A power system evaluation method and system based on remote confirmation failure of a knife switch

Abstract

The application discloses a kind of power system evaluation method and system based on knife gate remote confirmation failure, belong to power system evaluation technical field.The application obtains the probability value of knife gate remote confirmation failure and constructs several operating scenarios according to this, so that the risk assessment of power system operation can consider the topological misjudgment risk caused by the interruption of knife gate position signal transmission, false alarm or confirmation timeout;Then, by constructing the probability density model of each operating index and mapping to the standard normal space to determine the equivalent correlation coefficient, the complex correlation structure actually existing between wind power output, photovoltaic output and load can be accurately described;On this basis, based on the standard normal operating index and equivalent correlation coefficient, the linearized power flow calculation is carried out to obtain the operating state variable probability density function, and then the conditional risk value under each scenario is obtained by combining the conditional risk assessment, and the comprehensive evaluation result is obtained based on the conditional risk value of each scenario, which improves the accuracy of power system operation risk assessment.

Description

Technical Field

[0001] This invention belongs to the field of power system assessment technology, and in particular relates to a power system assessment method and system based on remote confirmation of failure of disconnectors. Background Technology

[0002] The operation of power systems is becoming increasingly complex due to the large-scale integration of distributed power sources. The intermittency and volatility of wind and solar power output, coupled with the spatiotemporal uncertainty of electric vehicle charging loads, cause the power system's operating state to deviate from the deterministic operating mode of traditional single-source radial networks. Therefore, assessing the operational risks of the power system, thereby providing dispatching and operation personnel with a global awareness of the grid's security situation, is an important way to ensure the safe and stable operation of the power system.

[0003] Existing power system operation risk assessment methods primarily focus on conventional uncertainties such as wind and solar power output fluctuations and load randomness. They employ scenario simulation or analytical methods for probabilistic power flow calculations, combined with expected value-based risk indicators to achieve operational risk assessment. However, because disconnectors in power systems are prone to remote confirmation failures due to interruptions in location signal transmission, false alarms, or timeouts, the actual power system topology changes, altering the power flow distribution and introducing risks. Furthermore, current power flow calculations mostly rely on simple orthogonal transformations or independent assumptions, failing to accurately reflect the complex interrelationships between various factors such as wind power output, solar power output, and loads within the power system, thus affecting the accuracy of operational risk assessment. In addition, current expected value-based risk indicators struggle to accurately perceive the actual threats posed by extreme risks in the power system, leading to biases in operational risk assessments. Summary of the Invention

[0004] The present invention aims to provide a power system assessment method and system based on remote confirmation of failure of disconnectors, so as to solve the above-mentioned technical problems and improve the accuracy of power system operation risk assessment.

[0005] To address the aforementioned technical problems, embodiments of the present invention provide a power system assessment method based on remote confirmation of disconnector failure, comprising: Obtain the probability value of remotely confirmed failure of disconnectors in the power system and the operating data of each operating indicator; construct several operating scenarios of the power system based on the probability value of remotely confirmed failure of disconnectors; Based on the operational data, a probability density model is constructed for each of the operational indicators within the power system; The probability density models of various operational indicators are mapped to a preset standard normal space to determine the standard normal operational indicator corresponding to each operational indicator and the equivalent correlation coefficient between each pair of operational indicators. Based on the standard normal operating index and equivalent correlation coefficient, linearized power flow calculation is performed on the power system to obtain the probability density function of several operating state variables of the power system in each operating scenario; Based on the probability density function of the operating state variables, the power system is subjected to conditional risk assessment to obtain several conditional risk values ​​for the power system in each operating scenario; Based on several conditional risk values ​​of the power system in each of the aforementioned operating scenarios, the operational risk assessment results of the power system are obtained.

[0006] Understandably, compared to existing technologies, this invention obtains the probability value of remote confirmation failure of disconnectors and constructs several operating scenarios accordingly. This allows power system operation risk assessment to account for topology misjudgment risks caused by disconnector position signal transmission interruptions, false alarms, or confirmation timeouts. Next, by constructing probability density models of various operating indicators and mapping them to a standard normal space to determine equivalent correlation coefficients, it accurately characterizes the complex correlation structure actually existing between wind power output, photovoltaic output, and loads, avoiding the correlation distortion problem caused by traditional methods using independent assumptions or simple orthogonal transformations. Based on this, linearized power flow calculations are performed using standard normal operating indicators and equivalent correlation coefficients to obtain the probability density function of operating state variables. This is then combined with conditional risk assessment to obtain conditional risk values ​​for each scenario. Conditional risk assessment effectively overcomes the inherent defect of traditional expected value-based risk indicators being insensitive to extreme events. Finally, a comprehensive assessment result is obtained based on the conditional risk values ​​of each scenario, improving the accuracy of power system operation risk assessment.

[0007] As a preferred embodiment, the operating indicators include: distributed wind power output, distributed photovoltaic power output, electric vehicle charging power, and load forecasting error; the operating data for distributed wind power output includes: historical wind speed and real-time wind speed; the operating data for distributed photovoltaic power output includes: solar irradiance data and maximum distributed photovoltaic power output; the operating data for electric vehicle charging power includes: historical charging power; the operating data for load forecasting error includes: load forecasting error value; and the construction of a probability density model for each of the operating indicators within the power system based on the operating data includes: The historical wind speed is estimated by maximum likelihood to obtain the first shape parameter and the first scale parameter; Based on the real-time wind speed, the first shape parameter, and the first scale parameter, a two-parameter Weibull distribution model is performed on the distributed wind power output to obtain the probability density model of the distributed wind power output. Calculate the mean and variance of the light intensity data, and determine the second shape parameter and the third shape parameter based on the mean and variance of the light intensity data; Based on the maximum output of the distributed photovoltaic system, the second shape parameter, and the third shape parameter, a beta distribution model of the distributed photovoltaic power is performed to obtain the probability density model of the distributed photovoltaic power. The historical charging power is fitted based on a preset kernel density estimation method or a Gaussian mixture model to obtain a probability density model of the electric vehicle charging power. The load prediction error value is fitted with a normal distribution to obtain the probability density model of the load prediction error.

[0008] In the above scheme, considering the physical characteristics of distributed wind power output, the first shape parameter and the first scale parameter are obtained by maximum likelihood estimation of historical wind speed. Combined with real-time wind speed, a two-parameter Weibull distribution is used for modeling, which accurately characterizes the skewed distribution characteristics of wind speed and its nonlinear mapping relationship with wind power output. This provides a wind power output probability density model that conforms to actual physical laws for subsequent probabilistic power flow calculations. Regarding the statistical characteristics of distributed photovoltaic (PV) power output, the second shape parameter and the third shape parameter are determined by calculating the mean and variance of irradiance data. Combined with the maximum output of distributed PV, a beta distribution is used for modeling, which accurately describes the probability distribution of irradiance within a finite interval, overcoming the limitations of traditional methods. This paper addresses the unreasonable tail probability problem that arises when fitting bounded physical quantities with a state distribution. For the complex distribution characteristics of electric vehicle charging power, kernel density estimation or Gaussian mixture models are used to fit historical charging power, forming a complementary combination that can adapt to the distribution modeling needs under different charging behavior patterns. For load forecasting errors, a normal distribution with a mean of zero is used for fitting, fully utilizing the statistical law that load forecasting errors are symmetrically distributed around the predicted value and have a low probability of large errors. By applying probability density modeling methods that match the physical statistical characteristics of four different types of operating indicators, accurate basic data support is provided for power system operation risk assessment, improving the accuracy of power system operation risk assessment.

[0009] As a preferred embodiment, mapping the probability density models of various operational indicators to a preset standard normal space to determine the standard normal operational indicator corresponding to each operational indicator and the equivalent correlation coefficient between any two operational indicators includes: Integrating the probability density models of the various operational indicators yields the cumulative distribution functions of the various operational indicators. Based on the preset Natav transform algorithm, the cumulative distribution function is mapped to the preset standard normal space to obtain the standard normal operating index for each of the operating indices; Based on a preset correlation analysis algorithm, the correlation of the operational data of various operational indicators is calculated to obtain the correlation coefficient between each pair of operational indicators. Based on the standard normal operating index of each of the operating indicators, the correlation coefficient between each pair of operating indicators is equivalently transformed to obtain the equivalent correlation coefficient between each pair of operating indicators in the standard normal space.

