Quantitative analysis method for uncertain factors of grid-connected entities actively connected to distribution systems

Abstract

The present invention relates to a technical field of uncertainty analysis of power systems, and in particular to a quantitative analysis method for uncertainty factors of a grid-connected entity actively connected to a distribution system. By deploying sensor networks and communication networks, real-time data collection and preliminary analysis are achieved; multi-physics imaging technology is applied to perform equipment detection and defect identification to improve the level of equipment safety management; data integration and risk assessment methods are used, combined with user-side power usage data, to conduct a comprehensive risk assessment; risk matrix tools are used to perform comprehensive data analysis and risk rating, and optimize risk sorting and priority division; through time series analysis and simulation, the impact of uncertainty factors on system stability is quantitatively analyzed; and a global sensitivity analysis method is used to evaluate the relative impact of uncertainty factors. The present invention effectively solves the existing technology in the analysis of uncertainty factors in the distribution system, improves the risk assessment of the power grid, and ensures the stable and safe operation of the power system.

Description

Technical Field

[0001] The present invention relates to the technical field of uncertainty analysis of power systems, and in particular to a quantitative analysis method for uncertainty factors of a grid-connected entity actively connected to a power distribution system. Background Art

[0002] With the development of new energy power systems, especially the active integration of grid-connected entities, distribution systems are becoming increasingly complex. Existing technologies have limitations in handling this complexity. They are unable to effectively monitor and analyze the operating status of distribution systems in real time, resulting in inaccurate and in-time security management and risk assessment of the power grid.

[0003] The existing technology has the following deficiencies:

[0004] Lack of real-time data processing capabilities. Existing methods effectively collect and analyze the operating data of the distribution system in real time, resulting in risk assessment that is not timely and accurate enough; limitations of risk assessment methods. Existing risk assessment methods may not be comprehensive enough and cannot fully utilize all available data, especially power usage data on the user side; insufficient handling of uncertainty factors. Existing methods lack efficient data analysis tools and models to deal with the uncertainty factors brought about by the access of grid-connected entities; insufficient equipment maintenance and safety management. Existing methods lack detailed detection of physical defects and potential safety hazards of equipment, resulting in less than optimized equipment maintenance and safety management.

[0005] Therefore, we need a more efficient quantitative analysis method that can process and analyze large amounts of real-time data, identify the uncertainty factors brought about by the access of grid-connected entities, obtain the key uncertainty factors that comprehensively affect the safe and stable operation of the active distribution system, and support the safe and stable operation of the distribution system. Summary of the Invention

[0006] In view of the above deficiencies in the existing technology, the technical problem to be solved by the present invention is to provide a quantitative analysis method for the uncertainty factors of the active access of the grid-connected entity to the distribution system, so as to improve the efficiency and accuracy of the risk assessment of the distribution network system, and make complex grid data easier to understand and analyze.

[0007] In order to achieve the above objectives, the present invention adopts the following technical solutions:

[0008] A quantitative analysis method for uncertain factors of a grid-connected entity actively connected to a power distribution system comprises the following steps:

[0009] S1. Uncertainty modeling: Data related to uncertainties is collected through the distribution system database. All uncertainties that may affect the operation of the distribution system (such as load fluctuations, intermittent renewable energy, and equipment aging) are listed. A probabilistic model is established to quantify the uncertainty of these uncertainties. The probability model includes:

[0010] Use normal distribution to simulate load fluctuations, X l ~N(μ l ,σ l 2 ), N represents normal distribution, μ l is the mean value of the load, σ l 2 is the standard deviation;

[0011] Using Poisson distribution to simulate renewable energy intermittency, X p ~P(λ), P represents the Poisson distribution, λ is the expected value of renewable energy power generation per unit time;

[0012] Use exponential distribution to model equipment failure rate: X age ~E(β), where E stands for exponential distribution, describing the failure time of a device or component, i.e., the time interval between two failures. β is the inverse of the failure rate, representing the average number of failures per unit time.

[0013] S2. State space modeling: integrating the uncertainty factor model established in step S1, establishing state equations and observation equations, and constructing a unified state space model to describe the evolution of the power system state over time;

[0014] S3. Evaluate the probability of failure of the power system under the influence of uncertain factors, and simulate the state space model established in step S2 using Monte Carlo simulation to estimate the failure probability;

[0015] S4. Sensitivity analysis: Change the parameters of the state-space model constructed in step S2 and use the Sobol method to evaluate the impact of parameter changes on the system. Evaluate the impact of parameter changes in the state-space model on the system output and identify the uncertainties that have the greatest impact on the system.

