A power distribution network operation reliability evaluation method and system

Abstract

The application belongs to the technical field of power distribution network analysis, and provides a power distribution network operation reliability evaluation method and system, comprising: obtaining wind speed time series data of a to-be-evaluated period; constructing a power distribution network line reliability model considering wind speed time-varying evolution characteristics and fault repair time lag based on the obtained wind speed time series data; calculating time-varying fault probabilities of power transmission lines in different periods according to the constructed power distribution network line reliability model; obtaining optimal loss of load of the power distribution network in combination with the obtained time-varying fault probabilities; predicting the optimal loss of load of the power distribution network according to the obtained optimal loss of load of the power distribution network and an optimal loss of load of the power distribution network agent model; and calculating a power distribution network operation reliability index according to the obtained optimal loss of load of the power distribution network prediction value, so as to complete the evaluation of the power distribution network operation reliability.

Description

Technical Field

[0001] This invention belongs to the field of power distribution network analysis technology, and specifically relates to a method and system for assessing the operational reliability of power distribution networks. Background Technology

[0002] The statements in this section are merely background information related to the present invention and do not necessarily constitute prior art.

[0003] With the large-scale integration of distributed power sources such as wind and solar power into distribution networks, traditional distribution networks are gradually evolving into active distribution networks with multi-source coordinated operation. The output of new energy sources exhibits strong randomness and volatility, while user loads show significant time-varying characteristics. The combination of these two factors makes the distribution network prone to spatiotemporal mismatches between power sources and loads, as well as supply-demand imbalances, during operation, thus increasing system operational risks.

[0004] Overhead lines and other equipment in power distribution networks are exposed to the natural environment for extended periods, and their failure rate is significantly affected by meteorological factors such as wind speed. When wind speed increases, transmission lines may fail due to damage to towers, conductor swaying, or contact with vegetation, and the fault repair process typically involves time lags in location, dispatch, and emergency repair. Therefore, the reliability parameters of power distribution network components exhibit significant time-varying characteristics.

[0005] Existing methods for assessing the reliability of power distribution networks mainly include analytical methods and simulation methods; specifically: Analytical methods are prone to problems such as complex modeling, high dimensionality, and difficulty in solving when dealing with distribution networks containing complex closed-loop topologies and multiple distributed sources. While simulation methods, such as Monte Carlo simulation, have a wide range of applications, they require repeated optimal power flow calculations for a large number of random fault scenarios, which is computationally time-consuming and makes it difficult to meet the timeliness requirements of online evaluation during the operational phase.

[0006] In recent years, data-driven methods have been increasingly applied to power system reliability assessment, improving assessment efficiency by establishing a mapping relationship between system state and load shedding. However, existing methods are mostly based on traditional static reliability models, treating component failure rates as constants, making it difficult to accurately reflect the dynamic impact of time-varying environmental factors such as wind speed on line vulnerability. Using static failure rates in distribution network operation assessments can easily lead to deviations between training samples and actual operating conditions, thus affecting the assessment accuracy of data-driven models.

[0007] Therefore, there is an urgent need to propose a method for assessing the reliability of power distribution networks that can simultaneously take into account the spatiotemporal evolution characteristics of wind speed and the lag effect of fault repair, and has high accuracy and high efficiency. Summary of the Invention

[0008] To address the aforementioned issues, this invention proposes a method and system for assessing the operational reliability of distribution networks. This method fully considers the spatiotemporal evolution characteristics of wind speed, significantly improving assessment efficiency while ensuring assessment accuracy, and providing effective support for intraday rolling scheduling and online safety early warning of distribution networks.

[0009] According to some embodiments, the first solution of the present invention provides a method for assessing the operational reliability of a power distribution network, which adopts the following technical solution: A method for assessing the operational reliability of a power distribution network includes: Obtain wind speed time-series data for the period to be evaluated; Based on the acquired wind speed time series data, a reliability model for distribution network lines that takes into account the time-varying evolution characteristics of wind speed and the time lag of fault repair is constructed. Based on the constructed distribution network line reliability model, the time-varying failure probability of the transmission line at different time periods is calculated; By combining the obtained time-varying fault probabilities, the optimal load shedding amount of the distribution network is obtained; Predict the optimal load shedding of the distribution network based on the obtained optimal load shedding of the distribution network and the proxy model of the optimal load shedding of the distribution network; The reliability index of the distribution network is calculated based on the obtained optimal load shedding prediction value, and the reliability of the distribution network is evaluated.

[0010] As a further technical limitation, in the process of constructing the reliability model of the distribution network line, the instantaneous failure rate of the transmission line is obtained by combining the acquired wind speed time series data with the basic failure rate of the transmission line and the critical wind speed threshold for the dynamic deterioration of the failure rate of the transmission line. Based on the instantaneous fault rate of the transmission line, and combined with the conditional probability of fault repair of the transmission line, the state transition relationship between the normal state and the fault state of the transmission line is obtained, that is, the reliability model of the distribution network line taking into account the time-varying evolution characteristics of wind speed and the fault repair time lag is obtained.

