal feeder operating characteristics indicators are electrical features that deviate from the normal operating range, identified during the monitoring process in steps S1 to S2. Examples include feeder end-node voltage exceeding the upper limit or feeder reverse active power exceeding the limit. Let's take feeder end-node voltage exceeding the upper limit as a typical abnormal operating characteristic indicator: Assume that within this effective analysis segment, the feeder end-node... The actual voltage measurement value is The maximum allowable voltage at this node is [value missing]. The abnormal operating characteristic index is defined as the voltage exceeding the upper limit. .when If an operational anomaly is detected, the marginal contribution calculation process in this step is triggered. This involves all data in the distribution network... A group of distributed power sources constitutes the participants. For any subset of this set subset This represents the characteristic function values ​​of a selected group of distributed power sources. Defined as the voltage rise at the feeder end node caused by the combined action of distributed power sources within this subset in the distribution network; The calculation method is as follows: This subset The average active power output of each distributed source in the effective analysis segment is substituted into the distribution network power flow calculation model to calculate the corresponding feeder end node voltage value. Then, the calculated voltage value is subtracted from the reference voltage value when no distributed source is connected to obtain the subset. The corresponding voltage rise is used as the indicator change. By traversing all possible subsets of the participant set, the indicator change of the abnormal operating characteristic indicator corresponding to each participant subset can be determined, providing the characteristic function value basis for subsequent marginal contribution calculation.

[0030] S33. Based on the changes in indicators of each participant subset, the allocation rules of cooperative game are adopted to calculate the marginal contribution of each distributed power grid connection point to the abnormal operation characteristics indicators. Specifically, the allocation rule for cooperative game theory employs the Shapley value method, which can fairly allocate each participant's independent contribution to the overall outcome even when there are coupled interactions among multiple participants. Anomalies in the operating characteristics of a distributed power generation grid connection point marginal contribution The Shapley value is calculated using the Shapley formula as follows: in: Representing a subset The number of distributed power sources included. Indicates the first A distributed power source is added to the subset. The marginal increment of the end-node voltage caused by this; the Shapley value. For the first The marginal contribution of each distributed power source's grid connection point to the overall voltage exceeding the upper limit is calculated as follows: the sum of the marginal contributions of all distributed power sources equals the total anomaly index value. This allocation method can effectively isolate the coupling effects between distributed power sources, and truly reflect the degree of independent influence of each power source on abnormal operating characteristics.

[0031] S34. From all distributed power grid connection points, select those distributed power grid connection points with positive marginal contribution values ​​and positive contribution to abnormal operating characteristics; Specifically, after calculating the marginal contribution of each distributed power source's grid connection point in step S33, the marginal contribution is first filtered for positive or negative values. Under the allocation rules of cooperative game theory, the marginal contribution represents the independent contribution of each distributed power source to the abnormal operating characteristic indicators. When the marginal contribution of a distributed power source is positive, it indicates that the power source has a positive contributing effect on the abnormal operating characteristics, that is, its output change exacerbates the abnormal indicators; when the marginal contribution of a distributed power source is negative, it indicates that the power source has a suppressive or mitigating effect on the abnormal operating characteristics, that is, its output change reduces the degree of abnormality to some extent. For example, when the voltage of the feeder end node exceeds the upper limit, if the distributed power source located at the feeder front end has a large output, it may raise the voltage of the upstream node, which to some extent alleviates the degree of voltage exceeding the limit at the end node, and its marginal contribution is negative. This step marks distributed power grid-connected points with negative marginal contribution values ​​as sources of anomaly suppression and excludes them from the determination of the target distributed power cluster; only distributed power grid-connected points with positive marginal contribution values ​​that positively contribute to the abnormal operation characteristics are selected as candidates for subsequent sorting and accumulation.

[0032] S35. Sort the marginal contribution of each selected distributed power grid connection point in descending order of value; S36. The marginal contribution values ​​are accumulated sequentially according to the sorting order. When the accumulated value reaches or exceeds the preset accumulated contribution value threshold, all distributed power grid connection points participating in the accumulation are identified as the target distributed power cluster.

