Power distribution network operation risk assessment method and device, electronic equipment and storage medium

Abstract

The invention discloses a power distribution network operation risk assessment method and device, electronic equipment and a storage medium, and relates to the technical field of power systems or other related fields, and the method comprises the steps: obtaining the charging power data of N electric vehicles in a target time period, N being a positive integer; node load flow calculation is carried out based on the charging power data, and probabilistic load flow expressions of N electric vehicles are established; based on the load flow calculation result and the probabilistic load flow expression, establishing a Gaussian mixture distribution body to obtain a Gaussian mixture probability model; and acquiring real-time charging power data of the N electric vehicles, and performing operation risk assessment on the target power distribution network based on the real-time charging power data and the Gaussian mixture probability model to obtain an assessment result. According to the invention, the technical problem of low evaluation efficiency and accuracy caused by high calculation complexity of power distribution network operation risk evaluation when the electric vehicle accesses the power grid in a large scale in the prior art is solved.

Description

Technical Field

[0001] The present invention relates to the technical field of power systems, and in particular to a method and device for evaluating the operating risk of a distribution network, electronic equipment, and a storage medium. Background Art

[0002] The rapid development and popularization of electric vehicles (EVs) are changing the operation mode of traditional power systems. As more and more electric vehicles are connected to the power grid, the impact of their charging demand on the distribution network is becoming more and more significant. One is the immediate impact of electric vehicle charging behavior on the operation risk of the power grid, and the other is the long-term impact of electric vehicle charging behavior on the operation risk of the power grid. The immediate impact refers to the immediate impact of electric vehicle charging behavior on the operation risk of the power system, such as the impact of electric vehicle charging behavior on the local node voltage or the mean value of local node voltage of the power grid. The long-term impact refers to the long-term impact of electric vehicle charging behavior on the operation risk of the power system, such as the impact of electric vehicle charging behavior on the frequency stability of the power grid. The randomness and uncertainty of electric vehicle charging demand bring huge risk challenges to the planning and operation of the power system. Demand response and other means can adjust the power distribution of electric vehicles and smooth their operation risks. However, the potential risks brought to the power network by the optimized electric vehicle power distribution still need to be effectively evaluated.

[0003] To address this challenge, researchers proposed a probabilistic power flow (PPF) analysis method that quantifies the impact of electric vehicle charging demand on the distribution network by considering the uncertainty of electric vehicle charging demand. Probabilistic power flow analysis can not only describe the random state of the system, but also evaluate potential risks such as low voltage and line overload caused by random fluctuations. However, with the increase in the number of electric vehicles, computational complexity has become an urgent problem to be solved. Existing PPF methods may face the dilemma of balancing computational efficiency and accuracy when dealing with large-scale electric vehicle integration, but neither is ideal.

[0004] To address the above-mentioned problems, no effective solution has been proposed yet. Summary of the invention

[0005] The embodiments of the present invention provide a method and device for evaluating the operation risk of a distribution network, an electronic device, and a storage medium, so as to at least solve the technical problem in the related art that the computational complexity of the distribution network operation risk evaluation when electric vehicles are connected to the power grid on a large scale is high, resulting in low evaluation efficiency and accuracy.

[0006] According to one aspect of an embodiment of the present invention, a method for assessing the operation risk of a distribution network is provided, comprising: obtaining charging power data of N electric vehicles within a target time period, wherein the charging power data refers to node power injection data of a grid node to which the electric vehicle is connected, and N is a positive integer; performing node power flow calculation based on the charging power data, and establishing a probabilistic power flow expression for the N electric vehicles, wherein the probabilistic power flow expression obeys a Gaussian distribution and is used to describe the correlation change between the real-time voltage of the grid node and the power injection disturbance, and the power injection disturbance refers to the node power injection data related to the random charging behavior of the electric vehicle; establishing a Gaussian mixture distribution body based on the power flow calculation result and the probabilistic power flow expression to obtain a Gaussian mixture probability model, wherein the Gaussian mixture probability model is used to monitor the real-time voltage disturbance of any of the random charging behaviors on the relevant grid nodes; obtaining the real-time charging power data of the N electric vehicles, and performing an operation risk assessment on the target distribution network based on the real-time charging power data and the Gaussian mixture probability model to obtain an assessment result.

[0007] Furthermore, the step of performing node power flow calculation based on the charging power data includes: for each of the grid nodes, establishing a power flow equation at the grid node based on the active power, reactive power and node voltage in the node power injection data; for each of the power flow equations, determining a reference operating point, and linearizing the power flow equation at the reference operating point to obtain a linearized expression of the power flow equation, wherein the reference operating point refers to a steady-state operating point at which the system load and power generation are balanced; treating the power injection disturbance as a random variable, formatting the linearized expression, and obtaining the power flow calculation result in which the node voltage changes in real time with the power injection disturbance.

[0008] Furthermore, the step of establishing a probabilistic power flow expression for N electric vehicles includes: for each electric vehicle, quantifying each random charging behavior to obtain R charging behavior parameters, where R is a positive integer; taking each charging behavior parameter as a first distribution component, performing Gaussian distribution fitting on the charging power data, and determining a first component parameter during the fitting process, wherein the first component parameter includes at least: a mixing weight, a mean, and a standard deviation; determining a probabilistic power flow expression for the power injection disturbance of each electric vehicle based on the first component parameter and a preset Gaussian distribution expression to obtain N probabilistic power flow expressions.

[0009] Furthermore, a step of establishing a Gaussian mixture distribution body based on the power flow calculation results and the probabilistic power flow expression to obtain a Gaussian mixture probability model includes: taking each of the electric vehicles as a second distribution component, performing Gaussian mixture distribution fitting based on the power flow calculation results and the probabilistic power flow expression, and determining the second component parameters during the fitting process; determining the analytical expression of the Gaussian mixture probability model based on the second component parameters and a preset Gaussian distribution expression to obtain the Gaussian mixture probability model.

[0010] Furthermore, after a Gaussian mixture distribution body is established based on the flow calculation results and the probabilistic flow expression to obtain a Gaussian mixture probability model, it also includes: merging the N original distribution components in the Gaussian mixture probability model based on a preset component reduction algorithm to obtain a lightweight model of the Gaussian mixture probability model; for the R updated distribution components in the lightweight model, redetermining the component parameters in the Gaussian distribution expression of each updated distribution component, wherein R is a positive integer less than N.

[0011] Furthermore, the step of merging the N original distribution components in the Gaussian mixture probability model based on a preset component reduction algorithm includes: determining the number of reduced components m, where m is a positive integer and m is equal to NR; setting i equal to 1, and for the N original distribution components, calculating the first type divergence distance between each of the original distribution components and other distribution components, determining the minimum distance among all the first type divergence distances, merging the two original distribution components corresponding to the minimum distance to obtain an updated distribution component, and setting i equal to i+1; performing iterative merging, the step of iterative merging includes: when i is less than or equal to m, for all the updated distribution components and the original distribution components, calculating the i+1th type divergence distance between each distribution component and other distribution components, merging the two distribution components corresponding to the minimum distance among all the i+1th type divergence distances to obtain the updated distribution component, and setting i equal to i+1; or, when i is greater than m, stopping the iterative merging.