[0010] In the above scheme, the cumulative distribution function is obtained by integrating the probability density model of each operating indicator. Then, based on the Natav transform algorithm, the cumulative distribution function is mapped to the standard normal space. This can accurately transform the original random variables with arbitrary distribution forms into standard normal variables while completely preserving all the probability information of the original variables, avoiding the information loss problem caused by traditional normalization methods that only use the mean and variance for transformation. On this basis, the correlation of the operating data of each operating indicator is calculated based on the preset correlation analysis algorithm to obtain the correlation coefficient between each operating indicator in the original space. Then, the correlation coefficient is equivalently transformed based on the standard normal operating indicators to obtain the equivalent correlation coefficient in the standard normal space. Thus, while transforming non-normally correlated operating indicators into standard normal operating indicators, the probability distribution information and correlation structure of the original variables are completely preserved. This can accurately depict the complex correlation structure that actually exists between wind power output, photovoltaic power output and load, avoiding the correlation distortion problem caused by traditional methods using independence assumptions or simple orthogonal transformations, and further improving the accuracy of power system operation risk assessment.

[0011] As a preferred embodiment, the linearized power flow calculation of the power system based on the standard normal operating index and the equivalent correlation coefficient yields a probability density function of several operating state variables of the power system in each operating scenario, including: The standard normal operating index and equivalent correlation coefficient are recursively derived using the preset cumulative quantity generation function method to obtain the semi-invariants of various operating indices at each preset order. By performing an inverse Natav transformation on the semi-invariants of various operating indicators at each order, the semi-invariants of the nodal injection power of the power system at each order are obtained. Linearized power flow calculations are performed on the semi-invariants of the node injected power at each order to obtain the node voltage probability density function, branch power flow probability density function, and load shedding probability density function of the power system in each operating scenario. Based on the node voltage probability density function, branch power flow probability density function, and load shedding probability density function, several operating state variable probability density functions of the power system are formed for each operating scenario.

[0012] In the above scheme, the cumulative generation function method is used to recursively derive the standard normal operating indicators and equivalent correlation coefficients to obtain the semi-invariants of each operating indicator at each preset order. As a high-order statistical characteristic of random variables, the semi-invariants can completely characterize the skewness, kurtosis, and other distribution patterns of non-normally distributed variables, and their homogeneous additivity makes the propagation calculation under linear transformation extremely efficient. Then, the inverse Natav transformation is performed on the semi-invariants of each operating indicator at each order to obtain the semi-invariants of the power system node injection power at each order, ensuring that the statistical characteristics of the node injection power are completely transmitted to the power flow calculation stage. Finally, linearized power flow calculation is performed on the semi-invariants of the node injection power at each order to obtain the node voltage probability density function, branch power flow probability density function, and load shedding probability density function under each operating scenario, forming a set of probability density functions covering the entire power grid operating state. By combining the semi-invariant method with linearized power flow calculation, accurate and complete operating state probability distribution information is provided for subsequent conditional risk assessment, further improving the accuracy of power system operating risk assessment.

[0013] As a preferred embodiment, the linearized power flow calculation of the injected power at each order of semi-invariants is performed to obtain the node voltage probability density function, branch power flow probability density function, and load shedding probability density function of the power system in each operating scenario, including: Based on the preset linearized power flow equations, the semi-invariants of the node injection power at each order are mapped to obtain the semi-invariants of the node voltage and the branch power flow at each order in the power system. The mean and standard deviation of the node voltage in each operating scenario are determined based on the semi-invariants of the node voltage at each order within the power system. The node voltage mean and standard deviation for each operating scenario are fitted with a probability density function to obtain the node voltage probability density function of the power system for each operating scenario. The mean and standard deviation of the branch power flow in each operating scenario are determined based on the semi-invariants of the branch power flow at each order, and the mean and standard deviation of the branch power flow in each operating scenario are fitted based on the preset series expansion method to obtain the branch power flow probability density function of the power system in each operating scenario. The load shedding power data of the power system in each operating scenario is obtained, and the load shedding power data is fitted according to a preset kernel density estimation algorithm to obtain the load loss probability density function of the power system in each operating scenario.

[0014] In the above scheme, by pre-setting linearized power flow equations, the semi-invariants of node injected power at each order are mapped to the semi-invariants of node voltage and branch power flow at each order. This accurately transfers the statistical characteristics of the semi-invariants of node injected power at each order to node voltage and branch power flow. After determining the mean and standard deviation of the semi-invariants of node voltage at each order, probability density fitting is performed to obtain the node voltage probability density function, which can retain the complete statistical information carried by the semi-invariants. After determining the mean and standard deviation of the semi-invariants of branch power flow at each order, a pre-set series expansion method is used to fit the branch power flow probability density function. This can reconstruct non-normal probability density functions of arbitrary shapes under the condition that only the semi-invariants of each order are known, thus balancing fitting accuracy and computational simplicity. The kernel density estimation algorithm is used to fit the load shedding power data. It can accurately characterize the actual probability distribution characteristics of the load shedding without pre-setting any distribution form, ensuring the fitting accuracy of the probability density functions of various operating state variables. This provides a reliable input for the accurate calculation of subsequent conditional risk values ​​and improves the accuracy of power system operation risk assessment.

[0015] As a preferred embodiment, the conditional risk assessment of the power system based on the probability density function of the operating state variables yields several conditional risk values ​​for the power system in each operating scenario, including: Based on the node voltage probability density function of the power system in each operating scenario, and combined with the preset node voltage limit, a voltage limit exceedance severity variable of the power system in each operating scenario is constructed. Based on the voltage exceedance severity variable, a voltage exceedance risk assessment is performed on the power system to obtain the voltage exceedance risk value of the power system in each operating scenario; Based on the branch power flow probability density function of the power system in each operating scenario, and combined with the preset maximum allowable power flow value of the branch, the branch overload severity variable of the power system in each operating scenario is constructed. Based on the branch overload severity variable, the branch overload condition risk assessment of the power system is performed to obtain the branch overload risk value of the power system in each operating scenario. The load loss risk value of the power system in each operating scenario is determined based on the load loss probability density function of the power system in each operating scenario.

[0016] In the above scheme, a voltage exceedance severity variable is constructed based on the node voltage probability density function and a preset node voltage limit. This transforms the node voltage of the power system into a single voltage exceedance severity variable, allowing the probability distribution of the severity variable to centrally reflect the degree of loss when the power system is in a voltage unsafe state. A branch overload severity variable is constructed based on the branch power flow probability density function and a preset maximum allowable power flow value, accurately reflecting the risk level of branch power flow overload in the power system. Furthermore, conditional risk assessment is performed based on the voltage exceedance severity variable, the branch overload severity variable, and the load shedding probability density function, respectively. Conditional risk assessment can calculate the expected value of tail losses for random variables, effectively amplifying the tail contribution of the severity variable under extreme fault scenarios. This avoids the inherent deficiency of traditional expected value indicators that overwhelm tail risks in high-probability safe scenarios, thus improving the accuracy of power system operation risk assessment.

[0017] As a preferred embodiment, the step of obtaining the operational risk assessment result of the power system based on several conditional risk values ​​of the power system in each operational scenario includes: Based on the probability value of the remote confirmation failure of the disconnect switch, the voltage over-limit risk value of the power system in each of the operating scenarios is weighted and fused to obtain the comprehensive voltage over-limit risk value of the power system. Based on the probability value of the remote confirmation failure of the disconnect switch, the branch overload risk value of the power system in each of the operating scenarios is weighted and fused to obtain the comprehensive branch overload risk value of the power system. Based on the probability value of the remote confirmation of failure of the disconnect switch, the load loss risk value of the power system in each of the operating scenarios is weighted and fused to obtain the comprehensive load loss risk value of the power system. The weighting coefficients of the voltage over-limit comprehensive risk value, branch overload comprehensive risk value, and load shedding comprehensive risk value are obtained, and the comprehensive risk value of voltage over-limit comprehensive risk value, branch overload comprehensive risk value, and load shedding comprehensive risk value are weighted and summed to obtain the comprehensive operation risk value of the power system; The comprehensive operational risk value is compared with the preset risk threshold to obtain the operational risk assessment result of the power system.

[0018] In the above scheme, based on the probability value of remotely confirmed failure of the disconnector, the risk values ​​of voltage over-limit, branch overload, and load shedding are weighted and fused to obtain the comprehensive risk values ​​of voltage over-limit, branch overload, and load shedding, respectively. The probability value of remotely confirmed failure of the disconnector is directly embedded as a risk quantification factor into the calculation process of the comprehensive risk value through a weighted fusion mechanism, so that the remotely confirmed failure of the disconnector can be transformed into a quantifiable risk contribution. Then, the weight coefficients of the comprehensive risk values ​​of voltage over-limit, branch overload, and load shedding are obtained and weighted summed to obtain the comprehensive operational risk value. The weight allocation can adaptively reflect the relative importance of each risk type under different operating scenarios. The comprehensive operational risk value is compared with the preset risk threshold to obtain the operational risk assessment result, which improves the accuracy of power system operational risk assessment.