[0016] Furthermore, the method for constructing the state space model includes:

[0017] S21. Using the model established for the uncertainty factors in step S1, select key response indicators that can characterize changes in the system state. The key response indicators can quantify the impact of the uncertainty factors on the system;

[0018] Key response indicators include: node voltage amplitude and phase angle, which reflect the voltage stability of the power system; generator output power, which indicates the power supply capacity; line flow, which reflects the condition of power flowing through the system; load fluctuation, which affects the supply and demand balance of the power grid; the accuracy of renewable energy power generation prediction, which affects power grid scheduling and planning; equipment failure rate, which affects the reliability and maintenance requirements of the system.

[0019] S22. Determine the state variable and select the key response indicator as the state variable. The state variable should be able to directly or indirectly reflect the influence of the uncertainty factor;

[0020] S23. Establish a state equation to establish the dynamic relationship between state variables. The state equation is expressed as a set of differential equations or difference equations that describe the changes of state variables over time. The state equation formula is:

[0021]

[0022] X(t)=[V t (t),P g (t),I l (t),P l (t),P r (t),λ(t),f(t),...];

[0023] In the formula, is the derivative of the state vector, which represents the rate of change of the state over time; A is the state matrix, which describes the dynamic relationship between state variables; X(t) is the state variable, which is the minimum information set describing the current state of the system; B is the input matrix, which links the control input to the state change; U(t) is the control input vector, including power input, load change, etc.; ω(t) is the process noise, which represents the internal random disturbance that the model cannot capture; V t (t) represents the voltage amplitude and phase angle of each node in the power system, P g (t) represents the output power of each generator in the system at time t; I l (t) represents the current flowing through each line in the system, which usually includes the real part (active power flow) and the imaginary part (reactive power flow); P l (t) represents the total load in the system or the load of each node in the time period t; P r (t) represents the predicted power generation from renewable energy sources such as wind and solar energy; λ(t) represents the rate at which equipment in the system may fail; f(t) represents the operating frequency of the power system;

[0024] S24. Establish observation equations to establish the relationship between state variables and observable variables. Observable variables include node voltage, line current, power, etc. The observation equations are expressed as a set of algebraic equations. The observation equation formula is:

[0025] Z(t)=CX(t)+DU(t)+υ(t);

[0026] X(t)=[V t (t),P g (t),I l (t),P l (t),P r (t),λ(t),f(t),...];

[0027] The observation model allows estimating state variables from observable variables, integrating data from different sensors, and by taking into account the observation noise, the observation model handles the uncertainty in the actual measurements;

[0028] Z(t) is the observation vector, which contains a set of variables that can be directly measured from the power system, such as node voltage, line current, power, etc.; C is the observation matrix, which is a coefficient matrix used to map state variables to the observation space. It defines how state variables affect observable variables; D is the direct transfer matrix, which is used to directly map control inputs to the observation space. The direct transfer matrix describes the direct impact of control inputs on observable variables; υ(t) is the observation noise, which represents random errors and external interference in the measurement process. The observation noise is usually assumed to be Gaussian white noise with zero mean and a certain variance.

[0029] S25. Construct a state-space model by combining the state equation and the observation equation to construct a state-space model. The state-space model formula is expressed as:

[0030]

[0031] is the derivative of the state vector, X(t) is the state variable, A is the state matrix, B is the input matrix, U(t) is the control input vector, ω(t) is the process noise, Z(t) is the observation vector, and υ(t) is the observation noise.

[0032] S3. Evaluate the probability of failure of the power system under the influence of uncertain factors, and simulate the state space model established in step S2 using Monte Carlo simulation to estimate the failure probability;

[0033] Furthermore, a Monte Carlo simulation is performed based on the state space model established in step S2 to estimate the failure probability. The specific method includes:

[0034] S31. Sampling uncertainties: using the probability model established in step S1, randomly sampling each uncertainty factor (such as load fluctuation, renewable energy intermittency, equipment aging, etc.), and using a pseudo-random number generator to generate sample data based on these probability models;