[0012] Furthermore, the step of performing an operation risk assessment on the target distribution network based on the real-time charging power data and the Gaussian mixture probability model to obtain an assessment result includes: inputting the real-time charging power data into the Gaussian mixture probability model, and having the model output the node voltage change of each of the grid nodes; for each of the grid nodes, calculating the voltage over-limit probability of the grid node based on the safety voltage threshold and the node voltage change, and obtaining N voltage over-limit probabilities; performing an operation risk assessment on the target distribution network based on the N voltage over-limit probabilities, and obtaining the assessment result, wherein the assessment result at least includes: the overall voltage over-limit probability of the target distribution network.

[0013] According to another aspect of an embodiment of the present invention, there is also provided an evaluation device for the operation risk of a distribution network, comprising: an acquisition unit, for acquiring charging power data of N electric vehicles within a target time period, wherein the charging power data refers to the node power injection data of the grid node to which the electric vehicle is connected, and N is a positive integer; a first establishment unit, for performing node power flow calculation based on the charging power data, and establishing a probabilistic power flow expression of the N electric vehicles, wherein the probabilistic power flow expression obeys a Gaussian distribution and is used to describe the correlation change between the real-time voltage of the grid node and the power injection disturbance, and the power injection disturbance refers to the node power injection data related to the random charging behavior of the electric vehicle; a second establishment unit, for establishing a Gaussian mixture distribution body based on the power flow calculation result and the probabilistic power flow expression, and obtaining a Gaussian mixture probability model, wherein the Gaussian mixture probability model is used to monitor the real-time voltage disturbance of any of the random charging behaviors on the relevant grid nodes; an evaluation unit, for acquiring the real-time charging power data of the N electric vehicles, and performing an operation risk evaluation on the target distribution network based on the real-time charging power data and the Gaussian mixture probability model to obtain an evaluation result.

[0014] Furthermore, the first establishment unit includes: an establishment module, which is used to establish the power flow equation at each power grid node based on the active power, reactive power and node voltage in the node power injection data; a linearization processing module, which is used to determine the reference operating point for each power flow equation, and linearize the power flow equation at the reference operating point to obtain a linearized expression of the power flow equation, wherein the reference operating point refers to a steady-state operating point at which the system load and power generation are balanced; a format deformation module, which is used to treat the power injection disturbance as a random variable, format-deform the linearized expression, and obtain the power flow calculation result in which the node voltage changes in real time with the power injection disturbance.

[0015] Furthermore, the first establishment unit also includes: a quantification module, which is used to quantify each random charging behavior for each of the electric vehicles to obtain R charging behavior parameters, where R is a positive integer; a first fitting module, which is used to use each of the charging behavior parameters as a first distribution component, perform Gaussian distribution fitting on the charging power data, and determine the first component parameters during the fitting process, wherein the first component parameters include at least: a mixing weight, a mean, and a standard deviation; a first determination module, which is used to determine the probability flow expression of the power injection disturbance of each of the electric vehicles based on the first component parameters and a preset Gaussian distribution expression to obtain N probability flow expressions.

[0016] Furthermore, the second establishing unit includes: a second fitting module, used to take each of the electric vehicles as a second distribution component, perform Gaussian mixture distribution fitting based on the power flow calculation results and the probabilistic power flow expression, and determine the second component parameters during the fitting process; a second determination module, used to determine the analytical expression of the Gaussian mixture probability model based on the second component parameters and a preset Gaussian distribution expression to obtain the Gaussian mixture probability model.

[0017] Furthermore, the distribution network operation risk assessment device also includes: a merging module, used to merge the N original distribution components in the Gaussian mixture probability model based on a preset component reduction algorithm to obtain a lightweight model of the Gaussian mixture probability model; a third determination module, used to redetermine the component parameters in the Gaussian distribution expression of each of the R updated distribution components in the lightweight model, wherein R is a positive integer less than N.

[0018] Further, the merging module includes: a determination submodule, used to determine the number of reduced components m, where m is a positive integer and m is equal to NR; a first merging submodule, used to set i equal to 1, for the N original distribution components, calculate the first type of divergence distance between each of the original distribution components and other distribution components, determine the minimum distance among all the first type of divergence distances, merge the two original distribution components corresponding to the minimum distance to obtain an updated distribution component, and set i equal to i+1; a second merging submodule, used to perform iterative merging, the iterative merging step includes: when i is less than or equal to m, for all the updated distribution components and the original distribution components, calculate the i+1th type of divergence distance between each distribution component and other distribution components, merge the two distribution components corresponding to the minimum distance among all the i+1th type of divergence distances to obtain the updated distribution component, and set i equal to i+1; a stopping submodule, used to stop performing the iterative merging when i is greater than m.

[0019] Furthermore, the evaluation unit includes: an input module, used to input the real-time charging power data into the Gaussian mixture probability model, and the model outputs the node voltage change of each of the grid nodes; a calculation module, used to calculate the voltage over-limit probability of the grid node for each of the grid nodes based on the safety voltage threshold and the node voltage change, and obtain N voltage over-limit probabilities; an evaluation module, used to perform an operation risk assessment on the target distribution network based on the N voltage over-limit probabilities, and obtain the assessment result, wherein the assessment result at least includes: the overall voltage over-limit probability of the target distribution network.

[0020] According to another aspect of an embodiment of the present invention, a computer-readable storage medium is further provided, wherein the computer-readable storage medium includes a stored computer program, wherein when the computer program is running, the device where the computer-readable storage medium is located is controlled to execute any one of the above-mentioned methods for assessing distribution network operation risks.

[0021] According to another aspect of an embodiment of the present invention, there is also provided an electronic device, comprising one or more processors and a memory, wherein the memory is used to store one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement any one of the above-mentioned methods for assessing distribution network operation risks.

[0022] In the present invention, a method for evaluating the operation risk of a distribution network is proposed. First, the charging power data of N electric vehicles in a target time period are obtained, wherein the charging power data refers to the node power injection data of the power grid node to which the electric vehicle is connected, and N is a positive integer. Then, the node flow calculation is performed based on the charging power data, and a probabilistic flow expression of the N electric vehicles is established, wherein the probabilistic flow expression obeys a Gaussian distribution and is used to describe the correlation change between the real-time voltage of the power grid node and the power injection disturbance, and the power injection disturbance refers to the node power injection data related to the random charging behavior of the electric vehicle. Then, based on the flow calculation result and the probabilistic flow expression, a Gaussian mixture distribution body is established to obtain a Gaussian mixture probability model, wherein the Gaussian mixture probability model is used to monitor the real-time voltage disturbance of any random charging behavior on the relevant power grid node. Finally, the real-time charging power data of the N electric vehicles are obtained, and the operation risk of the target distribution network is evaluated based on the real-time charging power data and the Gaussian mixture probability model to obtain an evaluation result.