[0019] Accordingly, this invention provides a power system assessment system based on remote confirmation of disconnector failure, comprising: a power system data acquisition module, a probability density model construction module, a standard normal space mapping module, a linearized power flow calculation module, a conditional risk assessment module, and a power system operation risk assessment module; The power system data acquisition module is used to acquire the probability value of remotely confirmed failure of disconnectors in the power system and the operating data of each operating indicator; and to construct several operating scenarios of the power system based on the probability value of remotely confirmed failure of disconnectors. The probability density model building module is used to build a probability density model for each of the operating indicators in the power system based on the operating data; The standard normal space mapping module is used to map the probability density models of various operating indicators to a preset standard normal space, so as to determine the standard normal operating indicator corresponding to each operating indicator and the equivalent correlation coefficient between each pair of operating indicators; The linearized power flow calculation module is used to perform linearized power flow calculation on the power system based on the standard normal operating index and the equivalent correlation coefficient, and obtain the probability density function of several operating state variables of the power system in each operating scenario. The conditional risk assessment module is used to perform conditional risk assessment on the power system based on the probability density function of the operating state variables, and to obtain several conditional risk values ​​of the power system in each operating scenario. The power system operation risk assessment module is used to obtain the operation risk assessment result of the power system based on several conditional risk values ​​of the power system in each operation scenario.

[0020] As a preferred embodiment, the probability density model construction module includes: a probability density model construction unit; In the probability density model construction unit, the operating indicators include: distributed wind power output, distributed photovoltaic power output, electric vehicle charging power, and load forecasting error; the operating data of distributed wind power output includes: historical wind speed and real-time wind speed; the operating data of distributed photovoltaic power output includes: solar irradiance data and maximum distributed photovoltaic power output; the operating data of electric vehicle charging power includes: historical charging power; and the operating data of load forecasting error includes: load forecasting error value. The probability density model building unit is used to perform maximum likelihood estimation of the historical wind speed to obtain the first shape parameter and the first scale parameter; Based on the real-time wind speed, the first shape parameter, and the first scale parameter, a two-parameter Weibull distribution model is performed on the distributed wind power output to obtain the probability density model of the distributed wind power output. Calculate the mean and variance of the light intensity data, and determine the second shape parameter and the third shape parameter based on the mean and variance of the light intensity data; Based on the maximum output of the distributed photovoltaic system, the second shape parameter, and the third shape parameter, a beta distribution model of the distributed photovoltaic power is performed to obtain the probability density model of the distributed photovoltaic power. The historical charging power is fitted based on a preset kernel density estimation method or a Gaussian mixture model to obtain a probability density model of the electric vehicle charging power. The load prediction error value is fitted with a normal distribution to obtain the probability density model of the load prediction error.

[0021] As a preferred embodiment, the standard normal space mapping module includes: a standard normal space mapping unit; The standard normal space mapping unit is used to integrate the probability density models of various operating indicators to obtain the cumulative distribution functions of various operating indicators. Based on the preset Natav transform algorithm, the cumulative distribution function is mapped to the preset standard normal space to obtain the standard normal operating index for each of the operating indices; Based on a preset correlation analysis algorithm, the correlation of the operational data of various operational indicators is calculated to obtain the correlation coefficient between each pair of operational indicators. Based on the standard normal operating index of each of the operating indicators, the correlation coefficient between each pair of operating indicators is equivalently transformed to obtain the equivalent correlation coefficient between each pair of operating indicators in the standard normal space.

[0022] Understandably, compared to existing technologies, this system, by acquiring the probability value of remote confirmation failure of disconnectors and constructing several operational scenarios accordingly, enables power system operation risk assessment to account for topology misjudgment risks caused by disconnector position signal transmission interruptions, false alarms, or confirmation timeouts. Next, by constructing probability density models of various operational indicators and mapping them to a standard normal space to determine equivalent correlation coefficients, it accurately characterizes the complex correlation structure actually existing between wind power output, photovoltaic output, and loads, avoiding the correlation distortion problem caused by traditional methods using independent assumptions or simple orthogonal transformations. Based on this, linearized power flow calculations are performed using standard normal operational indicators and equivalent correlation coefficients to obtain the probability density function of operational state variables. This is then combined with conditional risk assessment to obtain conditional risk values ​​for each scenario. Conditional risk assessment effectively overcomes the inherent defect of traditional expected value-based risk indicators being insensitive to extreme events. Finally, a comprehensive assessment result is obtained based on the conditional risk values ​​of each scenario, improving the accuracy of power system operation risk assessment. Attached Figure Description

[0023] Figure 1 A flowchart illustrating the steps of a power system assessment method based on remote confirmation of disconnector failure provided in this embodiment of the invention; Figure 2 This is a schematic diagram of the structure of a power system assessment system based on remote confirmation of failure of a disconnector, provided as an embodiment of the present invention. Detailed Implementation

[0024] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0025] Example 1 Please refer to Figure 1 , Figure 1 The flowchart of a power system assessment method based on remote confirmation of failure of disconnector provided in an embodiment of the present invention includes steps S101 to S106.

[0026] Step S101: Obtain the probability value of remote confirmation failure of disconnectors in the power system and the operating data of each operating indicator; construct several operating scenarios of the power system based on the probability value of remote confirmation failure of disconnectors.

[0027] Step S102: Construct a probability density model for each of the operating indicators in the power system based on the operating data.

[0028] Step S103: Map the probability density models of various operating indicators to a preset standard normal space to determine the standard normal operating indicator corresponding to each operating indicator and the equivalent correlation coefficient between each pair of operating indicators.

[0029] Step S104: Based on the standard normal operating index and the equivalent correlation coefficient, perform linearized power flow calculation on the power system to obtain the probability density function of several operating state variables of the power system in each operating scenario.

[0030] Step S105: Perform conditional risk assessment on the power system based on the probability density function of the operating state variables to obtain several conditional risk values ​​of the power system in each operating scenario.

[0031] Step S106: Based on several conditional risk values ​​of the power system in each of the operating scenarios, obtain the operating risk assessment result of the power system.

[0032] This embodiment obtains the probability value of remote confirmation failure of disconnectors and constructs several operating scenarios accordingly, enabling power system operation risk assessment to account for topology misjudgment risks caused by disconnector position signal transmission interruptions, false alarms, or confirmation timeouts. Next, by constructing probability density models of various operating indicators and mapping them to a standard normal space to determine equivalent correlation coefficients, it accurately depicts the complex correlation structure actually existing between wind power output, photovoltaic output, and loads, avoiding correlation distortion problems caused by traditional methods using independent assumptions or simple orthogonal transformations. Based on this, linearized power flow calculations are performed using standard normal operating indicators and equivalent correlation coefficients to obtain the probability density function of operating state variables. This is then combined with conditional risk assessment to obtain conditional risk values ​​for each scenario. Conditional risk assessment effectively overcomes the inherent insensitivity of traditional expected value-based risk indicators to extreme events. Finally, a comprehensive assessment result is obtained based on the conditional risk values ​​of each scenario, improving the accuracy of power system operation risk assessment.

[0033] In one optional embodiment, the probability value of remote confirmation failure of disconnectors within the power system and the operational data of each operational indicator are obtained; wherein, the operational indicators include: distributed wind power output, distributed photovoltaic power output, electric vehicle charging power, and load prediction error; the operational data of distributed wind power output includes: historical wind speed and real-time wind speed; the operational data of distributed photovoltaic power output includes: solar irradiance data and maximum distributed photovoltaic power output; Before calculating the probability density model of the operational indicators, the operational data for each indicator needs to be preprocessed; firstly, for the... The operational data of various operational indicators, using Criteria for removing outliers The outlier determination formula of the criterion is: ;in, For the first A single parameter data point for a certain operational indicator. The mean of the parameters, The standard deviation is given. Next, the data is normalized using the Min-Max Normalization algorithm. The normalization formula is as follows: ; This represents the minimum value of the running data; This indicates the maximum value of the running data; This represents the normalized running data; finally, this embodiment uses mean imputation to complete the data, and the mean of the mean imputation method is... The calculation formula is as follows, based on the valid data within the sliding window: ; The data within the window is non-abnormal. This represents the number of valid data points.

[0034] The above preprocessing process can be implemented using other algorithms according to actual needs, and this embodiment does not impose too many limitations here.

[0035] The probability value of remote confirmation of failure of the switch is defined as follows: It refers to the knife switch having The probability of remote confirmation failure occurs; therefore, this embodiment constructs two operating scenarios for the power system based on this, including: a normal operating scenario and a remote confirmation failure scenario for disconnectors; the normal operating scenario is denoted as... Its probability of occurrence is denoted as The scenario where the switch fails remotely is recorded as follows: Its probability of occurrence is denoted as .