[0033] Specifically, the preset cumulative contribution threshold is a judgment parameter set based on actual engineering experience, used to define the cumulative contribution ratio of distributed power sources that play a dominant role in the abnormal operation characteristics, with a value range of 70% to 90%. The selected positive marginal contributions are sorted in descending order, and then accumulated sequentially starting from the largest marginal contribution at the top of the list. When the accumulated value first reaches or exceeds the product of the preset cumulative contribution threshold and the total abnormal indicator value, all distributed power source grid-connected points participating in this accumulation are identified as the target distributed power source cluster. This target distributed power source cluster is the main disturbance source group causing this abnormal operation characteristics; the sum of the contributions of its members to the abnormal indicators accounts for the vast majority of the overall abnormality. Subsequent step S4 will use this target distributed power source cluster as the starting point to reconstruct the disturbance propagation path analysis.

[0034] S4. Timestamp-align and spatial topology-associate the first sampling frequency data on the transformer side and the second sampling frequency data on the feeder side corresponding to the target distributed power generation cluster, and reconstruct the disturbance propagation path of the abnormal operation characteristics from the target distributed power generation cluster to other nodes on the distribution network feeder. After locating the target distributed power generation cluster causing the abnormal operating characteristics in step S3, the complete propagation process of the abnormality from the disturbance source to other nodes on the distribution network feeder is reconstructed. This enables visualized tracing of the abnormal operating characteristic propagation path, providing dispatchers with complete information on the entire evolution of the abnormality. Please refer to [link to relevant documentation]. Figures 4 to 5 Specifically, it includes the following: S411. Acquire the first sampling frequency data collected by the intelligent measurement terminal on the target distributed power cluster area side, and the second sampling frequency data collected by the feeder terminal unit on the distribution network feeder, wherein the sampling frequency corresponding to the first sampling frequency data is higher than the sampling frequency corresponding to the second sampling frequency data. Specifically, the first sampling frequency data is acquired through intelligent measurement terminals installed at the grid connection points of each distributed power source in the target distributed power source cluster. This first sampling frequency data includes low-voltage side voltage time-series data and active power time-series data for the distribution area. The second sampling frequency data is acquired through feeder terminal units deployed at key nodes on the distribution network feeders. This second sampling frequency data includes feeder node voltage time-series data and feeder node active power time-series data. The first sampling frequency is from 1 to 10 times per second, preferably 1Hz. This high-frequency data can precisely depict the instantaneous changes in grid connection point voltage and power on a second-level time scale. The second sampling frequency is from 1 to 5 minutes per minute, preferably 1 / 60Hz. This low-frequency data reflects the overall trend of electrical quantities at the feeder nodes. The sampling frequencies of the two types of data differ by orders of magnitude, therefore subsequent interpolation and alignment processing is required to achieve synchronization in the time dimension.

[0035] S412. Using the sampling time of the second sampling frequency data as a reference, perform interpolation calculation on the first sampling frequency data that falls within the preset time window to obtain equivalent data that is aligned with the sampling time of the second sampling frequency data; Specifically, the preset time window is a sliding window range determined based on the difference in sampling frequencies between the two types of data and empirical values ​​of the time scale for distribution network disturbance propagation. Since the sampling interval for the second sampling frequency data is larger than that for the first sampling frequency data, an alignment window needs to be defined for each sampling moment of the second sampling frequency data. For each sampling moment of the second sampling frequency data... Define an area centered at that moment with a width of time window ,in The preferred time window is set to 10 to 30 seconds. This window width is slightly larger than the sampling interval of the first sampling frequency data, ensuring that the window contains enough first sampling frequency data points for interpolation calculations, while simultaneously capturing the complete fluctuation characteristics of the first sampling frequency data near that moment. All first sampling frequency data points falling within this time window will be included. Using linear interpolation, the result at time t is calculated. The equivalent voltage and equivalent power values ​​at the location are shown. In a physical sense, this equivalent data represents the equivalent value of the high-frequency sampling data on the transformer side at the low-frequency sampling time on the feeder side, thus establishing a precise correspondence between the two types of data in the time dimension. This provides a time-synchronized data foundation for subsequent time-series causal analysis of disturbance propagation paths.

[0036] S413. Based on the topology information of the distribution network feeder, the aligned first sampling frequency data is associated and bound with the second sampling frequency data at the corresponding spatial location.