[0035] S32. Set the number of simulations. Use the relationship between the confidence interval and the uncertainty factor sample to estimate the required number of simulations and the range. The number of simulations should be large enough to reduce the variance of the simulation results and improve the stability and credibility of the results. The number of simulations is estimated using the following formula:

[0036]

[0037] N s Estimated number of simulations; Z α / 2 is the value corresponding to the confidence level of the normal distribution; for the common 95% confidence level, Z α / 2 It is about 1.96, which means that under normal distribution, the proportion of areas within 1.96 standard deviations from the mean is about 95%; E m It is the acceptable error range of the simulation results, also known as the half-width of the confidence interval. It means that we hope that the error between the simulation results and the true value does not exceed this range. m The smaller it is, the higher our requirement for simulation accuracy is. Accordingly, the number of simulations required N is s The larger the N s refers to the dimension of the state vector, that is, the number of state variables;

[0038] S33 simulates system operation, and for each set of sample data of uncertain factors, inputs the sample data into the state space model established in step S2 to obtain the predicted values ​​of the state variables and observed variables of the power system under the given uncertainty factor sample;

[0039] S34. Determine whether the system has failed. Based on the preset failure criteria, determine whether the system has failed under the current uncertainty factor sample. The failure criteria judgment formula is:

[0040]

[0041] X(t)=[V t (t),P g (t),I l (t),P l (t),P r (t),λ(t),f(t),...]

[0042] The failure conditions satisfied by X(t) include:

[0043] V min ≤V i (t)≤V max ;

[0044] P g,i,min (t)≤Pg,i (t)≤P g,i,max (t);

[0045] I l,i,max ≥I l,i (t);

[0046] P l,min ≤P l (t)≤P l,max ;

[0047] P r,min ≤P r (t)≤P r,max ;

[0048] λ(t)≤λ max ;

[0049] f min ≤f(t)≤f max ;

[0050] V min and V max are the minimum and maximum safety thresholds of the node voltage, V i (t) is the voltage of each node; P g,i,min (t) and P g,i,max (t) are the minimum and maximum output power limits of the ith generator; I l,i,max is the power flow capacity of the i-th line; P l,min and P l,max are the minimum and maximum safety thresholds of the load, P r,min and P r,max are the minimum and maximum safety thresholds for renewable energy generation; λ max is the maximum safety threshold of equipment failure rate; f min and f max are the minimum and maximum safety thresholds of the system frequency;

[0051] S35. Count the number of failures and record the number of system failures;

[0052] The number of system failures in all simulation samples in step S34 is counted, and the formula is expressed as:

[0053]

[0054] N s is the total number of simulations, F i is the failure logic variable of the i-th simulation;

[0055] S36. Calculate the failure probability. Based on the number of failures and the total number of simulations, calculate the failure probability of the system under the influence of uncertain factors. The calculation formula is:

[0056]

[0057] N F is the number of failures, N s is the total number of simulations.

[0058] S4. Sensitivity analysis: Change the parameters of the state-space model constructed in step S2 and use the Sobol method to evaluate the impact of parameter changes on the system. Evaluate the impact of parameter changes in the state-space model on the system output and identify the uncertainties that have the greatest impact on the system.

[0059] S41. Parameter definition: Determine the initial estimated range for each parameter based on historical data and the physical meaning of the model;

[0060] S42. Conduct sensitivity analysis experiments: Based on the state-space model constructed in step S2, systematically change the value of each parameter of the state-space model, conduct a series of simulation experiments, and record the results;

[0061] S43. Calculate sensitivity index: Based on the experimental results of S42, calculate the sensitivity index to evaluate the impact of the parameters on the system. The calculation formula for the sensitivity index is:

[0062]

[0063] Among them, K i In the fixed parameter, Var(P F |K i |) is the fixed parameter K i Variance of failure probability at time t; Var(P F ) is the total variance of the failure probability when all parameters are varied;

[0064] S44. Analysis results: Based on the size of the sensitivity index, determine which uncertainties have the greatest impact on the system and which parameters have the greatest impact on the probability of system failure, that is, identify sensitive factors.

[0065] In summary, due to the adoption of the above technical solution, the beneficial technical effects of the invention are:

[0066] 1. Improve the accuracy of risk assessment

[0067] By establishing a probabilistic model to quantify uncertainties (such as load fluctuations, intermittent renewable energy, and equipment aging), accurate input data is provided for risk assessment. A state-space model is constructed to integrate the dynamic characteristics of the power system's state over time, enabling risk assessment based on the system's actual dynamic behavior. Monte Carlo simulation is used to estimate failure probabilities, accounting for the combined impact of multiple uncertainties and improving the accuracy of risk assessment results.