[0011] As a further technical limitation, in the process of obtaining the optimal load shedding amount of the distribution network, the operating state of the distribution network is sampled based on the obtained time-varying fault probability, combined with load fluctuation, distributed power generation output fluctuation and generator availability, and the optimal load shedding amount of the distribution network under each sampled state is calculated in combination with the optimal power flow model.

[0012] Furthermore, the sampling of the distribution network operation status includes continuous random sampling of node loads and the upper limit of available output of distributed wind turbine units; Discrete fault sampling is performed on the line operation status and generator set operation status; and continuous power features are concatenated with discrete topology features to form an input feature vector.

[0013] As a further technical limitation, the reliability indicators of the distribution network operation include Expected Energy Not Supplied (EENS) and Loss of Load Probability (LOLP).

[0014] It should be noted that the total number of scenarios in the Monte Carlo sampling is... , No. The system load shedding amount in each sampling scenario is: The current assessment section's duration step is... The expected power shortage EENS can then be expressed as: ; The probability of load loss (LOLP) can be expressed as: ; in, This represents an indicator function, which takes the value 1 when the condition within the parentheses is true, and 0 otherwise.

[0015] As a further technical limitation, the optimal load shedding proxy model for the distribution network adopts a regression model based on the Extreme Gradient Boosting (XGBoost) algorithm.

[0016] It should be noted that the XGBoost algorithm is used to construct the optimal load shedding surrogate model for the distribution network; specifically: The wind speed forecast, load forecast, and initial topology state are input into the system, and the scenario to be evaluated is constructed based on the obtained time-varying line fault probability, combined with state sampling and input feature vectors. The generated scenario to be evaluated is input into the trained XGBoost proxy model to quickly predict the optimal load loss in each scenario. The prediction results for all scenarios are used to statistically analyze operational reliability indicators.

[0017] According to some embodiments, the second aspect of the present invention provides a power distribution network operation reliability assessment system, which adopts the following technical solution: A power distribution network operation reliability assessment system, comprising: The acquisition module is configured to acquire wind speed time-series data for the period to be evaluated; The module is configured to build a reliability model of the distribution network lines that takes into account the time-varying evolution characteristics of wind speed and the time lag of fault repair based on the acquired wind speed time series data. The calculation module is configured to calculate the time-varying failure probability of the transmission line at different time periods based on the constructed distribution network line reliability model; and to obtain the optimal load shedding amount of the distribution network by combining the obtained time-varying failure probability. The prediction module is configured to predict the optimal load shedding of the distribution network based on the obtained optimal load shedding of the distribution network and the optimal load shedding surrogate model of the distribution network; The evaluation module is configured to calculate the distribution network operation reliability index based on the obtained optimal load shedding prediction value of the distribution network, and complete the evaluation of the distribution network operation reliability.

[0018] As a further technical limitation, in the process of constructing the reliability model of the distribution network line, the instantaneous failure rate of the transmission line is obtained by combining the acquired wind speed time series data with the basic failure rate of the transmission line and the critical wind speed threshold for the dynamic deterioration of the failure rate of the transmission line. Based on the instantaneous fault rate of the transmission line, and combined with the conditional probability of fault repair of the transmission line, the state transition relationship between the normal state and the fault state of the transmission line is obtained, that is, the reliability model of the distribution network line taking into account the time-varying evolution characteristics of wind speed and the fault repair time lag is obtained.

[0019] As a further technical limitation, in the process of obtaining the optimal load shedding amount of the distribution network, the operating state of the distribution network is sampled based on the obtained time-varying fault probability, combined with load fluctuation, distributed power generation output fluctuation and generator availability, and the optimal load shedding amount of the distribution network under each sampled state is calculated in combination with the optimal power flow model.

[0020] Furthermore, the sampling of the distribution network operation status includes continuous random sampling of node loads and the upper limit of available output of distributed wind turbines; discrete fault sampling of line operation status and generator operation status; and concatenation of continuous power features and discrete topology features to form an input feature vector.

[0021] As a further technical limitation, the reliability indicators of the distribution network operation include the expected power shortage and the probability of load loss.

[0022] It should be noted that the total number of scenarios in the Monte Carlo sampling is... , No. The system load shedding amount in each sampling scenario is: The current assessment section's duration step is... The expected power shortage EENS can then be expressed as: ; The probability of load loss (LOLP) can be expressed as: ; in, This represents an indicator function, which takes the value 1 when the condition within the parentheses is true, and 0 otherwise.

[0023] As a further technical limitation, the optimal load shedding proxy model for the distribution network adopts a regression model based on the extreme gradient boosting algorithm.

[0024] It should be noted that the XGBoost algorithm is used to construct the optimal load shedding surrogate model for the distribution network; specifically: The wind speed forecast, load forecast, and initial topology state are input into the system, and the scenario to be evaluated is constructed based on the obtained time-varying line fault probability, combined with state sampling and input feature vectors. The generated scenario to be evaluated is input into the trained XGBoost proxy model to quickly predict the optimal load loss in each scenario. The prediction results for all scenarios are used to statistically analyze operational reliability indicators.