[0037] Specifically, the topology information of the distribution network feeders describes the structured data of the electrical connections between nodes on the feeders and the spatial locations of each node on the feeders. This information can be obtained from the distribution network geographic information system and includes radial connections with the feeder outlet as the root node; the access locations of each distributed generation grid connection point on the feeder; the deployment locations of each feeder terminal unit; and the upstream and downstream hierarchical relationships between nodes. Based on this topology information, the data of each distributed generation grid connection point, aligned to the transformer substation side, is linked to the second sampling frequency data of the feeder monitoring node with the nearest electrical distance. If a distributed generation grid connection point is located exactly between two adjacent feeder monitoring nodes in the topology, the association weight is allocated according to the ratio of the electrical distance from the grid connection point to the two monitoring nodes, so that the association binding result can accurately reflect the true electrical proximity relationship of each node in the distribution network topology.

[0038] In this embodiment, after completing the timestamp alignment and spatial topology association, the disturbance propagation path of the abnormal operating characteristics from the target distributed power generation cluster to other nodes on the distribution network feeder is further restored, specifically including the following: S421. Extract the voltage drop start time of each grid-connected point in the target distributed power cluster during the abnormal occurrence period from the equivalent data after timestamp alignment and spatial topology association; Specifically, the moment when the abnormal performance indicator in step S3 first exceeds the preset limit. Based on this, the preset observation period is traced back. Forming observation intervals ,in The value is determined based on the typical time scale of disturbance propagation in the distribution network, preferably between 60 and 120 seconds. This observation interval can cover the complete time span from the start of the disturbance source change to the abnormal index exceeding the limit. Within this abnormal observation interval, the aligned equivalent voltage time series data corresponding to each grid-connected point in the target distributed power generation cluster are analyzed. Perform a point-by-point scan. Voltage sag initiation time. The criterion for judgment is: from Backtracking, find the first condition that satisfies continuity. Voltage change rate at each sampling point All are negative and the cumulative voltage drop is Exceeding the preset threshold The sampling time, among which The preferred value is 3 to 5 to eliminate random fluctuations at individual sampling points; drop threshold Preferably, the value is taken as 3% to 5% of the rated voltage at the grid connection point to distinguish between normal minor voltage fluctuations and significant voltage drops caused by changes in distributed power output. The sampling time that meets the above conditions... This represents the initial time of voltage drop at the grid connection point during this abnormal event. The magnitude of the voltage drop is the cumulative change. .

[0039] S422. Starting from each grid connection point in the target distributed power cluster, along the topology information towards the feeder outlet, compare the order of the voltage drop start times of adjacent monitoring nodes level by level to determine the direction of disturbance propagation. Specifically, using the distribution network feeder topology information obtained in step S413, a feeder node adjacency matrix or adjacency table data structure is constructed. This data structure explicitly records the parent and child node relationships for each node. For each grid-connected point in the target distributed power generation cluster, the location of the feeder branch node to which the grid-connected point is connected is first located in the topology information, serving as the starting point for disturbance propagation analysis. Secondly, the process traces upstream along the parent node pointers in the topology: assuming the current node is... Query its parent node ,Will Add the traversal path sequence and update the current node. This process is repeated until the current node becomes the feeder exit node, at which point the traversal terminates. During the traversal, only node records belonging to feeder monitoring nodes are retained; intermediate nodes that are not monitoring nodes are only used as a basis for topology jumps and are not included in subsequent comparison sequences. After the traversal is complete, for each target distributed power grid connection point, a node sequence can be obtained that starts from that connection point and proceeds through each upstream feeder monitoring node to the feeder exit node, serving as a candidate path set for disturbance propagation analysis.