[0068] 2. Enhance the system's real-time response capabilities

[0069] Probabilistic models can be updated in real time to reflect the latest system status, providing data support for rapid response. State-space models allow real-time monitoring of system status and rapid prediction of potential changes, providing a basis for real-time decision-making. Simulations can be completed within a reasonable time, enabling real-time risk assessment and response.

[0070] 3. Optimize equipment maintenance and safety management

[0071] Probabilistic models identify uncertainties such as equipment aging, predict equipment failure rates, and provide data support for maintenance planning. State-space models simulate the impact of equipment failures on system status, helping to develop preventive measures and optimize safety management. Sensitivity analysis identifies the uncertainties that have the greatest impact on the system, providing a basis for equipment maintenance and safety management. BRIEF DESCRIPTION OF THE DRAWINGS

[0072] Figure 1 Flowchart of the quantitative analysis method for the uncertainties of the grid-connected entities actively connected to the distribution system. DETAILED DESCRIPTION

[0073] In order to make the purpose, technical solutions and advantages of the invention more clear, the invention is further described in detail below in conjunction with the embodiments. It should be understood that the specific embodiments described here are only used to explain the invention and are not used to limit the invention.

[0074] like Figure 1 The method for quantitatively analyzing uncertainties of a grid-connected entity actively connecting to a power distribution system comprises the following steps:

[0075] S1. Uncertainty factor modeling: Uncertain factors are collected from the distribution system database. Probabilistic models are developed for these uncertain factors (such as load fluctuations, intermittent renewable energy, and equipment aging). The specific impact of each uncertainty factor on the distribution system is analyzed. Establishing a probabilistic model includes:

[0076] Use normal distribution to simulate load fluctuations: X l ~N(μ l ,σ l2 ), N represents normal distribution, μ l is the mean value of the load, σ l 2 is the standard deviation;

[0077] Modeling renewable energy intermittency using Poisson distribution: X p ~P(λ), P represents the Poisson distribution, λ is the expected value of renewable energy power generation per unit time;

[0078] Use exponential distribution to model equipment failure rate: X age ~E(β), where E represents the exponential distribution and describes the failure time of a device or component, i.e., the time interval between two failures. β is the inverse of the failure rate, which represents the average number of failures per unit time.

[0079] S2. State space modeling: Integrate the uncertainty factor model established in step S1, establish state equations and observation equations, and construct a unified state space model to describe the evolution of the power system state over time.

[0080] Furthermore, the method for constructing the state space model includes:

[0081] S21. Using the model established for the uncertainty factors in step S1, based on the understanding of the impact of the uncertainty factors, select key response indicators that can characterize the change in the system state. The key response indicators can quantify the impact of the uncertainty factors on the system.

[0082] Based on the probability model in S1, key response indicators such as node voltage and generator output power are selected as state variables, and state equations describing the dynamic relationship between state variables are established. Combining the state equations with observation data, observation equations are constructed to form a complete state space model, which integrates the dynamic characteristics of the power system and provides a theoretical basis for real-time monitoring and prediction.

[0083] Key response indicators include: node voltage amplitude and phase angle, which reflect the voltage stability of the power system; generator output power, which indicates the power supply capacity; line flow, which reflects the condition of power flowing through the system; load fluctuation, which affects the supply and demand balance of the power grid; the accuracy of renewable energy power generation prediction, which affects power grid scheduling and planning; equipment failure rate, which affects the reliability and maintenance requirements of the system.

[0084] S22. Determine state variables and select the key response indicators as state variables. State variables should be able to directly or indirectly reflect the influence of uncertainty factors. State variables may include node voltage amplitude and phase angle, generator output power, etc.

[0085] S23. Establish a state equation to establish the dynamic relationship between state variables. The state equation is expressed as a set of differential equations or difference equations that describe the changes of state variables over time. The state equation formula is:

[0086]

[0087] X(t)=[V t (t),P g (t),I l (t),P l (t),P r (t),λ(t),f(t),...];

[0088] The state equation combines state variables, state matrices, control inputs, input matrices, and process noise to describe the evolution of system states over time. The state equation captures the time-varying dynamic characteristics of the power system, such as changes in voltage and frequency. Through control inputs, the system state is adjusted to achieve system stability and optimization. The introduction of process noise enables the model to handle uncertainties in actual operation.