[0036] In this embodiment, the operating indicators include: distributed wind power output, distributed photovoltaic power output, electric vehicle charging power, and load forecasting error; the operating data for distributed wind power output includes: historical wind speed and real-time wind speed; the operating data for distributed photovoltaic power output includes: solar irradiance data and maximum distributed photovoltaic power output; the operating data for electric vehicle charging power includes: historical charging power; the operating data for load forecasting error includes: load forecasting error value; the step of constructing a probability density model for each of the operating indicators within the power system based on the operating data includes: The historical wind speed is estimated by maximum likelihood to obtain the first shape parameter and the first scale parameter; Based on the real-time wind speed, the first shape parameter, and the first scale parameter, a two-parameter Weibull distribution model is performed on the distributed wind power output to obtain the probability density model of the distributed wind power output. Calculate the mean and variance of the light intensity data, and determine the second shape parameter and the third shape parameter based on the mean and variance of the light intensity data; Based on the maximum output of the distributed photovoltaic system, the second shape parameter, and the third shape parameter, a beta distribution model of the distributed photovoltaic power is performed to obtain the probability density model of the distributed photovoltaic power. The historical charging power is fitted based on a preset kernel density estimation method or a Gaussian mixture model to obtain a probability density model of the electric vehicle charging power. The load prediction error value is fitted with a normal distribution to obtain the probability density model of the load prediction error.

[0037] In this embodiment, for distributed wind power output, the historical wind speed is subjected to maximum likelihood estimation (MLE) to obtain the first shape parameter. and first scale parameters Then, a two-parameter Weibull distribution is used to describe the randomness of wind speed-induced distributed wind power output. Therefore, the probability density model of distributed wind power output is expressed as: ; in, Real-time wind speed; Calculate the mean and variance of the light intensity data, and use the mean of the light intensity data as the second shape parameter. The variance of the light intensity data is used as the third shape parameter. The maximum output of distributed photovoltaic power is defined as follows: Therefore, this embodiment uses beta distribution modeling to obtain the probability density model of the distributed photovoltaic power output, which is expressed as: ; in, This is the gamma function.

[0038] For electric vehicle charging power, since kernel density estimation does not require a predefined distribution form and can flexibly characterize the multimodal or skewed characteristics of the actual charging power distribution, the probability density model of electric vehicle charging power described using kernel density estimation (KDE) is expressed as follows: ; in, Indicates the charging power of electric vehicles; For bandwidth parameters, The Gaussian kernel function; The total amount of historical charging power data; Indicates the first Historical charging power; If a Gaussian Mixture Model (GMM) is used for fitting, then the probability density model of the electric vehicle charging power can be expressed as: ; in, This represents the number of Gaussian components (usually 2 or 3). , , The first The weights, mean, and standard deviation of each component can be obtained iteratively from historical charging power using the expectation-maximization algorithm. It is a mathematical constant; Finally, the load forecasting error value is fitted with a normal distribution to obtain the probability density model of the load forecasting error, which is expressed as: ; in, This indicates the load forecasting error value; The standard deviation is denoted as .

[0039] In the above embodiments, considering the physical characteristics of distributed wind power output, the first shape parameter and the first scale parameter are obtained by maximum likelihood estimation of historical wind speeds. Combined with real-time wind speed, a two-parameter Weibull distribution is used for modeling, which accurately characterizes the skewed distribution characteristics of wind speed and its nonlinear mapping relationship with wind power output. This provides a wind power output probability density model that conforms to actual physical laws for subsequent probabilistic power flow calculations. Regarding the statistical characteristics of distributed photovoltaic power output, the second shape parameter and the third shape parameter are determined by calculating the mean and variance of irradiance data. Combined with the maximum output of distributed photovoltaic power, a beta distribution is used for modeling, which accurately describes the probability distribution of irradiance within a finite interval, overcoming the limitations of traditional methods. This paper addresses the unreasonable tail probability problem that arises when fitting bounded physical quantities with a state distribution. For the complex distribution characteristics of electric vehicle charging power, kernel density estimation or Gaussian mixture models are used to fit historical charging power, forming a complementary combination that can adapt to the distribution modeling needs under different charging behavior patterns. For load forecasting errors, a normal distribution with a mean of zero is used for fitting, fully utilizing the statistical law that load forecasting errors are symmetrically distributed around the predicted value and have a low probability of large errors. By applying probability density modeling methods that match the physical statistical characteristics of four different types of operating indicators, accurate basic data support is provided for power system operation risk assessment, improving the accuracy of power system operation risk assessment.

[0040] In this embodiment, mapping the probability density models of various operational indicators to a preset standard normal space to determine the standard normal operational indicator corresponding to each operational indicator and the equivalent correlation coefficient between each pair of operational indicators includes: Integrating the probability density models of the various operational indicators yields the cumulative distribution functions of the various operational indicators. Based on the preset Natav transform algorithm, the cumulative distribution function is mapped to the preset standard normal space to obtain the standard normal operating index for each of the operating indices; Based on a preset correlation analysis algorithm, the correlation of the operational data of various operational indicators is calculated to obtain the correlation coefficient between each pair of operational indicators. Based on the standard normal operating index of each of the operating indicators, the correlation coefficient between each pair of operating indicators is equivalently transformed to obtain the equivalent correlation coefficient between each pair of operating indicators in the standard normal space.

[0041] In an optional embodiment, the cumulative distribution function of each of the various operational indicators is obtained by integrating the probability density models of those indicators. ,in, Indicates the first The cumulative distribution function of various operating indicators; Then, the Nataf transform algorithm is used to map the cumulative distribution function to a preset standard normal space (which is the core carrier of the Nataf transform), to obtain the standard normal operating index for each of the aforementioned operating indices; specifically, it can be expressed as: ; in, It is the inverse function of the standard normal cumulative distribution function; Indicates the first The standard normal operating index corresponds to the standard normal operating index in the standard normal space, where the mean is 0 and the variance is 1.

[0042] Next, the correlation analysis algorithm was set as the Pearson correlation coefficient method. For distributed wind power output, historical wind speed was selected for correlation calculation; for distributed photovoltaic power output, solar irradiance data was selected for correlation calculation; for electric vehicle charging power, historical charging power was selected for correlation calculation; and for load forecasting error, the load forecasting error value was selected for correlation calculation. Thus, the correlation coefficient between each pair of the aforementioned operating indicators was calculated. ,in Indicates the first Various operational indicators; Based on the standard normal operating indicators for each of the aforementioned operating indicators, an equivalent transformation is performed on the correlation coefficients between each pair of the aforementioned operating indicators to obtain the equivalent correlation coefficients between each pair of the aforementioned operating indicators in the standard normal space. The equivalent conversion process is expressed as follows: ; in, and They represent the first The mean and standard deviation of various operating indicators; and They represent the first The mean and standard deviation of various operating indicators. It is the joint probability density function of the two-dimensional standard normal distribution.

[0043] It should be noted that the Nataf Transformation algorithm is a Nataf distribution transformation, a classic transformation algorithm in the field of probability. It can transform correlated random variables with arbitrary marginal distributions into independent standard normal random variables. While keeping the marginal distribution of variables unchanged, it handles the correlation between non-normal variables through equivalent correlation coefficients. It is widely used in engineering scenarios such as structural reliability analysis and stochastic simulation.

[0044] In the above embodiments, the cumulative distribution function is obtained by integrating the probability density model of each operating indicator. Then, the cumulative distribution function is mapped to the standard normal space based on the Natav transform algorithm. This can accurately convert the original random variables with arbitrary distribution forms into standard normal variables while completely preserving all the probability information of the original variables, avoiding the information loss problem caused by traditional normalization methods that only use the mean and variance for conversion. On this basis, the correlation of the operating data of each operating indicator is calculated based on the preset correlation analysis algorithm to obtain the correlation coefficient between each operating indicator in the original space. Then, the correlation coefficient is equivalently transformed based on the standard normal operating indicators to obtain the equivalent correlation coefficient in the standard normal space. Thus, while converting non-normally correlated operating indicators into standard normal operating indicators, the probability distribution information and correlation structure of the original variables are completely preserved. This can accurately depict the complex correlation structure that actually exists between wind power output, photovoltaic power output and load, avoiding the correlation distortion problem caused by the use of independence assumptions or simple orthogonal transformations in traditional methods, and further improving the accuracy of power system operation risk assessment.