[0040] For each pair of adjacent feeder monitoring nodes on each candidate path, perform a time-series causality determination. and In terms of topological relationships for The parent node, i.e. Closer to the feeder outlet side, For the lower-level nodes, obtain the voltage sag start time of each of the two nodes from the voltage sag start time data extracted in step S421. and Compare the chronological order of the two: If This indicates that the voltage drop occurs at the upper-level node before the lower-level node, and the disturbance propagates from the upper-level node to the lower-level node. At this point, a determination is made. To disturb the upstream node, To disturb downstream nodes; if This indicates that a voltage drop occurs first at the lower-level node, and the disturbance propagates backward from the lower level to the upper level. In other words, the disturbance source is located downstream of this branch. At this point, it is determined that... To disturb the upstream node, To disturb downstream nodes; if ,in To account for the allowable threshold of data synchronization error, a value of 0.5 to 2 seconds is preferred. If the voltage drop occurs almost simultaneously, the two nodes are marked as synchronization disturbance nodes, indicating that they are at the same wavefront position of disturbance propagation.

[0041] Through the above step-by-step comparison, the directional relationship of disturbance propagation between each pair of adjacent nodes is clarified. It should be noted that when the determination result is that the voltage drop of the lower-level node occurs before that of the upper-level node, it indicates that the disturbance propagates from the lower level to the upper level on this local path. The target distributed power cluster is the source of the disturbance, and the voltage drop it causes spreads step by step along the topology path. The temporal sequence of each pair of adjacent nodes together constitutes a complete propagation chain from the disturbance source to the feeder outlet.

[0042] S423. Based on the disturbance propagation direction between adjacent nodes, construct a directed topology graph of disturbance propagation starting from the target distributed power generation cluster and ending at the feeder outlet node of the distribution network feeder, as the disturbance propagation path.

[0043] Specifically, the vertex set in the topology graph includes each grid-connected point of the target distributed power cluster; all feeder monitoring nodes and feeder exit nodes participating in the traversal. The rule for constructing the directed edge set in the graph is: for each pair of adjacent nodes with a clearly defined upstream and downstream relationship ( , Add a line from the graph to the graph. point to The directed edges represent the direction of propagation of operational anomalies from upstream to downstream nodes. If a pair of nodes is identified as synchronous disturbance nodes, a bidirectional edge is added between them, or an undirected edge is used to mark the special synchronization relationship. The resulting directed topology graph of disturbance propagation exhibits the following structural features: the grid-connected node corresponding to the target distributed power generation cluster serves as the disturbance source node with a small or zero in-degree; the feeder outlet node serves as the disturbance sink node with a zero out-degree; any directed path from the disturbance source node to the sink node corresponds to a possible spatial trajectory of disturbance propagation. This directed topology graph visually demonstrates the complete propagation chain of operational anomalies from the distributed power generation side to the substation bus side, including key information such as the location of the disturbance source; the sequence of intermediate monitoring nodes passed through by the disturbance; the directional relationship of the disturbance propagation; and the final impact range. The directed topology diagram of the disturbance propagation is identified as the result of the disturbance propagation path analysis for this operational anomaly. It can be presented graphically on the human-machine interface of the dispatch automation system, enabling dispatchers to clearly understand where the anomaly originated, along what path it spread, and which areas it ultimately affected.

[0044] S5. By integrating the quasi-steady-state operation segment, the target distributed power source cluster, and the disturbance propagation path, the results of the operational characteristic impact analysis of the distribution network are generated.

[0045] This section fuses and integrates the information from the quasi-steady-state operation segment captured in S2, the target distributed power cluster located in S3, and the disturbance propagation path reconstructed in S4, generating a structured operational characteristic impact analysis report. This provides dispatchers with complete analytical conclusions encompassing three stages: steady-state screening, root cause localization, and path reconstruction. The report includes the following: 1. Extract steady-state characteristic parameters from quasi-steady-state operation segments, extract the identifiers of each power source and the marginal contribution of each power source from the target distributed power source cluster, and extract the disturbance node sequence and time series features from the disturbance propagation path; 2. The extracted steady-state characteristic parameters, the identifiers of each power source and the marginal contribution of that power source, the perturbation node sequence and time series characteristics are assembled according to the preset association rules to generate a multi-dimensional impact analysis report that includes steady-state analysis, anomaly tracing and propagation process.