[0025] According to some embodiments, a third aspect of the present invention provides a computer-readable storage medium, employing the following technical solution: A computer-readable storage medium having a program stored thereon, which, when executed by a processor, implements the steps in a power distribution network operation reliability assessment method as described in the first aspect of the present invention.

[0026] According to some embodiments, the fourth aspect of the present invention provides an electronic device, which adopts the following technical solution: An electronic device includes a memory, a processor, and a program stored in the memory and running on the processor, wherein the processor executes the program to implement the steps in a power distribution network operation reliability assessment method as described in the first aspect of the present invention.

[0027] According to some embodiments, the fifth aspect of the present invention provides a computer program product, which adopts the following technical solution: A computer program product includes software code, wherein the program in the software code performs the steps of a power distribution network operation reliability assessment method as described in the first aspect of the present invention.

[0028] Compared with the prior art, the beneficial effects of the present invention are as follows: Based on the correlation between wind speed and line failure rate, this invention constructs a reliability model for distribution network lines that takes into account the time-varying evolution characteristics of wind speed and the time lag of fault repair, and obtains the time-varying failure probability of lines in each time period. This invention combines the source-load fluctuation characteristics of the distribution network with the random fault state of components, and uses a method that combines Monte Carlo sampling and optimal power flow calculation to generate multi-condition operating samples and construct a training dataset. This invention constructs an optimal load shedding proxy model based on the extreme gradient boosting algorithm, and uses the training dataset to mine the nonlinear mapping relationship between the system operating state and the optimal load shedding amount; This invention utilizes a trained surrogate model to quickly predict the load shedding of the operational section to be evaluated, and calculates distribution network operation reliability indicators based on the prediction results, including at least the expected power shortage and the probability of load shedding. This invention significantly improves assessment efficiency while ensuring assessment accuracy, providing effective support for daily rolling scheduling and online safety early warning of power distribution networks. Attached Figure Description

[0029] The accompanying drawings, which form part of this embodiment, are used to provide a further understanding of this embodiment. The illustrative embodiments and their descriptions are used to explain this embodiment and do not constitute an improper limitation of this embodiment.

[0030] Figure 1 This is a flowchart of a power distribution network operation reliability assessment method according to Embodiment 1 of the present invention; Figure 2 This is a structural block diagram of a power distribution network operation reliability assessment system according to Embodiment 2 of the present invention. Detailed Implementation

[0031] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0032] It should be noted that the following detailed descriptions are exemplary and intended to provide further illustration of the invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0033] It should be noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the scope of exemplary embodiments according to the invention. As used herein, the singular form is intended to include the plural form as well, unless the context clearly indicates otherwise. Furthermore, it should be understood that when the terms "comprising" and / or "including" are used in this specification, they indicate the presence of features, steps, operations, devices, components, and / or combinations thereof.

[0034] In this invention, terms such as "upper," "lower," "left," "right," "front," "back," "vertical," "horizontal," "side," and "bottom" indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. These terms are used only to facilitate the description of the structural relationships of the various components or elements of this invention and do not specifically refer to any component or element in this invention. They should not be construed as limiting the invention.

[0035] In this invention, terms such as "fixed connection," "connected," and "linked" should be interpreted broadly, indicating a fixed connection, an integral connection, or a detachable connection; a direct connection or an indirect connection through an intermediate medium. Those skilled in the art can determine the specific meaning of these terms in this invention based on the specific circumstances, and they should not be construed as limitations on the invention.

[0036] Where there is no conflict, the embodiments and features in the embodiments of the present invention can be combined with each other.

[0037] Terminology Explanation: Wind speed time series data: A set of wind speed measurements recorded continuously in chronological order. It is used to characterize the strong correlation between timestamps and wind speed values ​​and is often used for meteorological analysis, wind power forecasting, and disaster early warning.

[0038] Wind speed time-varying evolution characteristics: The dynamic law of wind speed changing over time is mainly manifested as non-stationarity, multi-scale pulsation and periodic fluctuation.

[0039] Fault Repair Time: Mean Time To Repair (MTTR) is the average time required for equipment to recover from a fault to the completion of repairs and restoration of normal function. It covers the entire process of fault detection, diagnosis, repair implementation, and system recovery. The lower the value, the higher the maintainability and reliability of the system.

[0040] Time-varying faults: Faults whose characteristics (such as amplitude, frequency, probability of occurrence, etc.) change dynamically with time or system operating conditions. The core of these faults is that the fault signals are non-stationary and time-varying, which causes traditional diagnostic methods based on steady-state assumptions to fail.

[0041] Load loss: Load demand that the system cannot meet due to a shortage or failure of generating capacity. It is a negative indicator used to characterize the loss of power supply reliability.

[0042] Distribution network operation reliability: This involves assessing the power supply reliability of existing or newly designed distribution networks under the conditions of existing lines and equipment, with the aim of determining power supply quality and guiding power grid transformation and planning.

[0043] Expected power shortage: The expected amount of electricity demand to be reduced due to a shortage of generating capacity or grid constraints in a power system within a given time period; it is an indicator that measures the expected amount of electricity that the power system cannot meet user demand, and falls under the category of adequacy in power system reliability assessment. It is one of the commonly used indicators for assessing the reliability of generation and transmission systems.