[0023] In the present invention, charging power data of N electric vehicles in a specific time period are collected to reflect the power injection situation when the electric vehicles are connected to the grid node. The charging behavior of the electric vehicles is closely related to the operation state of the grid through node flow calculation. A probability flow expression obeying Gaussian distribution is constructed to accurately describe the dynamic change relationship between the real-time voltage of the grid node and the charging power disturbance. Further, these probability flow expressions are integrated into a Gaussian mixture distribution body to form a Gaussian mixture probability model, which can effectively monitor and predict the disturbance effect of random charging behavior on the grid node voltage. The randomness of the electric vehicle charging power is modeled using a Gaussian mixture model. The node voltage and the power injection disturbance are directly related through an analytical expression, avoiding the complex calculation process of a large number of Monte Carlo simulations in the traditional method. The real-time charging power data is combined with the optimized GMM model to quickly evaluate the operation risk of large-scale electric vehicle access to the distribution network, which can significantly reduce the calculation complexity and maintain the accuracy of the risk assessment. This solves the technical problem that the calculation complexity of the distribution network operation risk assessment when electric vehicles are connected to the grid on a large scale in the related technology is high, resulting in low evaluation efficiency and accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of this application. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0025] Figure 1 is a flow chart of an optional method for evaluating distribution network operation risk according to an embodiment of the present invention;

[0026] Figure 2 is a flow chart of an optional probabilistic power flow analysis method based on an improved Gaussian mixture model according to an embodiment of the present invention;

[0027] Figure 3 is a schematic diagram of an optional distribution network operation risk assessment device according to an embodiment of the present invention;

[0028] Figure 4 The present invention is a hardware structure block diagram of an electronic device (or mobile device) for a method for evaluating the operation risk of a distribution network according to an embodiment of the present invention. DETAILED DESCRIPTION

[0029] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0030] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0031] To facilitate those skilled in the art to understand the present invention, some terms or nouns involved in the embodiments of the present invention are explained below:

[0032] Improved Gaussian Mixture Model (GMM) is a probabilistic statistical model used to describe complex data distributions composed of multiple Gaussian distributions. In the present invention, GMM is improved to more effectively fit the randomness of electric vehicle charging power, and the accurate modeling of the optimized probability distribution of electric vehicle charging power is achieved by adjusting the number and parameters of Gaussian components in the model.

[0033] Probabilistic Power Flow, PPF, probability flow analysis, in the power system, when the input parameters (such as load, power output) are random, the method of analyzing the probability distribution of system states (such as node voltage, line power flow), PPF can evaluate the stability of the power system under different operating conditions, especially when dealing with uncertain sources (such as electric vehicle charging demand), it can predict and quantify potential voltage limit and line overload risks. In the present invention, it is used to evaluate the impact of the optimized electric vehicle charging power distribution on the distribution network.

[0034] Gaussian Mixture Model Reduction Algorithm, GMM-R, Gaussian mixture model simplification algorithm, an optimization algorithm used to simplify the number of components of the Gaussian mixture model (GMM) while retaining the accuracy of the original model as much as possible. It merges multiple Gaussian components into fewer components through an iterative process and uses KL divergence to measure information loss to ensure a balance between accuracy and complexity after model simplification.

[0035] Kullback-Leibler Divergence, KL Divergence, Kullback-Leibler Divergence, a statistic that measures the difference between two probability distributions, describes the information loss from one distribution to another, and is usually used to evaluate the degree of model fit or the effect of information compression. In the present invention, KL divergence is used as a metric in the GMM-R algorithm to quantify the information loss in the process of merging Gaussian components, with the goal of minimizing information loss, ensuring that the model can still accurately characterize the randomness and uncertainty of electric vehicle charging power after simplification, thereby providing a reliable data basis for the analysis of power system operation risks.

[0036] The following embodiments of the present invention can be applied to various power systems / applications / equipment that need to deal with large-scale electric vehicle charging load uncertainty and perform real-time or predictive risk assessment, and can implement an electric vehicle charging load risk assessment system based on a Gaussian mixture probability model.

[0037] Specifically, it can be applied to smart grid management systems. In smart grids, the following embodiments can be embedded in the grid management system to help monitor and predict in real time the power system operation risks caused by electric vehicle charging behaviors, such as voltage over-limit, line overload, etc. By analyzing the probability distribution of electric vehicle charging power, it can respond quickly, adopt demand response strategies or adjust the grid operation mode to maintain the stability and safety of the power system.

[0038] The following format example of the present invention can also be applied to the electric vehicle charging station operation platform. The charging station operator can optimize the charging schedule through the method provided by the present invention to reduce the instantaneous impact on the power grid, improve the operating efficiency of the charging station and the acceptance capacity of the power grid. At the same time, the platform can provide users with charging suggestions based on the risk assessment results of the power system, such as charging during the low electricity price period, to reduce the user's charging cost and the operation risk of the power grid.

[0039] The following format example of the present invention can also be applied to distributed energy management systems. In a power system that includes a large number of distributed energy sources (such as solar energy, wind energy) and electric vehicles, the method provided by the present invention can evaluate and optimize the overall energy supply and demand balance, ensure the stable operation of the power system, and maximize the utilization rate of renewable energy and the charging efficiency of electric vehicles.

[0040] The following format examples of the present invention can also be applied to power market trading scenarios. The power market can design a more sophisticated trading mechanism based on the method provided by the present invention, and guide electric vehicle users to charge during off-peak hours by adjusting electricity prices in real time, thereby reducing the load pressure on the power grid and improving the flexibility and efficiency of the power market.

[0041] The present invention is described in detail below in conjunction with various embodiments.

[0042] Embodiment 1

[0043] According to an embodiment of the present invention, an embodiment of a method for evaluating the operating risks of a distribution network is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0044] The implementation subject of the embodiment of the present invention can be a smart grid management system, an electric vehicle charging station operation platform or a distributed energy management system, as well as various power systems, applications and equipment that need to deal with the uncertainty of large-scale electric vehicle charging loads and perform real-time or predictive risk assessment. By combining the improved Gaussian mixture model (GMM) and probabilistic power flow (PPF) analysis technology, by adopting the charging power distribution optimized by GMM modeling, establishing a probabilistic power flow analysis framework, and proposing the component reduction algorithm GMM-R, etc., effective assessment and management of the operating risks of electric vehicles when they are connected to the power grid on a large scale are achieved.

[0045] In the specific implementation process, the embodiment of the present invention optimizes the GMM model through the GMM-R algorithm to quickly calculate the risk consequences of large-scale electric vehicles connecting to the power grid, provide real-time decision support for distribution network operators, and avoid the problem of low calculation efficiency in traditional methods.

[0046] The method of the embodiment of the present invention can be combined with real-time charging power data to dynamically adjust the charging strategy of electric vehicles to reduce potential risks to the power grid, and realize intelligent coordination and optimization between the power system and the charging needs of electric vehicles. It has high flexibility and versatility for both the transformation of existing power grids and the construction of future smart grids. It is not only suitable for electric vehicles, but can also be extended to uncertainty modeling and risk assessment scenarios of other distributed energy sources and loads.

[0047] The embodiments of the present invention are described in detail below in conjunction with specific implementation steps.

[0048] Figure 1 is a flow chart of an optional method for evaluating the operation risk of a distribution network according to an embodiment of the present invention, such as Figure 1 As shown, the method comprises the following steps:

[0049] Step S101, obtaining charging power data of N electric vehicles in a target time period, wherein the charging power data refers to node power injection data of a power grid node to which the electric vehicle is connected, and N is a positive integer.

[0050] Specifically, step S101 is essentially to quantify the power injection into the node (i.e., the charging location of the electric vehicle) in the power system. The charging power data refers to the power demand of each electric vehicle for its grid node within a certain time period. This demand reflects the uncertainty of the charging behavior of electric vehicles. That is, since the charging time and charging power of electric vehicles can vary with factors such as user behavior, battery status, and electricity price strategy, collecting and analyzing these data is crucial to understanding how electric vehicle charging affects the operation of the power grid.