[0046] Specifically, steady-state characteristic parameters are extracted from the effective analysis segments output in step S2. These parameters include the start and end timestamps of the effective analysis segments, their duration, the average voltage level at the feeder outlet during quasi-steady-state operation, the average reverse load rate of the feeder, and the average total output of each distributed power source. From the target distributed power source cluster information output in step S3, the identifiers of each power source and its marginal contribution are extracted. The identifier of each power source is the grid connection point number, and the marginal contribution is the independent contribution value of each power source to the abnormal operating characteristic indicators calculated in step S33. From the disturbance propagation directed topology graph output in step S4, the disturbance node sequence and timing features are extracted. This disturbance node sequence is the ordered arrangement of each feeder monitoring node along the disturbance propagation path from each grid connection point of the target distributed power source cluster to the feeder outlet node. The timing features include the voltage drop start time and voltage drop magnitude of each node. The preset association rules are based on a pre-configured data organization template according to the actual needs of distribution network operation characteristic analysis. This template defines the logical correspondence between various types of extracted information: steady-state characteristic parameters are used as the basic background layer of the report to describe the quasi-steady-state operating conditions when the anomaly occurs; the identifiers of each power source and the marginal contribution of the power source are arranged in descending order of contribution value, and associated with the corresponding abnormal operation characteristic indicators to form an anomaly tracing layer; the disturbance node sequence is arranged in sequence according to the disturbance propagation direction, and the voltage drop start time and drop amplitude of each node correspond to the spatial position in the topology to form a propagation process layer.

[0047] The above three layers of information are assembled according to association rules to generate a multidimensional impact analysis report. The report is output in a structured data format and includes various visualization formats such as tables, curves, and topology annotation diagrams. This enables dispatchers to fully understand the operational background of the anomaly, the dominant cause of the anomaly, and the diffusion process and final impact range of the anomaly on the feeder. This provides data support and decision-making basis for formulating targeted distributed power source control strategies and optimizing the operation mode of the distribution network.

[0048] Example 2 An embodiment of the distribution network operation characteristic impact analysis device of this application includes the following: The time-series data acquisition unit is used to acquire time-series data of distribution network feeder output voltage, reverse active power, grid connection point voltage of each distributed power source, and output of each distributed power source. The effective analysis segment determination unit is used to extract quasi-steady-state operation segments from time series data based on a preset electrical state change rate threshold and an adaptive duration threshold adjusted by the output fluctuation rate of distributed power sources, so as to obtain at least one effective analysis segment. The target distributed power generation cluster location unit is used to construct a distribution network source-load coupling sensitivity matrix for effective analysis segments, calculate the marginal contribution of each distributed power generation grid connection point to the abnormal feeder operation characteristics index based on the sensitivity matrix, and locate the target distributed power generation cluster that causes the abnormal operation characteristics based on the marginal contribution. The disturbance propagation path restoration unit is used to align the first sampling frequency data of the transformer side and the second sampling frequency data of the feeder side corresponding to the target distributed power cluster with timestamps and spatial topology association, and restore the disturbance propagation path of the abnormal operation characteristics from the target distributed power cluster to other nodes on the distribution network feeder. The unit for generating operational characteristic impact analysis results is used to synthesize quasi-steady-state operation segments, target distributed power source clusters, and disturbance propagation paths to generate operational characteristic impact analysis results for the distribution network.

[0049] For specific limitations regarding the device, please refer to the limitations of the method above, which will not be repeated here. Each module in the above device can be implemented entirely or partially through software, hardware, or a combination thereof. Each module can be embedded in or independent of the processor in the computer device in hardware form, or stored in the memory of the computer device in software form, so that the processor can call and execute the operations corresponding to each module.

[0050] It is understood that those skilled in the art can combine various implementation methods in the above embodiments under the guidance of the above examples to obtain technical solutions with multiple implementation methods.

[0051] The above description is merely a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this application should be included within the protection scope of this application.

Claims

1. A method for analyzing the impact of distribution network operation characteristics, characterized in that, include: Acquire time-series data of distribution network feeder output voltage, reverse active power, grid connection point voltage of each distributed power source, and output of each distributed power source; Based on a preset electrical state change rate threshold and an adaptive duration threshold adjusted by the output fluctuation rate of distributed power sources, the time series data is truncated into quasi-steady-state operation segments to obtain at least one effective analysis segment. For the effective analysis segment, a distribution network source-load coupling sensitivity matrix is ​​constructed. Based on the sensitivity matrix, the marginal contribution of each distributed power source grid connection point to the abnormal feeder operation characteristics index is calculated. Based on the marginal contribution, the target distributed power source cluster causing the abnormal operation characteristics is located. The first sampling frequency data on the transformer side and the second sampling frequency data on the feeder side corresponding to the target distributed power cluster are timestamped and spatially topologically associated to reconstruct the disturbance propagation path of the abnormal operating characteristics from the target distributed power cluster to other nodes on the distribution network feeder. Based on the aforementioned quasi-steady-state operation segment, target distributed power source cluster, and disturbance propagation path, the analysis results of the impact on the operation characteristics of the distribution network are generated.