[0044] The probability of load shedding, also known as the probability of insufficient power, is a core probabilistic indicator in power system reliability assessment. It is defined as the probability that the available capacity of the power generation system cannot meet the system's maximum annual load demand. Its calculation requires the construction of a generator outage capacity model and a load forecasting model, and the quantification of the total probability of insufficient system reserve capacity through a risk model. Together with the expected value of insufficient power (LOLE) and the expected value of insufficient electricity (EENS), it constitutes a complete power generation system reliability indicator system, but it only reflects the probability in the time dimension and does not involve the scale of power outages. In power system planning, it is widely used to assess the impact of the intermittency of new energy sources, guide the optimization of power source planning, and verify the effectiveness of the power grid security defense system.

[0045] Proxy model: An approximate model used to solve complex simulation or experimental problems. In fields such as engineering design optimization, it can replace the computationally expensive original model for rapid evaluation. Optimization based on proxy model is to establish a preliminary model through limited high-precision experiments, and then use the model to predict better points for further experiments. Depending on the model type and problem complexity, it may converge to a local or global optimum, or it may not be able to find the optimum.

[0046] Extreme gradient boosting algorithm: an efficient and scalable machine learning algorithm implemented under the gradient boosting framework. Essentially, it constructs a series of decision trees in a sequential manner, with each new tree used to correct the errors of the previous tree. Finally, the prediction results of all trees are aggregated to form a strong prediction model.

[0047] Offline training: Training the model using pre-stored historical data without internet access or real-time data; the biggest feature is that all the data is provided at once, and the model does not interact with the outside world during training. Once trained, it can be used directly, which is particularly suitable for places with high data privacy requirements or poor network environment.

[0048] Online assessment: Based on real-time data, it performs dynamic quantitative analysis and prediction of the stable service capability of operating equipment, systems or software in the near future; it is not a one-time static inspection, but a continuous monitoring, early warning and optimization process; it "diagnoses" the health status of equipment through real-time sensor data, shifting from passive emergency repair to proactive maintenance.

[0049] DC Optimal Power Flow (DC-OPF): An optimization calculation method in power system analysis based on linearization assumptions. It determines the optimal active power output of each generator and the power flow distribution of branches by minimizing generation costs or other objective functions while satisfying the physical constraints of the power grid. DC-OPF is a simplification of AC Optimal Power Flow (AC-OPF), transforming the nonlinear AC power flow equations into linear equations, which greatly reduces the computational complexity.

[0050] Optimal load shedding is an emergency control strategy that determines the optimal load shedding amount for each node under the premise of satisfying system stability constraints (such as transient stability, frequency stability, and voltage stability) in the event of a power system failure or power deficit.

[0051] Intraday rolling dispatch: a key link in the multi-timescale optimization of the power system, refers to dynamically adjusting the unit output and energy storage strategies for the next short period of time based on the day-ahead plan, with a rolling cycle of 15 minutes to 1 hour, using ultra-short-term forecast data, in order to smooth out the fluctuations of new energy sources and track real-time load changes.

[0052] Operational reliability indicators are quantitative standards used to measure the continuous, safe, and high-quality power supply capability of a power system under actual operating conditions. Unlike traditional reliability assessments during the planning phase, they focus more on the dynamic impact of real-time operating conditions (such as equipment health status, load level, and weather conditions) on reliability.

[0053] Power flow calculation: By solving a set of nonlinear equations, the global state of the power system during steady-state operation is determined; it is used to check whether equipment is overloaded, that is, to determine the load rate of components such as transformers and lines under specific operating modes and to provide early warnings; it is used to assess power quality, that is, to check whether the voltage of each node is within the allowable range and to analyze voltage stability; it is used to optimize operating modes, that is, to provide dispatchers with a basis for adjusting generator output and reactive power compensation schemes to ensure the economical and safe operation of the system.

[0054] Example 1 Embodiment 1 of this invention introduces a method for evaluating the operational reliability of a power distribution network.

[0055] like Figure 1 The method for assessing the operational reliability of a power distribution network, as shown, includes: Obtain wind speed time-series data for the period to be evaluated; Based on the acquired wind speed time series data, a reliability model for distribution network lines that takes into account the time-varying evolution characteristics of wind speed and the time lag of fault repair is constructed. Based on the constructed distribution network line reliability model, the time-varying failure probability of the transmission line at different time periods is calculated; By combining the obtained time-varying fault probabilities, the optimal load shedding amount of the distribution network is obtained; Predict the optimal load shedding of the distribution network based on the obtained optimal load shedding of the distribution network and the proxy model of the optimal load shedding of the distribution network; The reliability index of the distribution network is calculated based on the obtained optimal load shedding prediction value, and the reliability of the distribution network is evaluated.

[0056] As one or more implementation methods, the process of constructing a line reliability model that takes into account the spatiotemporal evolution characteristics of wind speed in this embodiment adopts the following flow: "wind speed time series data input - wind speed failure rate correlation function - line instantaneous failure rate calculation - repair time log-normal distribution modeling - fault repair conditional probability calculation - line state transition probability update - line comprehensive failure probability output".