[0051] Step S102, performing node power flow calculation based on charging power data, and establishing a probabilistic power flow expression for N electric vehicles, wherein the probabilistic power flow expression obeys a Gaussian distribution and is used to describe the correlation change between the real-time voltage of the grid node and the power injection disturbance, and the power injection disturbance refers to the node power injection data related to the random charging behavior of the electric vehicle.

[0052] It should be noted that power flow calculation is the basis of power system analysis, which is used to determine the steady-state operation state of the power grid under given load conditions. The voltage, phase angle and power flow of each node in the power grid are calculated by solving the nonlinear power flow equations, so as to understand the operation characteristics of the power system. In the embodiment of the present invention, the power flow calculation of N electric vehicle access nodes not only takes into account the power demand of electric vehicle charging, but also evaluates the potential impact of these demands on the grid node voltage and line power flow.

[0053] Specifically, the power flow equation can be described as w=f(x), where w represents the active and reactive power injection of the node, x represents the node voltage angle and amplitude, and f(·) represents the power flow equation.

[0054] Optionally, the step of performing node power flow calculation based on charging power data in step S102 includes: for each grid node, establishing a power flow equation at the grid node based on the active power, reactive power and node voltage in the node power injection data; for each power flow equation, determining a reference operating point, and linearizing the power flow equation at the reference operating point to obtain a linearized expression of the power flow equation, wherein the reference operating point refers to a steady-state operating point at which the system load and power generation are balanced; treating the power injection disturbance as a random variable, formatting the linearized expression, and obtaining a power flow calculation result in which the node voltage changes in real time with the power injection disturbance.

[0055] It should be noted that the reference operating point refers to the steady-state operating point where the system load and power generation are balanced under normal operation of the power system. In this state, the system operating parameters (such as node voltage and line power flow) are stable and meet all operating constraints. In the embodiment of the present invention, determining the reference operating point is the starting point of the power flow calculation, which provides a reference condition for the subsequent linearization processing, making the sensitivity analysis of uncertain factors (such as electric vehicle charging power) more reliable.

[0056] Linearizing the power flow equation at the reference operating point is essentially a local approximation of a complex nonlinear system to simplify the calculation process. The linearization step of the embodiment of the present invention can be selected to perform Taylor series expansion at the reference operating point, retain the first-order terms, ignore the second-order and higher-order nonlinear terms, and obtain a linearized expression of the power flow equation.

[0057] Specifically, the concept of Jacobian matrix is ​​introduced, and the power flow equation can be simplified into a linear equation through the Jacobian matrix: Δw = -JΔx, where J is the Jacobian coefficient matrix, Δw represents the random power injection disturbance of the node, and Δx represents the corresponding change of the node voltage.

[0058] Furthermore, the uncertainty quantity (i.e., power injection disturbance) associated with the random charging behavior of electric vehicles, which causes the power injection of power grid nodes to deviate from the reference operating point, is taken as a random variable. It is necessary to use the sensitivity matrix to format the linearized expression and link the random disturbance of electric vehicle charging power with the node voltage change.

[0059] Specifically, the power flow equation after the format deformation is: Δx=(-J) -1 Δw=SΔw, where S is the sensitivity matrix, which is the inverse matrix of the Jacobian matrix.

[0060] Considering the randomness and uncertainty of the charging demand of electric vehicles, the embodiment of the present invention uses GMM to model the probability distribution of its charging power. Optionally, the step of establishing the probability flow expression of N electric vehicles in step S102 includes: for each electric vehicle, quantifying each random charging behavior to obtain R charging behavior parameters, where R is a positive integer; taking each charging behavior parameter as the first distribution component, performing Gaussian distribution fitting on the charging power data, and determining the first component parameter during the fitting process, wherein the first component parameter at least includes: mixing weight, mean and standard deviation; determining the probability flow expression of the power injection disturbance of each electric vehicle based on the first component parameter and the preset Gaussian distribution expression, and obtaining N probability flow expressions. GMM can accurately capture complex distributions, including multi-peak and asymmetric characteristics, thereby more accurately reflecting the actual impact of electric vehicle charging behavior on the power system.

[0061] It should be noted that charging behavior parameters refer to statistical parameters directly related to the charging behavior of electric vehicles, including but not limited to charging start time, charging duration, charging power, etc., which are used to reflect the characteristics of different charging behaviors, that is, the randomness and uncertainty of electric vehicles charging at different time points and charging power. By quantifying R charging behavior parameters, the multi-dimensional characteristics of electric vehicle charging demand can be captured, providing a richer data foundation for subsequent probability distribution modeling.

[0062] Furthermore, each charging behavior parameter is regarded as an independent first distribution component, which is a random variable obeying a Gaussian distribution. The first component parameters refer to the statistical parameters determined in each first distribution component of the GMM, including the mixing weight, mean, and standard deviation. The mixing weight reflects the relative importance of each Gaussian component in the population, and the mean and standard deviation describe the central position of the component and the degree of discreteness of the data, respectively.

[0063] It should be noted that establishing the probabilistic power flow expression of N electric vehicles, following the rules of Gaussian distribution, is a key step in combining the power flow calculation results with the randomness of electric vehicle charging power. In the power system, due to the randomness of load and other factors, parameters such as node voltage and line power flow will also fluctuate over time. The probabilistic power flow expression can establish a direct connection between the uncertainty of electric vehicle charging power and the uncertainty of grid node voltage state, thereby providing a quantitative basis for risk assessment. Through the probabilistic power flow expression, the statistical relationship between node voltage and power injection disturbance can be described, that is, the probability of a certain voltage state in the presence of uncertainty, and then the grid operation risk caused by electric vehicle charging behavior can be predicted and controlled, providing guarantee for the safe, reliable and efficient operation of the power system.

[0064] Step S103, a Gaussian mixture distribution is established based on the power flow calculation result and the probability power flow expression to obtain a Gaussian mixture probability model, wherein the Gaussian mixture probability model is used to monitor the real-time voltage disturbance of any random charging behavior on the relevant power grid nodes.

[0065] Optionally, step S103 includes: taking each electric vehicle as a second distribution component, performing Gaussian mixture distribution fitting based on the power flow calculation results and the probability power flow expression, and determining the second component parameters during the fitting process; determining the analytical expression of the Gaussian mixture probability model based on the second component parameters and a preset Gaussian distribution expression to obtain the Gaussian mixture probability model.

[0066] It should be noted that each electric vehicle is regarded as the second distribution component in the Gaussian mixture model, corresponding to the impact of the random disturbance of the electric vehicle charging power on the grid node voltage. The second component parameters refer to the mixing weight, mean and standard deviation determined in the process of fitting the Gaussian mixture distribution for each electric vehicle as the second distribution component. They are used to describe the statistical characteristics of each electric vehicle in the Gaussian mixture probability model. The mixing weight represents the proportion of the electric vehicle's contribution to the voltage disturbance in the population, and the mean and standard deviation reflect the average value and fluctuation degree of the voltage disturbance respectively.

[0067] The Gaussian mixture probability model is a composite probability distribution composed of multiple Gaussian distributions, where each Gaussian distribution corresponds to a probability flow expression of an electric vehicle. The establishment of the Gaussian mixture probability model can monitor the real-time voltage disturbance of any random charging behavior on the relevant grid nodes, providing an effective means of quantitatively evaluating the operation risk of the power system. The analytical expression of the Gaussian mixture probability model is a mathematical form derived from the second component parameter and the preset Gaussian distribution expression. It supports the direct calculation of the contribution of each electric vehicle charging behavior to the grid node voltage disturbance, as well as the probability distribution of the grid node voltage disturbance under the joint action of all electric vehicles, avoiding complex numerical simulation processes and greatly improving the computational efficiency and accuracy of risk assessment.