2. The method for analyzing the impact of distribution network operation characteristics according to claim 1, characterized in that, The process of extracting quasi-steady-state operation segments from the time-series data, based on a preset electrical state change rate threshold and an adaptive duration threshold adjusted by the output fluctuation rate of distributed power sources, includes: A multi-dimensional electrical state vector characterizing the operating state of the distribution network at each sampling moment is constructed from the time-series data of the feeder outlet voltage, reverse active power, and voltage at each distributed power source grid connection point. Calculate the change amplitude of the multidimensional electrical state vector at adjacent sampling times, and use it as the electrical state vector change rate; When the rate of change of the electrical state vector is detected to be less than the preset threshold for the rate of change of the electrical state, the current sampling time is recorded as the candidate start time, and the rate of change of the electrical state vector at subsequent sampling times is continuously monitored. When the rate of change of the electrical state vector at each adjacent sampling time is less than the electrical state change rate threshold within a duration starting from the candidate start time, and the duration reaches the adaptive duration threshold adjusted by the output fluctuation rate of the distributed power source, the time series data from the candidate start time to the current time is extracted as a valid analysis segment.

3. The method for analyzing the impact of distribution network operation characteristics according to claim 2, characterized in that, The adaptive duration threshold is determined in the following way: Calculate the standard deviation of the total active power output of each of the distributed power generation grid connection points during the current analysis period, and use it as a quantitative value of the output fluctuation of the distributed power generation. Based on the quantized value, an adaptive duration threshold is determined through a preset negative correlation function relationship; wherein, the larger the quantized value, the smaller the determined adaptive duration threshold.

4. The method for analyzing the impact of distribution network operation characteristics according to claim 2, characterized in that, During continuous monitoring, if the rate of change of the electrical state vector at any sampling moment is found to be greater than the electrical state change rate threshold before the duration reaches the adaptive duration threshold, the monitoring record starting from that candidate start moment is abandoned, and continuous monitoring is re-executed with the sampling moment where the rate of change exceeds the limit as the new monitoring start moment. When a sampling moment where the rate of change of the electrical state vector is less than the electrical state change rate threshold is subsequently detected, the sampling moment where the rate of change exceeds the limit is taken as the new candidate start moment.

5. The method for analyzing the impact of distribution network operation characteristics according to claim 1, characterized in that, For the effective analysis segment, a distribution network source-load coupling sensitivity matrix is ​​constructed. Based on the sensitivity matrix, the marginal contribution of each distributed power source connection point to the abnormal indicators of feeder operation characteristics is calculated, including: Based on the distribution network topology and impedance parameters corresponding to the effective analysis segment, a distribution network source-load coupling sensitivity matrix is ​​constructed to characterize the relationship between the active power output changes at each distributed power source grid connection point and the voltage changes at each key node. In response to the presence of abnormal feeder operating characteristics indicators within the effective analysis segment, each of the distributed power generation grid connection points is divided into multiple participant subsets, and the change in the abnormal operating characteristics indicators caused by the combined action of each participant subset in the distribution network is determined. Based on the changes in indicators of each subset of participants, the marginal contribution of each distributed power grid connection point to the abnormal operating characteristic indicators is calculated using the allocation rules of cooperative game theory.

6. The method for analyzing the impact of distribution network operation characteristics according to claim 5, characterized in that, The method of locating the target distributed power cluster causing abnormal operating characteristics based on the marginal contribution includes: From all the distributed power grid connection points, select those distributed power grid connection points with positive marginal contribution values ​​and positive contribution to the abnormal operating characteristics. The marginal contribution of each selected distributed power grid connection point is sorted in descending order of value; The marginal contribution values ​​are accumulated sequentially according to the sorting order. When the accumulated value reaches or exceeds the preset cumulative contribution threshold, all distributed power grid connection points that participated in the accumulation are identified as the target distributed power cluster.