[0057] This embodiment obtains wind speed time-series data for the period to be evaluated and establishes a time-varying failure rate model for the line based on the correlation between wind speed and line failure rate.

[0058] In this embodiment, the basic fault rate of the line under normal operating conditions is assumed to be... ,time The wind speed is The critical wind speed threshold that causes the line failure rate to deteriorate dynamically is The instantaneous failure rate of the line It can be represented as: ; in, Indicates the basic failure rate of the line; Indicates time wind speed; Indicates the critical wind speed threshold; , and This represents the empirical vulnerability parameter obtained by fitting historical meteorological data and fault data. It is the proportional coefficient before the exponential term, used to adjust the amplification of the failure rate caused by wind speed; It is the wind speed index coefficient, used to characterize the rate at which the failure rate increases with increasing wind speed; It is a constant correction term in the exponential fitting function, used to correct the baseline offset of the fitting curve. These three parameters are not independent physical quantities, but together determine the shape of the fitting curve of line failure rate changing with wind speed.

[0059] In this embodiment, the line fault rate is adjusted in real time according to wind speed changes, based on the instantaneous line fault rate, so that the reliability modeling of the transmission line can reflect the dynamic impact of wind speed disturbances.

[0060] This embodiment considers that the emergency repair process after a line fault usually includes multiple stages such as fault location, material allocation and on-site construction operations. The repair time has obvious time lag and randomness. Therefore, a log-normal distribution is used to model the line repair time.

[0061] Let the random variable of repair time be... Then its probability density function It can be represented as: ; in, and Let represent the mean and standard deviation parameters of the log-normal distribution of repair time, respectively; its cumulative distribution function is denoted as . This means that the repair time is a random variable. The cumulative distribution function is used to represent the cumulative distribution of power after a line fault occurs, over time. The probability of completing the repair within a certain timeframe.

[0062] Based on the risk function theory, when the fault has lasted for a period of time... Under the condition of, the next time step The conditional probability of completing the repair within the time limit It can be represented as: ; Meanwhile, the normal line has a time step The probability of a new failure occurring within. It can be represented as: ; Based on this, this embodiment constructs the state transition relationship between the normal state and the fault state of the line.

[0063] set up Indicates time The line is in normal condition and the duration is The state probability, Indicates time The line is in a fault state and the duration is Given the state probabilities, its recursive update equations can be expressed as: ; Therefore, the line at time... The overall failure probability, i.e., the line unavailability rate. Its expression is: .

[0064] The results obtained This will serve as the basis for sampling the line operating status in the subsequent system status sampling phase.

[0065] As one or more implementation methods, this embodiment constructs a distribution network system state sampling and training sample generation; that is, after obtaining the time-varying fault probability of each line, a multi-condition sample set of the distribution network for surrogate model training is further constructed.

[0066] In this embodiment, basic data such as network topology, node load, wind power output, line parameters, and generator parameters are first input, and then the operation scenario is generated from both continuous and discrete variable levels.

[0067] For the continuous variable portion, let the system reference load vector be... , No. The random perturbation coefficient vector corresponding to each scenario is: Then the first Actual load injection vector in each scenario It can be represented as: ; in, This represents the matrix dot product.

[0068] For the power generation side, let the first... The theoretical maximum available power output vector of the generator set under each scenario is: The generator operating state vector is Then the actual available power generation capacity vector in the corresponding scenario It can be represented as: ; in, Indicates the total number of nodes. Indicates the total number of generators. D represents the total number of lines, and D represents the dimension of the feature vector.

[0069] For each sampling scenario, the corresponding optimal load loss amount needs to be obtained through the physical model as the supervised learning label.

[0070] Let the first Nodes in various scenarios The optimal load shedding amount is Then the optimal load loss label for this scenario It can be represented as: ; After repeated sampling across numerous scenarios, all input feature vectors are stacked row-wise to form a training feature matrix. , can be represented as: ; Corresponding label vector It can be represented as: ; This embodiment consists of a sample generation process of "inputting basic data - continuous variable sampling - discrete variable sampling - constructing scene input feature vectors - forming single scene operating status - repeated sampling - multi-condition training sample set".

[0071] As one or more implementation methods, this embodiment establishes an optimal load shedding solution model based on optimal DC power flow.

[0072] During the training sample generation phase, a feature vector needs to be input for each scenario, and the corresponding optimal load shedding amount is solved using the DC optimal power flow model. This step is used to complete the process of converting system operating status to labels. The physical mapping is the foundation for subsequent agent model training.

[0073] For any sampling scenario, establish an optimization model with the objective of minimizing the total system load shedding.

[0074] Let the total number of nodes be ,node The shear load is The actual active power output vector of the generator is The active power flow vector of the line is The node voltage phase angle vector is The optimal load shedding objective function can then be expressed as: ; Let the node-branch correlation matrix be... The generator access matrix is The node demand load vector is The maximum available capacity vector of the generator is The maximum transmission capacity vector of the line is The generator operating state Boolean vector is The Boolean vector of the line's operating status is Then the DC optimal power flow constraint set can be expressed as: ; For lines that remain connected, linearized phase angle constraints must also be applied.