[0068] The above step S103 not only converts the charging behavior of each electric vehicle into the second distribution component independently, but also integrates these components into a unified probability model through Gaussian mixture distribution fitting, which can monitor and predict the random changes of node voltage when large-scale electric vehicles are connected to the power grid in real time.

[0069] In order to solve the computational complexity problem caused by large-scale electric vehicle access, an embodiment of the present invention innovatively proposes a GMM-R algorithm, which measures and reduces the number of Gaussian components in the GMM through KL divergence, while maintaining the statistical properties and accuracy of the model, significantly improving the computational efficiency and model scalability under large-scale data processing. Optionally, after executing step S103, it also includes: merging the N original distribution components in the Gaussian mixture probability model based on a preset component reduction algorithm to obtain a lightweight model of the Gaussian mixture probability model; for the R updated distribution components in the lightweight model, re-determining the component parameters in the Gaussian distribution expression of each updated distribution component, where R is a positive integer less than N.

[0070] Optionally, the step of merging N original distribution components in the Gaussian mixture probability model based on a preset component reduction algorithm includes: determining the number of reduced components m, where m is a positive integer and m is equal to NR; setting i equal to 1, and for the N original distribution components, calculating the first type divergence distance between each original distribution component and other distribution components, determining the minimum distance among all first type divergence distances, merging the two original distribution components corresponding to the minimum distance to obtain an updated distribution component, and setting i equal to i+1; performing iterative merging, the step of iterative merging includes: when i is less than or equal to m, for all updated distribution components and original distribution components, calculating the i+1th type divergence distance between each distribution component and other distribution components, merging the two distribution components corresponding to the minimum distance among all i+1th type divergence distances to obtain an updated distribution component, and setting i equal to i+1; or, when i is greater than m, stopping the iterative merging.

[0071] Step S104, obtaining real-time charging power data of N electric vehicles, and performing an operation risk assessment on the target distribution network based on the real-time charging power data and a Gaussian mixture probability model to obtain an assessment result.

[0072] Optionally, step S104 includes: inputting real-time charging power data into a Gaussian mixture probability model, and having the model output the node voltage change of each grid node; for each grid node, calculating the voltage over-limit probability of the grid node based on the safety voltage threshold and the node voltage change, and obtaining N voltage over-limit probabilities; performing an operation risk assessment on the target distribution network based on the N voltage over-limit probabilities, and obtaining an assessment result, wherein the assessment result at least includes: the overall voltage over-limit probability of the target distribution network.

[0073] It should be noted that real-time charging power data refers to the actual measured value of each electric vehicle's charging power at a specific moment or time period. It is based on real-time monitoring and recording of electric vehicle charging behavior, reflecting the power injection status of the grid node at the current moment. Inputting real-time charging power data into the Gaussian mixture probability model can instantly evaluate the real-time impact of electric vehicle charging on the grid node voltage and provide instant feedback for the safe monitoring of power system operation. When the real-time charging power data is input into the Gaussian mixture probability model, the model can calculate the node voltage change of each grid node based on the previously established second distribution component parameters and analytical expressions.

[0074] The safety voltage threshold is the upper and lower limit standards set for node voltage during power system operation to ensure that the voltage level is within the safe range for equipment and users. It is used to define voltage over-limit situations. When the node voltage changes beyond this threshold, the power system operation will face problems such as equipment damage, power outages or reduced power quality.

[0075] Based on the safety voltage threshold and the node voltage change of each grid node, the voltage over-limit probability refers to the probability that the grid node voltage exceeds the safety threshold under given charging conditions, which is used to quantify the grid operation risk caused by electric vehicle charging. The operation risk assessment based on N voltage over-limit probabilities refers to the comprehensive consideration of the voltage over-limit risks of all grid nodes and the quantitative analysis of the overall operation risk of the target distribution network.

[0076] Through the above steps S101 and S104, the charging power data of N electric vehicles in the target time period can be obtained first, wherein the charging power data refers to the node power injection data of the power grid node to which the electric vehicle is connected, and N is a positive integer. Then, the node flow calculation is performed based on the charging power data, and a probabilistic flow expression of the N electric vehicles is established, wherein the probabilistic flow expression obeys a Gaussian distribution and is used to describe the correlation change between the real-time voltage of the power grid node and the power injection disturbance. The power injection disturbance refers to the node power injection data related to the random charging behavior of the electric vehicle. Then, based on the flow calculation result and the probabilistic flow expression, a Gaussian mixture distribution body is established to obtain a Gaussian mixture probability model, wherein the Gaussian mixture probability model is used to monitor the real-time voltage disturbance of any random charging behavior on the relevant power grid nodes. Finally, the real-time charging power data of the N electric vehicles are obtained, and the operation risk assessment of the target distribution network is performed based on the real-time charging power data and the Gaussian mixture probability model to obtain an assessment result.

[0077] In an embodiment of the present invention, charging power data of N electric vehicles in a specific time period are collected to reflect the power injection situation when the electric vehicles are connected to the grid node. The charging behavior of the electric vehicles is closely associated with the operation state of the grid through node power flow calculation. A probability power flow expression obeying Gaussian distribution is constructed to accurately describe the dynamic change relationship between the real-time voltage of the grid node and the charging power disturbance. Optionally, these probability power flow expressions are integrated into a Gaussian mixture distribution body to form a Gaussian mixture probability model, which can effectively monitor and predict the disturbance effect of random charging behavior on the grid node voltage. The randomness of the electric vehicle charging power is modeled using a Gaussian mixture model. The node voltage and the power injection disturbance are directly associated through an analytical expression, avoiding the complex calculation process of a large number of Monte Carlo simulations in the traditional method. The real-time charging power data is combined with the optimized GMM model to quickly evaluate the operation risk of large-scale electric vehicle access to the distribution network, which can significantly reduce the calculation complexity while maintaining the accuracy of the risk assessment, thereby solving the technical problem that the calculation complexity of the distribution network operation risk assessment when electric vehicles are connected to the grid on a large scale in the related technology is high, resulting in low evaluation efficiency and accuracy.

[0078] The present invention is described below in conjunction with another specific implementation manner.

[0079] The implementation mode of the present invention focuses on the potential risks of randomness and uncertainty of large-scale access of electric vehicles to the power system, evaluates the risk consequences of the charging power of electric vehicles optimized by means such as demand response, proposes a probabilistic power flow analysis method based on an improved Gaussian mixture model (GMM), and calculates the operating risk on this basis, aiming to reduce the complexity of probabilistic power flow calculation when large-scale electric vehicles are connected and retain a high calculation accuracy.

[0080] Figure 2 is a flowchart of an optional probabilistic power flow analysis method based on an improved Gaussian mixture model according to an embodiment of the present invention, such as Figure 2 As shown in the figure, the method includes the following processes: optimizing the probability distribution of electric vehicle charging power; establishing the power flow equation; using the Gaussian mixture model to model the optimized random charging power of electric vehicles; giving the analytical expression of the probability power flow; reducing the complexity of the probability power flow calculation; and risk analysis. The following will introduce them one by one.

[0081] 1. Optimize the probability distribution of electric vehicle charging power.