[0045] In this embodiment, the linearized power flow calculation of the power system based on the standard normal operating index and the equivalent correlation coefficient, to obtain the probability density function of several operating state variables of the power system in each operating scenario, includes: The standard normal operating index and equivalent correlation coefficient are recursively derived using the preset cumulative quantity generation function method to obtain the semi-invariants of various operating indices at each preset order. By performing an inverse Natav transformation on the semi-invariants of various operating indicators at each order, the semi-invariants of the nodal injection power of the power system at each order are obtained. Linearized power flow calculations are performed on the semi-invariants of the node injected power at each order to obtain the node voltage probability density function, branch power flow probability density function, and load shedding probability density function of the power system in each operating scenario. Based on the node voltage probability density function, branch power flow probability density function, and load shedding probability density function, several operating state variable probability density functions of the power system are formed for each operating scenario.

[0046] In the above embodiments, the cumulative generation function method is used to recursively derive the standard normal operating indicators and equivalent correlation coefficients to obtain the semi-invariants of each operating indicator at each preset order. As a high-order statistical characteristic of random variables, the semi-invariants can completely characterize the skewness, kurtosis, and other distribution patterns of non-normally distributed variables, and their homogeneity and additivity make the propagation calculation under linear transformation extremely efficient. Then, the inverse Natav transformation is performed on the semi-invariants of each operating indicator at each order to obtain the semi-invariants of the power system node injection power at each order, ensuring that the statistical characteristics of the node injection power are completely transmitted to the power flow calculation stage. Finally, linearized power flow calculation is performed on the semi-invariants of the node injection power at each order to obtain the node voltage probability density function, branch power flow probability density function, and load shedding probability density function under each operating scenario, forming a set of probability density functions covering the entire power grid operating state. By combining the semi-invariant method with linearized power flow calculation, accurate and complete operating state probability distribution information is provided for subsequent conditional risk assessment, further improving the accuracy of power system operating risk assessment.

[0047] In this embodiment, the linearized power flow calculation of the injected power at each order of semi-invariants to obtain the node voltage probability density function, branch power flow probability density function, and load shedding probability density function of the power system in each operating scenario includes: Based on the preset linearized power flow equations, the semi-invariants of the node injection power at each order are mapped to obtain the semi-invariants of the node voltage and the branch power flow at each order in the power system. The mean and standard deviation of the node voltage in each operating scenario are determined based on the semi-invariants of the node voltage at each order within the power system. The node voltage mean and standard deviation for each operating scenario are fitted with a probability density function to obtain the node voltage probability density function of the power system for each operating scenario. The mean and standard deviation of the branch power flow in each operating scenario are determined based on the semi-invariants of the branch power flow at each order, and the mean and standard deviation of the branch power flow in each operating scenario are fitted based on the preset series expansion method to obtain the branch power flow probability density function of the power system in each operating scenario. The load shedding power data of the power system in each operating scenario is obtained, and the load shedding power data is fitted according to a preset kernel density estimation algorithm to obtain the load loss probability density function of the power system in each operating scenario.

[0048] In the above embodiments, by pre-setting linearized power flow equations, the semi-invariants of node injected power at each order are mapped to the semi-invariants of node voltage and branch power flow at each order. This accurately transmits the statistical characteristics of the semi-invariants of node injected power at each order to the node voltage and branch power flow. After determining the mean and standard deviation of the semi-invariants of node voltage at each order, probability density fitting is performed to obtain the node voltage probability density function, which can retain the complete statistical information carried by the semi-invariants. After determining the mean and standard deviation of the semi-invariants of branch power flow at each order, a pre-set series expansion method is used to fit the branch power flow probability density function. This can reconstruct a non-normal probability density function of arbitrary shape under the condition that only the semi-invariants of each order are known, thus balancing fitting accuracy and computational simplicity. The kernel density estimation algorithm is used to fit the load shedding power data. It can accurately characterize the actual probability distribution characteristics of the load shedding without pre-setting any distribution form, ensuring the fitting accuracy of the probability density functions of various operating state variables. This provides a reliable input for the accurate calculation of subsequent conditional risk values ​​and improves the accuracy of power system operation risk assessment.

[0049] In an optional embodiment, the Cumulant Generating Function method is used to recursively derive the standard normal operating index and the equivalent correlation coefficient to obtain the semi-invariants of various operating indices at each preset order; wherein, the preset order includes, but is not limited to: first order, second order, third order, etc. Then, an inverse Natav transformation (the inverse process of the Natav transformation) is performed on the semi-invariants of the various operating indicators at each order to obtain the semi-invariants of the nodal injection power of the power system at each order. The linearized power flow equations are defined as follows: ; This represents the node-injected power disturbance vector of the power system; It is a Jacobian matrix; The vector represents the change in node voltage magnitude and phase angle. By mapping the semi-invariants of the injected power at each order to the node voltage at each order in the power system, the semi-invariants of the branch power flow at each order are obtained. Since the process of linearizing power flow calculation is already a mature technology, it will not be described in detail here.

[0050] Since the mean is first-order and the standard deviation is second-order, this embodiment defines the mean of the node voltage in each operating scenario from the semi-invariant of the node voltage at each order within the power system. and standard deviation Then, Gram-Charlier series expansion is used to fit the probability density of the node voltage in each operating scenario, resulting in the node voltage probability density function of the power system in each operating scenario, which is expressed as: ; This refers to node voltage variables; and Representing the scene respectively Next The mean and standard deviation of the voltage at each node; It is standardization Semi-invariants of order 1; It is the standard normal probability density function. For the first Hermite polynomial of order 1 Let the order be the series expansion. Similarly, the mean of the branch power flow in each operating scenario is determined based on the semi-invariants of the branch power flow at each order. and standard deviation ,definition Indicates the first Branch power flow; similarly, Gram-Charlier series expansion (i.e., series expansion method) is used to construct the branch power flow probability density function of the power system in each of the aforementioned operating scenarios. It is represented as: ; in, Represents branch power flow variables; and Scenes Next The mean and standard deviation of the power flow in each branch. For standardization Semi-invariants of order 1.

[0051] The load shedding power data of the power system in each operating scenario is obtained by sampling the power flow calculation under each operating scenario using the Latin Hypercube Sampling algorithm. Then, a kernel density estimation algorithm is used to fit the load shedding power data to obtain the load loss probability density function of the power system in each operating scenario, which is expressed as: ; in, For load shedding power variables; Representing a scene Next The load shearing power of each sample; The number of samples; This is the kernel function.

[0052] It should be noted that the Gram-Charlier series expansion is a classic series method in statistics used to approximate non-normal probability density functions. Based on the normal distribution, it modifies the normal distribution by introducing higher-order cumulants and Hermitian polynomials, and can accurately capture non-normal characteristics of the distribution such as skewness and kurtosis.

[0053] In this embodiment, the conditional risk assessment of the power system based on the probability density function of the operating state variables, to obtain several conditional risk values ​​for the power system in each operating scenario, includes: Based on the node voltage probability density function of the power system in each operating scenario, and combined with the preset node voltage limit, a voltage limit exceedance severity variable of the power system in each operating scenario is constructed. Based on the voltage exceedance severity variable, a voltage exceedance risk assessment is performed on the power system to obtain the voltage exceedance risk value of the power system in each operating scenario; Based on the branch power flow probability density function of the power system in each operating scenario, and combined with the preset maximum allowable power flow value of the branch, the branch overload severity variable of the power system in each operating scenario is constructed. Based on the branch overload severity variable, the branch overload condition risk assessment of the power system is performed to obtain the branch overload risk value of the power system in each operating scenario. The load loss risk value of the power system in each operating scenario is determined based on the load loss probability density function of the power system in each operating scenario.

[0054] In the above embodiments, a voltage exceedance severity variable is constructed based on the node voltage probability density function and a preset node voltage limit, transforming the power system node voltage into a single voltage exceedance severity variable. This allows the probability distribution of the severity variable to centrally reflect the degree of loss when the power system is in a voltage unsafe state. A branch overload severity variable is constructed based on the branch power flow probability density function and a preset maximum allowable power flow value, accurately reflecting the risk level of branch power flow overload in the power system. Furthermore, conditional risk assessment is performed based on the voltage exceedance severity variable, the branch overload severity variable, and the load shedding probability density function, respectively. Conditional risk assessment can perform expected value calculations for the tail losses of random variables, effectively amplifying the tail contribution of the severity variable under extreme fault scenarios. This avoids the inherent shortcomings of traditional expected value indicators that overwhelm tail risks in high-probability safe scenarios, thus improving the accuracy of power system operation risk assessment.