[0075] Set up branch roads The meritorious trend is The line reactance is ,node and nodes The voltage phase angles are respectively and Then the phase angle constraint can be expressed as: ; And the main network access node is taken as the phase angle reference point, that is: ; When the system experiences extreme multiple faults and becomes isolated from the main network, the power flow model with phase angle constraints may fail to converge due to the lack of a unified phase angle reference point. In this case, the phase angle constraints can be discarded, and only the power balance constraints and capacity boundary constraints can be retained. The optimal load shedding under this state can then be solved using a network flow model.

[0076] "Power balance constraints, generator output constraints, line capacity constraints, node load shedding constraints, phase angle constraints / island network flow processing" together constitute the solution framework for optimal load shedding labels.

[0077] As one or more implementation methods, this embodiment constructs a proxy model for evaluating the reliability of power distribution network operation based on the extreme gradient boosting algorithm.

[0078] After completing sample generation and label calculation, this embodiment uses the XGBoost algorithm to construct an optimal load shedding proxy model for the distribution network; specifically: Wind speed forecast, load forecast, and initial topology state are input into the system, and based on the obtained time-varying line fault probability, combined with state sampling and input feature vectors, a scenario to be evaluated is constructed. The scenario to be evaluated is then input into the trained XGBoost surrogate model to quickly predict the optimal load loss in each scenario. Based on the prediction results of all scenarios, operational reliability indicators are statistically analyzed.

[0079] For any input sample The load shedding prediction value output by the XGBoost proxy model It can be represented as: ; in, This represents the total number of regression trees. Indicates the first The mapping function corresponding to each regression tree, This represents the function space of the regression tree.

[0080] In the In the first iteration, the XGBoost objective function, after second-order Taylor expansion, can be expressed as: ; in, This represents the first derivative of the loss function with respect to the previous round of predictions. This represents the second derivative of the loss function with respect to the previous round of predictions. This represents the structural risk regularization term used to control model complexity. Through cross-validation, parameters such as tree depth, learning rate, and the number of base learners can be optimized to obtain a high-precision optimal load shedding surrogate model suitable for the mixed continuous-discrete state characteristics of distribution networks.

[0081] As one or more implementation methods, this embodiment realizes a two-stage operational reliability assessment consisting of offline training and online evaluation.

[0082] In this embodiment, the reliability assessment process includes an offline training phase and an online assessment phase; specifically: The offline training phase is mainly used for wind speed time-varying reliability modeling, system state sampling, optimal load shedding label calculation, and XGBoost model training and parameter optimization; the online evaluation phase calls the trained proxy model to quickly infer the multi-scenario states of the current or future operating sections and to statistically analyze the operational reliability indicators.

[0083] During the offline training phase, time-varying reliability of the line is modeled; subsequently, multi-condition sampling is performed, and the optimal load shedding amount under each sampling scenario is solved using the DC optimal power flow model to form a training sample set. ; Finally, the XGBoost model is trained and its parameters are optimized, and the optimal model is saved. Through the above offline process, a large amount of physics solution work that originally needed to be repeated in the online stage can be completed in advance.

[0084] During the online assessment phase, forecast information for the current or future operating sections is first obtained, including wind speed forecasts, load forecasts, and the initial network topology status. Then, online state sampling is performed in the same manner as in the offline phase to construct a set of states to be evaluated; Call the trained XGBoost model to quickly output the optimal load loss prediction value for each scenario; Calculate the expected power shortage EENS and the probability of load shedding LOLP based on the prediction results of all scenarios.

[0085] Let the total number of scenes in the Monte Carlo sampling be... , No. The system load shedding amount in each sampling scenario is: The current assessment section's duration step is... The expected power shortage EENS can then be expressed as: ; The probability of load loss (LOLP) can be expressed as: ; in, This represents an indicator function, which takes the value 1 when the condition within the parentheses is true, and 0 otherwise.

[0086] Through the above two-stage process of offline training and online evaluation, this embodiment transfers a large number of optimal power flow calculations that need to be repeatedly performed in the online stage in traditional methods to the offline stage. This allows the online stage to only perform state sampling, surrogate model inference, and index statistics, thus significantly improving the efficiency of distribution network operation reliability assessment while ensuring evaluation accuracy.

[0087] Based on the correlation between wind speed and line failure rate, this embodiment constructs a distribution network line reliability model that takes into account the time-varying evolution characteristics of wind speed and the time lag of fault repair, and obtains the time-varying failure probability of the line in each time period. This embodiment combines the characteristics of power distribution network source-load fluctuations and the random fault states of components, and uses a combination of Monte Carlo sampling and optimal power flow calculation to generate multi-condition operating samples and construct a training dataset. This embodiment constructs an optimal load shedding proxy model based on the extreme gradient boosting algorithm, and uses the training dataset to mine the nonlinear mapping relationship between the system operating state and the optimal load shedding amount; This embodiment utilizes the trained surrogate model to quickly predict the load loss of the operating section to be evaluated, and calculates the distribution network operation reliability index based on the prediction results, including at least the expected power shortage and the probability of load loss. This embodiment significantly improves assessment efficiency while ensuring assessment accuracy, providing effective support for daily rolling dispatching and online safety early warning of the distribution network.