[0082] Distribution network operators can adjust the probability distribution of charging power of electric vehicle users through demand response and other means. For example, distribution network operators release subsidized electricity prices, and electric vehicle users adjust their charging behavior with the goal of optimal economic cost, thereby achieving the purpose of probability distribution of charging power.

[0083] 2. Establish the tidal flow equation.

[0084] The power flow equation can be described as: w = f(x), where w represents the active and reactive power injection of the node, x represents the node voltage angle and amplitude, and f(·) represents the power flow equation.

[0085] The embodiment of the present invention can perform Taylor series expansion on the power flow equation at the reference working point and ignore the second-order and higher-order terms. For example, it can be simplified to: Δw = -JΔx, Δx = (-J) -1 Δw=SΔw, where J is the Jacobian matrix of the last iteration of the power flow calculation, S is the sensitivity matrix (i.e., the inverse matrix of the Jacobian matrix), Δw represents the random power injection disturbance of the node, and Δx represents the corresponding change of the node voltage.

[0086] It should be noted that considering the randomness of EV charging demand, Δw is regarded as a random variable.

[0087] 3. The optimized random charging power of electric vehicles is modeled using Gaussian mixture model.

[0088] Optimization control measures such as demand response can change the power probability distribution of electric vehicles. The Gaussian mixture model is used to model the random power injection of electric vehicles. The formula is: Among them, i represents the i-th node, K i Denotes Δw i The number of components of the Gaussian mixture model (GMM), π ik / μ ik / σ ik Respectively represent Δw i The mixture weight, mean, and standard deviation of the kth component of .

[0089] The GMM in the embodiment of the present invention can accurately fit complex data distribution (eg, multi-peak and arbitrarily shaped distribution) by adjusting the number and parameters of Gaussian components.

[0090] 4. Give the analytical expression of the probabilistic power flow.

[0091] By modeling power injection through GMM, the probability power flow can be modeled as a weighted mixture of Gaussian distributions, each with a different mean and variance. The charging power of each electric vehicle access node is modeled as a Gaussian mixture model, and an analytical expression for the probability power flow can be obtained.

[0092] 5. Reduce the complexity of probabilistic power flow calculation.

[0093] The excessive number of components in the Gaussian mixture model makes the model computationally intensive, which may lead to scalability issues. The embodiments of the present invention propose a component reduction algorithm to reduce the computational complexity of the GMM-based PPF analysis while maintaining accuracy.

[0094] Specifically, the Kulbeck-Leibler (KL) divergence is used to measure the distance between two Gaussian distributions. The original components of the GMM are assigned to the nearest reduced components through iterative optimization, and the parameters (weights, means, and variances) of these reduced components are updated. In each iteration, the KL divergence is used as a measure to minimize information loss, ensuring the effectiveness of the assignment and parameter update process, reducing the complexity of the model while retaining the distribution characteristics of the original data. Therefore, the simplified GMM is not only more computationally efficient, but also can accurately capture the intrinsic structure of the data while maintaining its statistical properties.

[0095] 6. Risk analysis.

[0096] Regarding random risk analysis, existing studies have considered a variety of risk criteria, including low voltage, line and transformer overload, voltage quality deterioration, etc. Among them, the low voltage criterion has become an important focus due to the heavy load on the power system caused by large-scale electric vehicle access and strict requirements under low voltage conditions. In the implementation of the present invention, the voltage change of node i is obtained by probabilistic flow calculation as Δx. Combined with the safety threshold Δx of the node voltage, the probability of the node voltage exceeding the limit can be further calculated as P(Δx>Δx), which is used to characterize the operation risk.

[0097] Compared with the prior art, the embodiment of the present invention proposes a probabilistic power flow analysis method based on an improved Gaussian mixture model (GMM), which can more accurately capture the randomness and uncertainty characteristics of large-scale electric vehicle charging demand. By using GMM to model the optimized electric vehicle charging power, not only the analytical expression of the probabilistic power flow is obtained, which significantly reduces the calculation complexity, but also the GMM-R algorithm is used to achieve efficient reduction of model components while maintaining high calculation accuracy. This breakthrough technology solves the contradiction between computational efficiency and accuracy that the existing PPF method may face when dealing with large-scale electric vehicle integration, and provides a powerful tool for risk assessment of power systems in real-time application scenarios.

[0098] The embodiments of the present invention can also instantly calculate the voltage changes of each grid node and the probability of voltage exceeding the limit based on real-time charging power data and a Gaussian mixture probability model, and then provide an overall operation risk assessment of the target distribution network. It can also respond to random changes in electric vehicle charging behavior in real time and provide dynamic risk monitoring. By quantifying the consequences of risks, such as the possibility of low voltage events and line overloads, power system operators can take more precise control measures and optimize the operation strategy of the power network, effectively improving the reliability and safety of the power system and realizing real-time monitoring and intelligent management of large-scale electric vehicle charging risks.

[0099] The present invention is described below in conjunction with another optional embodiment.

[0100] Embodiment 2

[0101] The distribution network operation risk assessment device provided in this embodiment includes multiple implementation units, each implementation unit corresponding to each implementation step in the above-mentioned embodiment 1.

[0102] Figure 3 is a schematic diagram of an optional distribution network operation risk assessment device according to an embodiment of the present invention, such as Figure 3 As shown, the device may include: an acquisition unit 31 , a first establishment unit 32 , a second establishment unit 33 , and an evaluation unit 34 .

[0103] The acquisition unit 31 is used to acquire charging power data of N electric vehicles in a target time period, wherein the charging power data refers to node power injection data of a power grid node to which the electric vehicle is connected, and N is a positive integer.

[0104] The first establishing unit 32 is used to perform node power flow calculation based on charging power data and establish a probabilistic power flow expression for N electric vehicles, wherein the probabilistic power flow expression obeys a Gaussian distribution and is used to describe the correlation change between the real-time voltage of the grid node and the power injection disturbance, and the power injection disturbance refers to the node power injection data related to the random charging behavior of the electric vehicle.

[0105] The second establishing unit 33 is used to establish a Gaussian mixture distribution body based on the power flow calculation result and the probability power flow expression to obtain a Gaussian mixture probability model, wherein the Gaussian mixture probability model is used to monitor the real-time voltage disturbance of any random charging behavior on the relevant power grid nodes.

[0106] The evaluation unit 34 is used to obtain the real-time charging power data of N electric vehicles, and perform an operation risk evaluation on the target distribution network based on the real-time charging power data and a Gaussian mixture probability model to obtain an evaluation result.

[0107] The above-mentioned distribution network operation risk assessment device can first obtain the charging power data of N electric vehicles in the target time period through the acquisition unit 31, wherein the charging power data refers to the node power injection data of the power grid node to which the electric vehicle is connected, and N is a positive integer, and then the node flow calculation is performed based on the charging power data through the first establishment unit 32, and a probability flow expression of the N electric vehicles is established, wherein the probability flow expression obeys the Gaussian distribution and is used to describe the correlation change between the real-time voltage of the power grid node and the power injection disturbance, and the power injection disturbance refers to the node power injection data related to the random charging behavior of the electric vehicle, and then a Gaussian mixture distribution body is established based on the flow calculation result and the probability flow expression through the second establishment unit 33 to obtain a Gaussian mixture probability model, wherein the Gaussian mixture probability model is used to monitor the real-time voltage disturbance of any random charging behavior on the relevant power grid node, and finally the real-time charging power data of the N electric vehicles is obtained through the evaluation unit 34, and the operation risk of the target distribution network is evaluated based on the real-time charging power data and the Gaussian mixture probability model to obtain an evaluation result.