7. The method for analyzing the impact of distribution network operation characteristics according to claim 1, characterized in that, The step of aligning the first sampling frequency data on the transformer substation side and the second sampling frequency data on the feeder side corresponding to the target distributed power cluster with timestamps and associating them with spatial topology includes: Acquire first sampling frequency data collected by the intelligent measurement terminal on the target distributed power cluster area side, and second sampling frequency data collected by the feeder terminal unit on the distribution network feeder, wherein the sampling frequency corresponding to the first sampling frequency data is higher than the sampling frequency corresponding to the second sampling frequency data. Using the sampling time of the second sampling frequency data as a reference, interpolation calculation is performed on the first sampling frequency data that falls within a preset time window to obtain equivalent data that is aligned with the sampling time of the second sampling frequency data; Based on the topology information of the power distribution network feeder, the aligned first sampling frequency data is associated and bound with the second sampling frequency data at the corresponding spatial location.

8. The method for analyzing the impact of distribution network operation characteristics according to claim 7, characterized in that, The restoration of the disturbance propagation path from the target distributed power cluster to other nodes on the distribution network feeder, as described above, includes: From the equivalent data after timestamp alignment and spatial topology association, extract the voltage drop start time of each grid-connected point in the target distributed power cluster during the abnormal occurrence period; Starting from each grid connection point in the target distributed power cluster, the order of the voltage drop start times of adjacent monitoring nodes is compared step by step along the topology information towards the feeder outlet to determine the disturbance propagation direction. Based on the disturbance propagation direction between adjacent nodes, a directed topology graph for disturbance propagation is constructed, starting from the target distributed power cluster and ending at the feeder outlet node of the distribution network feeder, as the disturbance propagation path.

9. The method for analyzing the impact of distribution network operation characteristics according to claim 1, characterized in that, The aforementioned quasi-steady-state operation segment; The target distributed generation cluster and disturbance propagation path are used to generate the impact analysis results on the operation characteristics of the distribution network, including: Steady-state characteristic parameters are extracted from the effective analysis segments in quasi-steady-state operation; the identifiers of each power source and the marginal contribution of each power source are extracted from the target distributed power source cluster; and the disturbance node sequence and time series features are extracted from the disturbance propagation path. The extracted steady-state characteristic parameters, the identifiers of each power source and the marginal contribution of that power source, the perturbation node sequence and time series characteristics are assembled according to preset association rules to generate a multi-dimensional impact analysis report that includes steady-state analysis, anomaly tracing and propagation process.

10. A device for analyzing the impact of power distribution network operation characteristics, characterized in that, The method for analyzing the impact of distribution network operation characteristics according to any one of claims 1 to 9 includes: The time-series data acquisition unit is used to acquire time-series data of distribution network feeder output voltage, reverse active power, grid connection point voltage of each distributed power source, and output of each distributed power source. The effective analysis segment determination unit is used to extract quasi-steady-state operation segments from the time series data based on a preset electrical state change rate threshold and an adaptive duration threshold adjusted by the output fluctuation rate of the distributed power source, so as to obtain at least one effective analysis segment. The target distributed power generation cluster location unit is used to construct a distribution network source-load coupling sensitivity matrix for the effective analysis segment, calculate the marginal contribution of each distributed power generation grid connection point to the abnormal feeder operation characteristics index based on the sensitivity matrix, and locate the target distributed power generation cluster that causes the abnormal operation characteristics based on the marginal contribution. The disturbance propagation path restoration unit is used to perform time stamp alignment and spatial topology association on the first sampling frequency data of the transformer side and the second sampling frequency data of the feeder side corresponding to the target distributed power cluster, and restore the disturbance propagation path of the abnormal operation characteristics from the target distributed power cluster to other nodes on the distribution network feeder. The unit for generating operational characteristic impact analysis results is used to integrate the quasi-steady-state operation segment, the target distributed power generation cluster, and the disturbance propagation path to generate operational characteristic impact analysis results for the distribution network.

Images