[0055] In this embodiment, obtaining the operational risk assessment result of the power system based on several conditional risk values ​​of the power system in each operational scenario includes: Based on the probability value of the remote confirmation failure of the disconnect switch, the voltage over-limit risk value of the power system in each of the operating scenarios is weighted and fused to obtain the comprehensive voltage over-limit risk value of the power system. Based on the probability value of the remote confirmation failure of the disconnect switch, the branch overload risk value of the power system in each of the operating scenarios is weighted and fused to obtain the comprehensive branch overload risk value of the power system. Based on the probability value of the remote confirmation of failure of the disconnect switch, the load loss risk value of the power system in each of the operating scenarios is weighted and fused to obtain the comprehensive load loss risk value of the power system. The weighting coefficients of the voltage over-limit comprehensive risk value, branch overload comprehensive risk value, and load shedding comprehensive risk value are obtained, and the comprehensive risk value of voltage over-limit comprehensive risk value, branch overload comprehensive risk value, and load shedding comprehensive risk value are weighted and summed to obtain the comprehensive operation risk value of the power system; The comprehensive operational risk value is compared with the preset risk threshold to obtain the operational risk assessment result of the power system.

[0056] In the above embodiments, based on the probability value of remotely confirmed failure of the disconnector, the voltage over-limit risk value, branch overload risk value, and load shedding risk value are weighted and fused to obtain the comprehensive risk values ​​of voltage over-limit, branch overload, and load shedding, respectively. The probability value of remotely confirmed failure of the disconnector is directly embedded as a risk quantification factor in the calculation process of the comprehensive risk value using a weighted fusion mechanism, so that the remotely confirmed failure of the disconnector can be transformed into a quantifiable risk contribution. Then, the weight coefficients of the comprehensive risk values ​​of voltage over-limit, branch overload, and load shedding are obtained and weighted summed to obtain the comprehensive operational risk value. The weight allocation can adaptively reflect the relative importance of each risk type under different operating scenarios. The comprehensive operational risk value is compared with a preset risk threshold to obtain the operational risk assessment result, which improves the accuracy of power system operational risk assessment.

[0057] In an alternative embodiment, due to the node voltage probability density function It represents the numerical distribution of node voltages, therefore it can be derived from the node voltage probability density function. Extracting scenes Next per-unit value of node voltage Simultaneously define node voltage limits, including the node voltage upper limit. and node voltage lower limit Define the number of node voltages as At this time, the severity of voltage exceeding the limit is variable. Represented as: ; Based on the voltage exceedance severity variable, a voltage exceedance risk assessment is performed on the power system to obtain the voltage exceedance risk value for the power system in each operating scenario. The process is as follows: First, the voltage exceedance severity variable... Construct the probability density function of voltage exceedance severity for the power system in each of the aforementioned operating scenarios. It is represented as: ; Therefore, the voltage over-limit risk value is expressed as: ; in, For the confidence level, a value of 0.95 is preferred; For the scene The severity of voltage over-limit at the confidence level Value at risk below; This is a variable representing the severity of voltage exceeding limits.

[0058] Similarly, due to the branch power flow probability density function This represents the numerical distribution of power flow in branch circuits, and therefore can be derived from the probability density function of the branch power flow. Extracting scenes Next The trend value of the branch road trend At the same time, define the maximum allowable power flow value for the branch. Therefore, the branch overload severity variable is expressed as: ; For the number of branch power flows; Based on this, the probability density function of the branch overload severity is calculated. It is represented as: ; Subsequently, based on the branch overload severity variable, a branch overload condition risk assessment is performed on the power system to obtain the branch overload risk value of the power system in each operating scenario, which is expressed as: ; in, For the scene The severity of the overload in the lower branch at the confidence level Value at risk below; For branch overload severity variables; For the load shedding risk value, it is based on the load shedding probability density function of the power system in each of the aforementioned operating scenarios. The loss-of-load risk value of the power system in each of the aforementioned operating scenarios is determined as follows: ; For the scene The load shedding is at a confidence level Value at risk.

[0059] Therefore, based on the probability value of the remote confirmation failure of the disconnector, the voltage over-limit risk value of the power system in each operating scenario is weighted and fused to obtain the comprehensive voltage over-limit risk value of the power system. It is represented as: ; This represents the voltage over-limit risk value for normal operating scenarios. The voltage over-limit risk value is used to confirm the failure scenario of the disconnect switch remotely; Based on the probability value of remotely confirmed failure of the disconnector, the branch overload risk values ​​of the power system in each operating scenario are weighted and fused to obtain the comprehensive branch overload risk value of the power system. It is represented as: ; The branch overload risk value is for normal operation scenarios; The branch overload risk value is used to confirm the failure scenario of the switch at a remote location; Based on the probability value of remotely confirmed failure of the disconnector, the load loss risk value of the power system in each of the aforementioned operating scenarios is weighted and fused to obtain the comprehensive load loss risk value of the power system. It is represented as: ; This represents the risk value for load loss during normal operation. To determine the underload risk value for remotely confirmed failure scenarios of the disconnect switch; This embodiment uses the entropy weight method to determine the weighting coefficients for the comprehensive risk values ​​of voltage over-limit, branch overload, and load shedding. Finally, the comprehensive risk values ​​of voltage over-limit, branch overload, and load shedding are weighted and summed to obtain the comprehensive operational risk value of the power system. It is represented as: ; in, This is the weighting coefficient for the overall risk value of voltage exceeding limits; The weighting coefficient for the comprehensive risk value of branch overload; This is the weighting coefficient for the comprehensive risk value of load loss.

[0060] Setting risk thresholds includes: , , and The overall operational risk value is greater than The operational risk assessment result is at the red high-risk level, and the comprehensive operational risk value of the power system is within [a certain range]. The operational risk assessment result is orange, indicating a higher risk level, and the comprehensive operational risk value of the power system is within [a certain range]. The operational risk assessment result is yellow (medium risk level), and the comprehensive operational risk value of the power system is less than or equal to... The operational risk assessment result is then classified as green and low-risk.

[0061] Example 2 Please refer to Figure 2 , Figure 2 A schematic diagram of a power system assessment system based on remote confirmation of failure of disconnector provided in an embodiment of the present invention includes: a power system data acquisition module 201, a probability density model construction module 202, a standard normal space mapping module 203, a linearized power flow calculation module 204, a conditional risk assessment module 205, and a power system operation risk assessment module 206. The power system data acquisition module 201 is used to acquire the probability value of remotely confirmed failure of disconnectors in the power system and the operating data of each operating indicator; and to construct several operating scenarios of the power system based on the probability value of remotely confirmed failure of disconnectors. The probability density model construction module 202 is used to construct a probability density model for each of the operating indicators in the power system based on the operating data; The standard normal space mapping module 203 is used to map the probability density models of various operating indicators to a preset standard normal space to determine the standard normal operating indicator corresponding to each operating indicator and the equivalent correlation coefficient between each pair of operating indicators. The linearized power flow calculation module 204 is used to perform linearized power flow calculation on the power system based on the standard normal operating index and the equivalent correlation coefficient, and obtain the probability density function of several operating state variables of the power system in each operating scenario. The conditional risk assessment module 205 is used to perform conditional risk assessment on the power system based on the probability density function of the operating state variables, and obtain several conditional risk values ​​of the power system in each operating scenario. The power system operation risk assessment module 206 is used to obtain the operation risk assessment result of the power system based on several conditional risk values ​​of the power system in each operation scenario.

[0062] In this embodiment, the probability density model construction module 202 includes: a probability density model construction unit; In the probability density model construction unit, the operating indicators include: distributed wind power output, distributed photovoltaic power output, electric vehicle charging power, and load forecasting error; the operating data of distributed wind power output includes: historical wind speed and real-time wind speed; the operating data of distributed photovoltaic power output includes: solar irradiance data and maximum distributed photovoltaic power output; the operating data of electric vehicle charging power includes: historical charging power; and the operating data of load forecasting error includes: load forecasting error value. The probability density model building unit is used to perform maximum likelihood estimation of the historical wind speed to obtain the first shape parameter and the first scale parameter; Based on the real-time wind speed, the first shape parameter, and the first scale parameter, a two-parameter Weibull distribution model is performed on the distributed wind power output to obtain the probability density model of the distributed wind power output. Calculate the mean and variance of the light intensity data, and determine the second shape parameter and the third shape parameter based on the mean and variance of the light intensity data; Based on the maximum output of the distributed photovoltaic system, the second shape parameter, and the third shape parameter, a beta distribution model of the distributed photovoltaic power is performed to obtain the probability density model of the distributed photovoltaic power. The historical charging power is fitted based on a preset kernel density estimation method or a Gaussian mixture model to obtain a probability density model of the electric vehicle charging power. The load prediction error value is fitted with a normal distribution to obtain the probability density model of the load prediction error.

[0063] In this embodiment, the standard normal space mapping module 203 includes: a standard normal space mapping unit; The standard normal space mapping unit is used to integrate the probability density models of various operating indicators to obtain the cumulative distribution functions of various operating indicators. Based on the preset Natav transform algorithm, the cumulative distribution function is mapped to the preset standard normal space to obtain the standard normal operating index for each of the operating indices; Based on a preset correlation analysis algorithm, the correlation of the operational data of various operational indicators is calculated to obtain the correlation coefficient between each pair of operational indicators. Based on the standard normal operating index of each of the operating indicators, the correlation coefficient between each pair of operating indicators is equivalently transformed to obtain the equivalent correlation coefficient between each pair of operating indicators in the standard normal space.