[0108] In an embodiment of the present invention, charging power data of N electric vehicles in a specific time period are collected to reflect the power injection situation when the electric vehicles are connected to the grid node. The charging behavior of the electric vehicles is closely associated with the grid operation state through node power flow calculation. A probability power flow expression obeying Gaussian distribution is constructed to accurately describe the dynamic change relationship between the real-time voltage of the grid node and the charging power disturbance. Optionally, these probability power flow expressions are integrated into a Gaussian mixture distribution body to form a Gaussian mixture probability model, which can effectively monitor and predict the disturbance effect of random charging behavior on the grid node voltage. The Gaussian mixture model is used to model the randomness of the electric vehicle charging power. The node voltage and the power injection disturbance are directly associated through an analytical expression, avoiding the complex calculation process of a large number of Monte Carlo simulations in the traditional method. The real-time charging power data is combined with the optimized GMM model to quickly evaluate the operation risk of large-scale electric vehicle access to the distribution network, which can significantly reduce the calculation complexity while maintaining the accuracy of the risk assessment, thereby solving the technical problem of high calculation complexity of the distribution network operation risk assessment when electric vehicles are connected to the grid on a large scale in the related technology, resulting in low evaluation efficiency and accuracy.

[0109] Optionally, the first establishment unit includes: an establishment module, which is used to establish the power flow equation at each power grid node based on the active power, reactive power and node voltage in the node power injection data; a linearization processing module, which is used to determine the reference operating point for each power flow equation, and linearize the power flow equation at the reference operating point to obtain a linearized expression of the power flow equation, wherein the reference operating point refers to a steady-state operating point at which the system load and power generation are balanced; a format deformation module, which is used to treat the power injection disturbance as a random variable, format-deform the linearized expression, and obtain a power flow calculation result in which the node voltage changes in real time with the power injection disturbance.

[0110] Optionally, the first establishment unit also includes: a quantification module, which is used to quantify each random charging behavior for each electric vehicle to obtain R charging behavior parameters, where R is a positive integer; a first fitting module, which is used to use each charging behavior parameter as a first distribution component, perform Gaussian distribution fitting on the charging power data, and determine the first component parameters during the fitting process, wherein the first component parameters include at least: mixing weight, mean and standard deviation; a first determination module, which is used to determine the probability flow expression of the power injection disturbance of each electric vehicle based on the first component parameters and a preset Gaussian distribution expression, and obtain N probability flow expressions.

[0111] Optionally, the second establishment unit includes: a second fitting module, used to take each electric vehicle as a second distribution component, perform Gaussian mixture distribution fitting based on the power flow calculation results and the probability power flow expression, and determine the second component parameters during the fitting process; a second determination module, used to determine the analytical expression of the Gaussian mixture probability model based on the second component parameters and a preset Gaussian distribution expression to obtain the Gaussian mixture probability model.

[0112] Optionally, the distribution network operation risk assessment device also includes: a merging module, used to merge N original distribution components in the Gaussian mixture probability model based on a preset component reduction algorithm to obtain a lightweight model of the Gaussian mixture probability model; a third determination module, used to redetermine the component parameters in the Gaussian distribution expression of each updated distribution component for the R updated distribution components in the lightweight model, where R is a positive integer less than N.

[0113] Optionally, the merging module includes: a determination submodule, used to determine the number of reduced components m, where m is a positive integer and m is equal to NR; a first merging submodule, used to set i equal to 1, and for N original distribution components, calculate the first type of divergence distance between each original distribution component and other distribution components, determine the minimum distance among all first type of divergence distances, merge the two original distribution components corresponding to the minimum distance to obtain an updated distribution component, and set i equal to i+1; a second merging submodule, used to perform iterative merging, the iterative merging step including: when i is less than or equal to m, for all updated distribution components and original distribution components, calculate the i+1th type of divergence distance between each distribution component and other distribution components, merge the two distribution components corresponding to the minimum distance among all i+1th type of divergence distances to obtain an updated distribution component, and set i equal to i+1; a stopping submodule, used to stop performing iterative merging when i is greater than m.

[0114] Optionally, the evaluation unit includes: an input module, used to input real-time charging power data into a Gaussian mixture probability model, and the model outputs the node voltage change of each grid node; a calculation module, used to calculate the voltage over-limit probability of each grid node based on the safety voltage threshold and the node voltage change, and obtain N voltage over-limit probabilities; an evaluation module, used to perform an operation risk assessment on the target distribution network based on the N voltage over-limit probabilities, and obtain an evaluation result, wherein the evaluation result at least includes: the overall voltage over-limit probability of the target distribution network.

[0115] The above-mentioned distribution network operation risk assessment device may also include a processor and a memory. The above-mentioned acquisition unit 31, the first establishment unit 32, the second establishment unit 33, the assessment unit 34, etc. are all stored in the memory as program units, and the processor executes the above-mentioned program units stored in the memory to realize corresponding functions.

[0116] The processor includes a kernel, which retrieves the corresponding program unit from the memory. One or more kernels can be set, and a Gaussian mixture distribution body is established based on the power flow calculation results and the probability power flow expression by adjusting the kernel parameters to obtain a Gaussian mixture probability model, wherein the Gaussian mixture probability model is used to monitor the real-time voltage disturbance of any random charging behavior on the relevant power grid nodes, obtain the real-time charging power data of N electric vehicles, and perform an operation risk assessment on the target distribution network based on the real-time charging power data and the Gaussian mixture probability model.

[0117] The above-mentioned memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM, and the memory includes at least one storage chip.

[0118] The present application also provides a computer program product, which, when executed on a data processing device, is suitable for executing a program that is initialized with the following method steps: obtaining charging power data of N electric vehicles within a target time period, wherein the charging power data refers to the node power injection data of the power grid node to which the electric vehicle is connected, and N is a positive integer; performing node flow calculation based on the charging power data, and establishing a probabilistic flow expression for the N electric vehicles, wherein the probabilistic flow expression obeys a Gaussian distribution and is used to describe the correlation change between the real-time voltage of the power grid node and the power injection disturbance, and the power injection disturbance refers to the node power injection data related to the random charging behavior of the electric vehicle; establishing a Gaussian mixture distribution body based on the flow calculation result and the probabilistic flow expression to obtain a Gaussian mixture probability model, wherein the Gaussian mixture probability model is used to monitor the real-time voltage disturbance of any random charging behavior on the relevant power grid node; obtaining the real-time charging power data of N electric vehicles, and performing an operation risk assessment on the target distribution network based on the real-time charging power data and the Gaussian mixture probability model to obtain an assessment result.

[0119] According to another aspect of an embodiment of the present invention, a computer-readable storage medium is further provided, the computer-readable storage medium including a stored computer program, wherein when the computer program is running, the device where the computer-readable storage medium is located is controlled to execute the distribution network operation risk assessment method of any one of the above-mentioned embodiments one.

[0120] According to another aspect of an embodiment of the present invention, there is also provided an electronic device, comprising one or more processors and a memory, the memory being used to store one or more programs, wherein when the one or more programs are executed by one or more processors, the one or more processors implement the method for assessing the operation risk of a distribution network of any one of the above-mentioned embodiments one.