[0089] In the formula, is the derivative of the state vector, which represents the rate of change of the state over time; A is the state matrix, which describes the dynamic relationship between state variables; X(t) is the state variable, which is the minimum information set describing the current state of the system; B is the input matrix, which links the control input to the state change; U(t) is the control input vector, including power input, load change, etc.; ω(t) is the process noise, which represents the internal random disturbance that the model cannot capture; V t (t) represents the voltage amplitude and phase angle of each node in the power system, P g (t) represents the output power of each generator in the system at time t; I l (t) represents the current flowing through each line in the system, which usually includes the real part (active power flow) and the imaginary part (reactive power flow); P l (t) represents the total load in the system or the load of each node at time t; P r (t) represents the predicted power generation from renewable energy sources such as wind and solar energy; λ(t) represents the rate at which equipment in the system may fail; f(t) represents the operating frequency of the power system;

[0090] S24. Establish observation equations to establish the relationship between state variables and observable variables. Observable variables include node voltage, line current, power, etc. The observation equations can be expressed as a set of algebraic equations. The observation equation formula is:

[0091] Z(t)=CX(t)+DU(t)+υ(t);

[0092] X(t)=[V t (t),P g (t),I l (t),P l (t),P r (t),λ(t),f(t),...];

[0093] The observation model allows estimating state variables from observable variables, integrating data from different sensors, and by taking into account the observation noise, the observation model handles the uncertainty in the actual measurements;

[0094] Z(t) is the observation vector, which contains a set of variables that can be directly measured from the power system, such as node voltage, line current, power, etc.; C is the observation matrix, which is a coefficient matrix used to map state variables to the observation space. It defines how state variables affect observable variables; D is the direct transfer matrix, which is used to directly map control inputs to the observation space. The direct transfer matrix describes the direct impact of control inputs on observable variables; υ(t) is the observation noise, which represents random errors and external interference in the measurement process. The observation noise is usually assumed to be Gaussian white noise with zero mean and a certain variance.

[0095] S25. Construct a state-space model by combining the state equation and the observation equation to construct a state-space model. The state-space model formula is expressed as:

[0096]

[0097] is the derivative of the state vector, X(t) is the state variable, A is the state matrix, B is the input matrix, U(t) is the control input vector, ω(t) is the process noise, Z(t) is the observation vector, and υ(t) is the observation noise.

[0098] S3. Evaluate the probability of failure of the power system under the influence of uncertain factors, and simulate the state space model established in step S2 using Monte Carlo simulation to estimate the failure probability;

[0099] Set the confidence level percentage for the Monte Carlo simulation and determine the number of simulations based on the required accuracy. Use a pseudo-random number generator to perform sampling based on the probability model to obtain sample data for uncertain factors. Input the sample data into the state space model to simulate system operation and predict state variables and observation variables. Determine the system state based on the failure criterion, count the number of failures, and calculate the failure probability.

[0100] Furthermore, the state space model established in step S2 is simulated using Monte Carlo simulation to estimate the failure probability. The specific method includes:

[0101] S31. Set the number of simulations. Use the relationship between the confidence interval and the sample size to estimate the required number of simulations and the range. The number of simulations should be large enough to reduce the variance of the simulation results and improve the stability and credibility of the results. The number of simulations is estimated using the following formula:

[0102]

[0103] N s Estimated number of simulations; Z α / 2 is the value corresponding to the confidence level of the normal distribution, for the common 95% confidence level; Z α / 2 It is about 1.96, which means that under normal distribution, the proportion of areas within 1.96 standard deviations from the mean is about 95%; E m It is the acceptable error range of the simulation results, also known as the half width of the confidence interval. It means that we hope the error between the simulation results and the true value does not exceed this range. m The smaller it is, the higher the requirement for simulation accuracy is. Accordingly, the number of simulations required N is s The larger the N s refers to the dimension of the state vector, that is, the number of state variables;

[0104] S32. Sampling Uncertain Factors. Using the probability model established in step S1, randomly sample each uncertainty factor (such as load fluctuation, renewable energy intermittency, equipment aging, etc.), and use a pseudo-random number generator to generate sample data based on these probability models;