[0064] In this embodiment, the linearized power flow calculation module 204 includes: a linearized power flow calculation unit; The linearized power flow calculation unit is used to recursively derive the standard normal operating index and equivalent correlation coefficient according to the preset cumulative quantity generation function method, so as to obtain the semi-invariants of various operating indices at each preset order. By performing an inverse Natav transformation on the semi-invariants of various operating indicators at each order, the semi-invariants of the nodal injection power of the power system at each order are obtained. Linearized power flow calculations are performed on the semi-invariants of the node injected power at each order to obtain the node voltage probability density function, branch power flow probability density function, and load shedding probability density function of the power system in each operating scenario. Based on the node voltage probability density function, branch power flow probability density function, and load shedding probability density function, several operating state variable probability density functions of the power system are formed for each operating scenario.

[0065] In this embodiment, the linearized power flow calculation unit includes: a linearized power flow calculation subunit; The linearized power flow calculation subunit is used to map the semi-invariants of the node injection power at each order based on the preset linearized power flow equation, so as to obtain the semi-invariants of the node voltage and the branch power flow at each order in the power system. The mean and standard deviation of the node voltage in each operating scenario are determined based on the semi-invariants of the node voltage at each order within the power system. The node voltage mean and standard deviation for each operating scenario are fitted with a probability density function to obtain the node voltage probability density function of the power system for each operating scenario. The mean and standard deviation of the branch power flow in each operating scenario are determined based on the semi-invariants of the branch power flow at each order, and the mean and standard deviation of the branch power flow in each operating scenario are fitted based on the preset series expansion method to obtain the branch power flow probability density function of the power system in each operating scenario. The load shedding power data of the power system in each operating scenario is obtained, and the load shedding power data is fitted according to a preset kernel density estimation algorithm to obtain the load loss probability density function of the power system in each operating scenario.

[0066] In this embodiment, the conditional risk assessment module 205 includes: a conditional risk assessment unit; The conditional risk assessment unit is used to construct a voltage exceedance severity variable for the power system in each operating scenario based on the node voltage probability density function of the power system in each operating scenario and in combination with a preset node voltage limit. Based on the voltage exceedance severity variable, a voltage exceedance risk assessment is performed on the power system to obtain the voltage exceedance risk value of the power system in each operating scenario; Based on the branch power flow probability density function of the power system in each operating scenario, and combined with the preset maximum allowable power flow value of the branch, the branch overload severity variable of the power system in each operating scenario is constructed. Based on the branch overload severity variable, the branch overload condition risk assessment of the power system is performed to obtain the branch overload risk value of the power system in each operating scenario. The load loss risk value of the power system in each operating scenario is determined based on the load loss probability density function of the power system in each operating scenario.

[0067] In this embodiment, the power system operation risk assessment module 206 includes: a power system operation risk assessment unit; The power system operation risk assessment unit is used to perform weighted fusion of the voltage over-limit risk value of the power system in each of the operation scenarios based on the probability value of the remote confirmation failure of the disconnector, so as to obtain the comprehensive voltage over-limit risk value of the power system. Based on the probability value of the remote confirmation failure of the disconnect switch, the branch overload risk value of the power system in each of the operating scenarios is weighted and fused to obtain the comprehensive branch overload risk value of the power system. Based on the probability value of the remote confirmation of failure of the disconnect switch, the load loss risk value of the power system in each of the operating scenarios is weighted and fused to obtain the comprehensive load loss risk value of the power system. The weighting coefficients of the voltage over-limit comprehensive risk value, branch overload comprehensive risk value, and load shedding comprehensive risk value are obtained, and the comprehensive risk value of voltage over-limit comprehensive risk value, branch overload comprehensive risk value, and load shedding comprehensive risk value are weighted and summed to obtain the comprehensive operation risk value of the power system; The comprehensive operational risk value is compared with the preset risk threshold to obtain the operational risk assessment result of the power system.

[0068] This embodiment obtains the probability value of remote confirmation failure of disconnectors and constructs several operating scenarios accordingly, enabling power system operation risk assessment to account for topology misjudgment risks caused by disconnector position signal transmission interruptions, false alarms, or confirmation timeouts. Next, by constructing probability density models of various operating indicators and mapping them to a standard normal space to determine equivalent correlation coefficients, it accurately depicts the complex correlation structure actually existing between wind power output, photovoltaic output, and loads, avoiding correlation distortion problems caused by traditional methods using independent assumptions or simple orthogonal transformations. Based on this, linearized power flow calculations are performed using standard normal operating indicators and equivalent correlation coefficients to obtain the probability density function of operating state variables. This is then combined with conditional risk assessment to obtain conditional risk values ​​for each scenario. Conditional risk assessment effectively overcomes the inherent insensitivity of traditional expected value-based risk indicators to extreme events. Finally, a comprehensive assessment result is obtained based on the conditional risk values ​​of each scenario, improving the accuracy of power system operation risk assessment.

[0069] This invention, through obtaining the probability value of remote confirmation failure of disconnectors and constructing several operating scenarios accordingly, enables power system operation risk assessment to account for topology misjudgment risks caused by disconnector position signal transmission interruptions, false alarms, or confirmation timeouts. Next, by constructing probability density models of various operating indicators and mapping them to a standard normal space to determine equivalent correlation coefficients, the complex correlation structure actually existing between wind power output, photovoltaic output, and loads can be accurately characterized, avoiding the correlation distortion problem caused by traditional methods using independent assumptions or simple orthogonal transformations. Based on this, linearized power flow calculations are performed using standard normal operating indicators and equivalent correlation coefficients to obtain the probability density function of operating state variables. This is then combined with conditional risk assessment to obtain conditional risk values ​​for each scenario. Conditional risk assessment effectively overcomes the inherent defect of traditional expected value-based risk indicators being insensitive to extreme events. Finally, a comprehensive assessment result is obtained based on the conditional risk values ​​of each scenario, improving the accuracy of power system operation risk assessment.

[0070] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above descriptions are merely specific embodiments of the present invention and are not intended to limit the scope of protection of the present invention. In particular, it should be noted that any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention for those skilled in the art.

Claims

1. A power system assessment method based on remote confirmation of disconnector failure, characterized in that, include: Obtain the probability value of remotely confirmed failure of disconnectors within the power system and the operating data of each operating indicator; Several operating scenarios of the power system are constructed based on the probability value of the remote confirmation failure of the disconnect switch; Based on the operational data, a probability density model is constructed for each of the operational indicators within the power system; The probability density models of various operational indicators are mapped to a preset standard normal space to determine the standard normal operational indicator corresponding to each operational indicator and the equivalent correlation coefficient between each pair of operational indicators. Based on the standard normal operating index and equivalent correlation coefficient, linearized power flow calculation is performed on the power system to obtain the probability density function of several operating state variables of the power system in each operating scenario; Based on the probability density function of the operating state variables, the power system is subjected to conditional risk assessment to obtain several conditional risk values ​​for the power system in each operating scenario; Based on several conditional risk values ​​of the power system in each of the aforementioned operating scenarios, the operational risk assessment results of the power system are obtained.

2. The power system assessment method based on remote confirmation of disconnector failure as described in claim 1, characterized in that, The operational indicators include: distributed wind power output, distributed photovoltaic power output, electric vehicle charging power, and load forecasting error; the operational data for distributed wind power output includes: historical wind speed and real-time wind speed; the operational data for distributed photovoltaic power output includes: solar irradiance data and maximum distributed photovoltaic power output; the operational data for electric vehicle charging power includes: historical charging power; the operational data for load forecasting error includes: load forecasting error value; the construction of a probability density model for each of the operational indicators within the power system based on the operational data includes: The historical wind speed is estimated by maximum likelihood to obtain the first shape parameter and the first scale parameter; Based on the real-time wind speed, the first shape parameter, and the first scale parameter, a two-parameter Weibull distribution model is performed on the distributed wind power output to obtain the probability density model of the distributed wind power output. Calculate the mean and variance of the light intensity data, and determine the second shape parameter and the third shape parameter based on the mean and variance of the light intensity data; Based on the maximum output of the distributed photovoltaic system, the second shape parameter, and the third shape parameter, a beta distribution model of the distributed photovoltaic power is performed to obtain the probability density model of the distributed photovoltaic power. The historical charging power is fitted based on a preset kernel density estimation method or a Gaussian mixture model to obtain a probability density model of the electric vehicle charging power. The load prediction error value is fitted with a normal distribution to obtain the probability density model of the load prediction error.