[0121] Figure 4 1 is a hardware structure block diagram of an electronic device (or mobile device) for a method for evaluating the operation risk of a distribution network according to an embodiment of the present invention. Figure 4 As shown, the electronic device may include one or more processors ( Figure 4 The processor 402a, processor 402b, ..., processor 402n are used to illustrate, and the processor may include but is not limited to a processing device such as a microprocessor MCU or a programmable logic device FPGA), a memory 404 for storing data. In addition, it may also include: a display, an input / output interface (I / O interface), a universal serial bus (USB) port (which may be included as one of the ports of the I / O interface), a network interface, a keyboard, a power supply and / or a camera. A person skilled in the art can understand that Figure 4The structure shown is only for illustration and does not limit the structure of the above electronic device. Figure 4 More or fewer components as shown, or with Figure 4 Different configurations are shown.

[0122] The serial numbers of the above embodiments of the present invention are only for description and do not represent the advantages or disadvantages of the embodiments.

[0123] In the above embodiments of the present invention, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0124] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. Among them, the device embodiments described above are only schematic. For example, the division of the units can be a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of units or modules, which can be electrical or other forms.

[0125] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple units. Some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0126] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0127] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, a server or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, magnetic disk or optical disk and other media that can store program codes.

[0128] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A method for evaluating the operation risk of a distribution network, characterized in that: include: Obtaining charging power data of N electric vehicles within a target time period, wherein the charging power data refers to node power injection data of a power grid node to which the electric vehicles are connected, and N is a positive integer; Perform node power flow calculation based on the charging power data, and establish a probability power flow expression of the N electric vehicles, wherein the probability power flow expression obeys a Gaussian distribution and is used to describe the correlation change between the real-time voltage of the grid node and the power injection disturbance, and the power injection disturbance refers to the node power injection data related to the random charging behavior of the electric vehicle; A Gaussian mixture distribution is established based on the power flow calculation result and the probability power flow expression to obtain a Gaussian mixture probability model, wherein the Gaussian mixture probability model is used to monitor the real-time voltage disturbance of any of the random charging behaviors on the relevant power grid nodes; Real-time charging power data of the N electric vehicles are obtained, and an operation risk assessment is performed on a target distribution network based on the real-time charging power data and the Gaussian mixture probability model to obtain an assessment result.

2. The evaluation method according to claim 1, characterized in that: The step of performing node power flow calculation based on the charging power data includes: For each of the power grid nodes, a power flow equation at the power grid node is established based on the active power, reactive power and node voltage in the node power injection data; For each of the power flow equations, a reference operating point is determined, and the power flow equation is linearized at the reference operating point to obtain a linearized expression of the power flow equation, wherein the reference operating point refers to a steady-state operating point at which system load and power generation are balanced; The power injection disturbance is taken as a random variable, and the linearized expression is formatted to obtain the power flow calculation result in which the node voltage changes in real time with the power injection disturbance.

3. The evaluation method according to claim 2, characterized in that: The step of establishing a probabilistic power flow expression of N electric vehicles comprises: For each of the electric vehicles, each random charging behavior is quantified to obtain R charging behavior parameters, where R is a positive integer; Taking each of the charging behavior parameters as a first distribution component, performing Gaussian distribution fitting on the charging power data, and determining first component parameters during the fitting process, wherein the first component parameters include at least: a mixing weight, a mean, and a standard deviation; The probability power flow expression of the power injection disturbance of each electric vehicle is determined based on the first component parameter and a preset Gaussian distribution expression, and N probability power flow expressions are obtained.

4. The evaluation method according to claim 1, characterized in that: The step of establishing a Gaussian mixture distribution body based on the power flow calculation result and the probabilistic power flow expression to obtain a Gaussian mixture probability model includes: Taking each of the electric vehicles as a second distribution component, performing Gaussian mixture distribution fitting based on the power flow calculation result and the probabilistic power flow expression, and determining the second component parameters during the fitting process; An analytical expression of the Gaussian mixture probability model is determined based on the second component parameter and a preset Gaussian distribution expression to obtain the Gaussian mixture probability model.

5. The evaluation method according to claim 1, characterized in that: After establishing a Gaussian mixture distribution body based on the power flow calculation result and the probability power flow expression to obtain a Gaussian mixture probability model, the method further includes: Based on a preset component reduction algorithm, N original distribution components in the Gaussian mixture probability model are merged to obtain a lightweight model of the Gaussian mixture probability model; For the R updated distribution components in the lightweight model, the component parameters in the Gaussian distribution expression of each updated distribution component are re-determined, where R is a positive integer less than N.

6. The evaluation method according to claim 1, characterized in that: The step of merging N original distribution components in the Gaussian mixture probability model based on a preset component reduction algorithm comprises: Determine the number of reductions m, where m is a positive integer and m is equal to NR; Let i be equal to 1, for the N original distribution components, calculate the first type divergence distance between each of the original distribution components and other distribution components, determine the minimum distance among all the first type divergence distances, merge the two original distribution components corresponding to the minimum distance to obtain an updated distribution component, and let i be equal to i+1; Perform iterative merging, the iterative merging step comprising: when i is less than or equal to m, for all the updated distribution components and the original distribution components, calculating the i+1th type of divergence distance between each distribution component and other distribution components, merging the two distribution components corresponding to the minimum distance among all the i+1th type of divergence distances to obtain the updated distribution component, and setting i equal to i+1; or, When i is greater than m, the iterative merging is stopped.

7. The evaluation method according to claim 1, characterized in that: The step of performing an operation risk assessment on a target distribution network based on the real-time charging power data and the Gaussian mixture probability model to obtain an assessment result comprises: Inputting the real-time charging power data into the Gaussian mixture probability model, and having the model output a node voltage change of each of the grid nodes; For each of the grid nodes, based on the safety voltage threshold and the node voltage change, the voltage over-limit probability of the grid node is calculated to obtain N voltage over-limit probabilities; An operation risk assessment is performed on the target distribution network based on the N voltage over-limit probabilities to obtain the assessment result, wherein the assessment result at least includes: an overall voltage over-limit probability of the target distribution network.

8. A distribution network operation risk assessment device, characterized in that: include: An acquisition unit, used to acquire charging power data of N electric vehicles within a target time period, wherein the charging power data refers to node power injection data of a power grid node to which the electric vehicle is connected, and N is a positive integer; A first establishing unit is used to perform node power flow calculation based on the charging power data, and establish a probability power flow expression of N electric vehicles, wherein the probability power flow expression obeys a Gaussian distribution and is used to describe the correlation change between the real-time voltage of the grid node and the power injection disturbance, and the power injection disturbance refers to the node power injection data related to the random charging behavior of the electric vehicle; A second establishing unit is used to establish a Gaussian mixture distribution body based on the power flow calculation result and the probability power flow expression to obtain a Gaussian mixture probability model, wherein the Gaussian mixture probability model is used to monitor the real-time voltage disturbance of any of the random charging behaviors on the relevant power grid nodes; An evaluation unit is used to obtain real-time charging power data of N electric vehicles, and to perform an operation risk evaluation on a target distribution network based on the real-time charging power data and the Gaussian mixture probability model to obtain an evaluation result.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes a stored computer program, wherein when the computer program is executed, the device where the computer-readable storage medium is located is controlled to execute the method for evaluating the operation risk of the distribution network according to any one of claims 1 to 7.

10. An electronic device, characterized in that: It includes one or more processors and a memory, wherein the memory is used to store one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement the distribution network operation risk assessment method as described in any one of claims 1 to 7.

Images