[0088] Example 2 Embodiment 2 of the present invention introduces a power distribution network operation reliability assessment system.

[0089] like Figure 2 The distribution network operation reliability assessment system shown includes: The acquisition module is configured to acquire wind speed time-series data for the period to be evaluated; The module is configured to build a reliability model of the distribution network lines that takes into account the time-varying evolution characteristics of wind speed and the time lag of fault repair based on the acquired wind speed time series data. The calculation module is configured to calculate the time-varying failure probability of the transmission line at different time periods based on the constructed distribution network line reliability model; and to obtain the optimal load shedding amount of the distribution network by combining the obtained time-varying failure probability. The prediction module is configured to predict the optimal load shedding of the distribution network based on the obtained optimal load shedding of the distribution network and the optimal load shedding surrogate model of the distribution network; The evaluation module is configured to calculate the distribution network operation reliability index based on the obtained optimal load shedding prediction value of the distribution network, and complete the evaluation of the distribution network operation reliability.

[0090] As one or more implementation methods, in the process of constructing the reliability model of the power distribution network line, the instantaneous failure rate of the transmission line is obtained by combining the acquired wind speed time series data with the basic failure rate of the transmission line and the critical wind speed threshold for the dynamic deterioration of the failure rate of the transmission line. Based on the instantaneous fault rate of the transmission line, and combined with the conditional probability of fault repair of the transmission line, the state transition relationship between the normal state and the fault state of the transmission line is obtained, that is, the reliability model of the distribution network line taking into account the time-varying evolution characteristics of wind speed and the fault repair time lag is obtained.

[0091] As one or more implementation methods, in the process of obtaining the optimal load shedding amount of the distribution network, the operating state of the distribution network is sampled based on the obtained time-varying fault probability, combined with load fluctuation, distributed power generation output fluctuation and generator availability, and the optimal load shedding amount of the distribution network under each sampled state is calculated in combination with the optimal power flow model.

[0092] As one or more implementation methods, the sampling of the power distribution network operation status includes continuous random sampling of node loads and the upper limit of available output of distributed wind turbine units; Discrete fault sampling is performed on the line operation status and generator set operation status; and continuous power features are concatenated with discrete topology features to form an input feature vector.

[0093] As one or more implementation methods, the reliability indicators of the power distribution network operation include the expected power shortage and the probability of load shedding.

[0094] It should be noted that the total number of scenarios in the Monte Carlo sampling is... , No. The system load shedding amount in each sampling scenario is: The current assessment section's duration step is... The expected power shortage EENS can then be expressed as: ; The probability of load loss (LOLP) can be expressed as: ; in, This represents an indicator function, which takes the value 1 when the condition within the parentheses is true, and 0 otherwise.

[0095] As a further technical limitation, the optimal load shedding proxy model for the distribution network adopts a regression model based on the Extreme Gradient Boosting (XGBoost) algorithm.

[0096] As one or more implementation methods, the optimal load shedding proxy model for the distribution network adopts a regression model based on the extreme gradient boosting algorithm.

[0097] It should be noted that the XGBoost algorithm is used to construct the optimal load shedding surrogate model for the distribution network; specifically: The wind speed forecast, load forecast, and initial topology state are input into the system, and the scenario to be evaluated is constructed based on the obtained time-varying line fault probability, combined with state sampling and input feature vectors. The generated scenario to be evaluated is input into the trained XGBoost proxy model to quickly predict the optimal load loss in each scenario. The prediction results for all scenarios are used to statistically analyze operational reliability indicators.

[0098] Based on the correlation between wind speed and line failure rate, this embodiment constructs a distribution network line reliability model that takes into account the time-varying evolution characteristics of wind speed and the time lag of fault repair, and obtains the time-varying failure probability of the line in each time period. This embodiment combines the characteristics of power distribution network source-load fluctuations and the random fault states of components, and uses a combination of Monte Carlo sampling and optimal power flow calculation to generate multi-condition operating samples and construct a training dataset. This embodiment constructs an optimal load shedding proxy model based on the extreme gradient boosting algorithm, and uses the training dataset to mine the nonlinear mapping relationship between the system operating state and the optimal load shedding amount; This embodiment utilizes the trained surrogate model to quickly predict the load loss of the operating section to be evaluated, and calculates the distribution network operation reliability index based on the prediction results, including at least the expected power shortage and the probability of load loss. This embodiment significantly improves assessment efficiency while ensuring assessment accuracy, providing effective support for daily rolling dispatching and online safety early warning of the distribution network.

[0099] The detailed steps are the same as those provided in Example 1 for evaluating the reliability of power distribution network operation, and will not be repeated here.

[0100] Example 3 Embodiment 3 of the present invention provides a computer-readable storage medium.

[0101] A computer-readable storage medium having a program stored thereon, which, when executed by a processor, implements the steps in a power distribution network operation reliability assessment method as described in Embodiment 1 of the present invention.