3. The power system assessment method based on remote confirmation of disconnector failure as described in claim 2, characterized in that, The step of mapping the probability density models of various operational indicators to a preset standard normal space to determine the standard normal operational indicator corresponding to each operational indicator and the equivalent correlation coefficient between each pair of operational indicators includes: Integrating the probability density models of the various operational indicators yields the cumulative distribution functions of the various operational indicators. Based on the preset Natav transform algorithm, the cumulative distribution function is mapped to the preset standard normal space to obtain the standard normal operating index for each of the operating indices; Based on a preset correlation analysis algorithm, the correlation of the operational data of various operational indicators is calculated to obtain the correlation coefficient between each pair of operational indicators. Based on the standard normal operating index of each of the operating indicators, the correlation coefficient between each pair of operating indicators is equivalently transformed to obtain the equivalent correlation coefficient between each pair of operating indicators in the standard normal space.

4. The power system assessment method based on remote confirmation of disconnector failure as described in claim 1, characterized in that, The linearized power flow calculation of the power system based on the standard normal operating index and equivalent correlation coefficient yields a probability density function of several operating state variables of the power system in each operating scenario, including: The standard normal operating index and equivalent correlation coefficient are recursively derived using the preset cumulative quantity generation function method to obtain the semi-invariants of various operating indices at each preset order. By performing an inverse Natav transformation on the semi-invariants of various operating indicators at each order, the semi-invariants of the nodal injection power of the power system at each order are obtained. Linearized power flow calculations are performed on the semi-invariants of the injected power at each order to obtain the node voltage probability density function, branch power flow probability density function, and load shedding probability density function of the power system in each operating scenario. Based on the node voltage probability density function, branch power flow probability density function, and load shedding probability density function, several operating state variable probability density functions of the power system are formed for each operating scenario.

5. The power system assessment method based on remote confirmation of disconnector failure as described in claim 4, characterized in that, The linearized power flow calculation of the injected power at each node at each order of semi-invariant yields the node voltage probability density function, branch power flow probability density function, and off-load probability density function of the power system in each operating scenario, including: Based on the preset linearized power flow equations, the semi-invariants of the node injection power at each order are mapped to obtain the semi-invariants of the node voltage and the branch power flow at each order in the power system. The mean and standard deviation of the node voltage in each operating scenario are determined based on the semi-invariants of the node voltage at each order within the power system. The node voltage mean and standard deviation for each operating scenario are fitted with a probability density function to obtain the node voltage probability density function of the power system for each operating scenario. The mean and standard deviation of the branch power flow in each operating scenario are determined based on the semi-invariants of the branch power flow at each order, and the mean and standard deviation of the branch power flow in each operating scenario are fitted based on the preset series expansion method to obtain the branch power flow probability density function of the power system in each operating scenario. The load shedding power data of the power system in each operating scenario is obtained, and the load shedding power data is fitted according to a preset kernel density estimation algorithm to obtain the load loss probability density function of the power system in each operating scenario.

6. A power system assessment method based on remote confirmation of failure of a disconnector as described in claim 4 or 5, characterized in that, The conditional risk assessment of the power system based on the probability density function of the operating state variables yields several conditional risk values ​​for the power system in each operating scenario, including: Based on the node voltage probability density function of the power system in each operating scenario, and combined with the preset node voltage limit, a voltage limit exceedance severity variable of the power system in each operating scenario is constructed. Based on the voltage exceedance severity variable, a voltage exceedance risk assessment is performed on the power system to obtain the voltage exceedance risk value of the power system in each operating scenario; Based on the branch power flow probability density function of the power system in each operating scenario, and combined with the preset maximum allowable power flow value of the branch, the branch overload severity variable of the power system in each operating scenario is constructed. Based on the branch overload severity variable, the branch overload condition risk assessment of the power system is performed to obtain the branch overload risk value of the power system in each operating scenario. The load loss risk value of the power system in each operating scenario is determined based on the load loss probability density function of the power system in each operating scenario.

7. The power system assessment method based on remote confirmation of failure of disconnectors as described in claim 6, characterized in that, The method of obtaining the operational risk assessment result of the power system based on several conditional risk values ​​of the power system in each operational scenario includes: Based on the probability value of the remote confirmation failure of the disconnect switch, the voltage over-limit risk value of the power system in each of the operating scenarios is weighted and fused to obtain the comprehensive voltage over-limit risk value of the power system. Based on the probability value of the remote confirmation failure of the disconnect switch, the branch overload risk value of the power system in each of the operating scenarios is weighted and fused to obtain the comprehensive branch overload risk value of the power system. Based on the probability value of the remote confirmation of failure of the disconnect switch, the load loss risk value of the power system in each of the operating scenarios is weighted and fused to obtain the comprehensive load loss risk value of the power system. The weighting coefficients of the voltage over-limit comprehensive risk value, branch overload comprehensive risk value, and load shedding comprehensive risk value are obtained, and the comprehensive risk value of voltage over-limit comprehensive risk value, branch overload comprehensive risk value, and load shedding comprehensive risk value are weighted and summed to obtain the comprehensive operation risk value of the power system; The comprehensive operational risk value is compared with the preset risk threshold to obtain the operational risk assessment result of the power system.

8. A power system assessment system based on remote failure confirmation of disconnectors, characterized in that, include: The system includes a power system data acquisition module, a probability density model construction module, a standard normal space mapping module, a linearized power flow calculation module, a conditional risk assessment module, and a power system operation risk assessment module. The power system data acquisition module is used to acquire the probability value of remotely confirmed failure of disconnectors in the power system and the operating data of each operating indicator; and to construct several operating scenarios of the power system based on the probability value of remotely confirmed failure of disconnectors. The probability density model building module is used to build a probability density model for each of the operating indicators in the power system based on the operating data; The standard normal space mapping module is used to map the probability density models of various operating indicators to a preset standard normal space, so as to determine the standard normal operating indicator corresponding to each operating indicator and the equivalent correlation coefficient between each pair of operating indicators; The linearized power flow calculation module is used to perform linearized power flow calculation on the power system based on the standard normal operating index and the equivalent correlation coefficient, and obtain the probability density function of several operating state variables of the power system in each operating scenario. The conditional risk assessment module is used to perform conditional risk assessment on the power system based on the probability density function of the operating state variables, and to obtain several conditional risk values ​​of the power system in each operating scenario. The power system operation risk assessment module is used to obtain the operation risk assessment result of the power system based on several conditional risk values ​​of the power system in each operation scenario.

9. A power system assessment system based on remote failure confirmation of disconnectors as described in claim 8, characterized in that, The probability density model construction module includes: a probability density model construction unit; In the probability density model construction unit, the operating indicators include: distributed wind power output, distributed photovoltaic power output, electric vehicle charging power, and load forecasting error; the operating data of distributed wind power output includes: historical wind speed and real-time wind speed; the operating data of distributed photovoltaic power output includes: solar irradiance data and maximum distributed photovoltaic power output; the operating data of electric vehicle charging power includes: historical charging power; and the operating data of load forecasting error includes: load forecasting error value. The probability density model building unit is used to perform maximum likelihood estimation of the historical wind speed to obtain the first shape parameter and the first scale parameter; Based on the real-time wind speed, the first shape parameter, and the first scale parameter, a two-parameter Weibull distribution model is performed on the distributed wind power output to obtain the probability density model of the distributed wind power output. Calculate the mean and variance of the light intensity data, and determine the second shape parameter and the third shape parameter based on the mean and variance of the light intensity data; Based on the maximum output of the distributed photovoltaic system, the second shape parameter, and the third shape parameter, a beta distribution model of the distributed photovoltaic power is performed to obtain the probability density model of the distributed photovoltaic power. The historical charging power is fitted based on a preset kernel density estimation method or a Gaussian mixture model to obtain a probability density model of the electric vehicle charging power. The load prediction error value is fitted with a normal distribution to obtain the probability density model of the load prediction error.

10. A power system assessment system based on remote failure confirmation of disconnectors as described in claim 9, characterized in that, The standard normal space mapping module includes: a standard normal space mapping unit; The standard normal space mapping unit is used to integrate the probability density models of various operating indicators to obtain the cumulative distribution functions of various operating indicators. Based on the preset Natav transform algorithm, the cumulative distribution function is mapped to the preset standard normal space to obtain the standard normal operating index for each of the operating indices; Based on a preset correlation analysis algorithm, the correlation of the operational data of various operational indicators is calculated to obtain the correlation coefficient between each pair of operational indicators. Based on the standard normal operating index of each of the operating indicators, the correlation coefficient between each pair of operating indicators is equivalently transformed to obtain the equivalent correlation coefficient between each pair of operating indicators in the standard normal space.

Images