[0102] The detailed steps are the same as those provided in Example 1 for evaluating the reliability of power distribution network operation, and will not be repeated here.

[0103] Example 4 Embodiment 4 of the present invention provides an electronic device.

[0104] An electronic device includes a memory, a processor, and a program stored in the memory and running on the processor. When the processor executes the program, it implements the steps in the power distribution network operation reliability assessment method as described in Embodiment 1 of the present invention.

[0105] The detailed steps are the same as those provided in Example 1 for evaluating the reliability of power distribution network operation, and will not be repeated here.

[0106] Example 5 Embodiment 5 of the present invention provides a computer program product.

[0107] A computer program product includes software code, wherein the program in the software code performs the steps of a power distribution network operation reliability assessment method as described in Embodiment 1 of the present invention.

[0108] The detailed steps are the same as those provided in Example 1 for evaluating the reliability of power distribution network operation, and will not be repeated here.

[0109] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code. The solutions in the embodiments of the present invention can be implemented using various computer languages, such as the object-oriented programming language Java and the interpreted scripting language JavaScript.

[0110] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0111] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0112] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0113] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of the invention.

[0114] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

[0115] The above description is merely a preferred embodiment of this practice and is not intended to limit the scope of this practice. Various modifications and variations can be made to this practice by those skilled in the art. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of this practice should be included within the protection scope of this practice.

Claims

1. A method for assessing the operational reliability of a power distribution network, characterized in that, include: Obtain wind speed time-series data for the period to be evaluated; Based on the acquired wind speed time series data, a reliability model for distribution network lines that takes into account the time-varying evolution characteristics of wind speed and the time lag of fault repair is constructed. Based on the constructed distribution network line reliability model, the time-varying failure probability of the transmission line at different time periods is calculated; By combining the obtained time-varying fault probabilities, the optimal load shedding amount of the distribution network is obtained; Predict the optimal load shedding of the distribution network based on the obtained optimal load shedding of the distribution network and the proxy model of the optimal load shedding of the distribution network; The reliability index of the distribution network is calculated based on the obtained optimal load shedding prediction value, and the reliability of the distribution network is evaluated.

2. The method for assessing the operational reliability of a distribution network as described in claim 1, characterized in that, In the process of constructing the reliability model of the distribution network line, the instantaneous failure rate of the transmission line is obtained by combining the acquired wind speed time series data with the basic failure rate of the transmission line and the critical wind speed threshold for the dynamic deterioration of the failure rate of the transmission line. Based on the obtained instantaneous failure rate of the transmission line, the state transition relationship between the normal state and the fault state of the transmission line is obtained by combining the conditional probability of the fault repair of the transmission line. That is, the reliability model of the distribution network line considering the time-varying evolution characteristics of wind speed and the time lag of fault repair is obtained.

3. The method for assessing the operational reliability of a distribution network as described in claim 1, characterized in that, In obtaining the optimal load shedding amount of the distribution network, the operating status of the distribution network is sampled based on the obtained time-varying fault probability, combined with load fluctuations, distributed power generation output fluctuations, and generator availability status. The optimal load shedding amount of the distribution network under each sampled status is then calculated using the optimal power flow model.

4. The method for assessing the operational reliability of a distribution network as described in claim 3, characterized in that, The sampling of the distribution network operation status includes continuous random sampling of node loads and the upper limit of available output of distributed wind turbines; discrete fault sampling of line operation status and generator operation status; and concatenation of continuous power features and discrete topology features to form an input feature vector.

5. The method for assessing the operational reliability of a distribution network as described in claim 1, characterized in that, The reliability indicators of the power distribution network include the expected power shortage and the probability of load loss.

6. The method for assessing the operational reliability of a distribution network as described in claim 1, characterized in that, The optimal load shedding proxy model for the distribution network adopts a regression model based on the extreme gradient boosting algorithm.

7. A power distribution network operation reliability assessment system, characterized in that, include: The acquisition module is configured to acquire wind speed time-series data for the period to be evaluated; The module is configured to build a reliability model of the distribution network lines that takes into account the time-varying evolution characteristics of wind speed and the time lag of fault repair based on the acquired wind speed time series data. The calculation module is configured to calculate the time-varying failure probability of transmission lines at different times based on the constructed distribution network line reliability model. By combining the obtained time-varying fault probabilities, the optimal load shedding amount of the distribution network is obtained; The prediction module is configured to predict the optimal load shedding of the distribution network based on the obtained optimal load shedding of the distribution network and the optimal load shedding surrogate model of the distribution network; The evaluation module is configured to calculate the distribution network operation reliability index based on the obtained optimal load shedding prediction value of the distribution network, and complete the evaluation of the distribution network operation reliability.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the steps of a power distribution network operation reliability assessment method as described in any one of claims 1-6.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the program, it implements the steps of a power distribution network operation reliability assessment method as described in any one of claims 1-6.

10. A computer program product, comprising software code, characterized in that, The program in the software code executes the steps of a power distribution network operation reliability assessment method as described in any one of claims 1-6.

Images