[0105] S33. Simulate system operation. For each set of sample data of uncertain factors, input the sample data into the state space model established in step S2 to obtain the predicted values ​​of the state variables and observed variables of the power system under the given uncertainty factor sample;

[0106] S34. Determine whether the system has failed. Based on the preset failure criteria, determine whether the system has failed under the current uncertainty factor sample. The failure criteria judgment formula is:

[0107]

[0108] X(t)=[V t (t),P g (t),I l (t),P l (t),P r (t),λ(t),f(t),...];

[0109] The failure conditions satisfied by X(t) include:

[0110] V min≤V i (t)≤V max ;

[0111] P g,i,min (t)≤P g,i (t)≤P g,i,max (t);

[0112] I l,i,max ≥I l,i (t);

[0113] P l,min ≤P l (t)≤P l,max ;

[0114] P r,min ≤P r (t)≤P r,max ;

[0115] λ(t)≤λ max ;

[0116] f min ≤f(t)≤f max ;

[0117] V min and V max are the minimum and maximum safety thresholds of the node voltage, V i (t) is the voltage of each node; P g,i,min (t) and P g,i,max (t) are the minimum and maximum output power limits of the ith generator; I l,i,max is the power flow capacity of the i-th line; P l,min and P l,max are the minimum and maximum safety thresholds of the load, P r,min and P r,max are the minimum and maximum safety thresholds for renewable energy generation; λ max is the maximum safety threshold of equipment failure rate; f min and f max are the minimum and maximum safety thresholds of the system frequency;

[0118] S35. Statistics of failures. Record the number of system failures;

[0119] The number of system failures in all simulation samples in step S34 is counted, and the formula is expressed as:

[0120]

[0121] N s is the total number of simulations, F i is the failure logic variable of the i-th simulation;

[0122] S36. Calculate the failure probability. Based on the number of failures and the total number of simulations, calculate the failure probability of the system under the influence of uncertain factors. The calculation formula is:

[0123]

[0124] N F is the number of failures, N s is the total number of simulations.

[0125] S4. Sensitivity analysis: Change the parameters of the state-space model constructed in step S2 and use the Sobol method to evaluate the impact of parameter changes on the system. Evaluate the impact of parameter changes in the state-space model on the system output and identify the uncertainties that have the greatest impact on the system.

[0126] S41. Parameter definition: Determine the initial estimated range for each parameter based on historical data and the physical meaning of the model;

[0127] S42. Conduct sensitivity analysis experiments: Based on the state-space model constructed in step S2, systematically change the value of each parameter of the state-space model, conduct a series of simulation experiments, and record the results;

[0128] S43. Calculate sensitivity index: Based on the experimental results of S42, calculate the sensitivity index to evaluate the impact of the parameters on the system. The calculation formula for the sensitivity index is:

[0129]

[0130] Among them, K i In the fixed parameter, Var(P F |K i |) is the fixed parameter K i Variance of failure probability at time t; Var(P F ) is the total variance of the failure probability when all parameters are varied;

[0131] S44. Analysis results: Based on the size of the sensitivity index, determine which uncertainties have the greatest impact on the system and which parameters have the greatest impact on the probability of system failure, that is, identify sensitive factors.

[0132] The above description is a preferred embodiment of the invention and is not intended to limit the invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the invention should be included in the scope of protection of the invention.

Claims

1. A quantitative analysis method for uncertainties in a grid-connected entity actively connected to a power distribution system, comprising the following steps: S1. Uncertainty factor modeling: Collect data related to uncertainties through the distribution system database, list all uncertainties that may affect the operation of the distribution system, and establish a probabilistic model for the uncertainties to quantify their uncertainty; S2. State space modeling: integrating the uncertainty factor model established in step S1, establishing state equations and observation equations, and constructing a unified state space model to describe the evolution of the power system state over time; S3. Evaluate the probability of failure of the power system under the influence of uncertain factors, and use Monte Carlo simulation based on the state space model established in step S2 to estimate the failure probability; S4. Sensitivity analysis: Change the parameters of the state-space model constructed in step S2 and use the Sobol method to evaluate the impact of parameter changes in the state-space model on the system output and identify the uncertainties that have the greatest impact on the system. The probability model established in step S1 includes: Use normal distribution to simulate load fluctuations: X l ~N(μ l ,σ l 2 ), N represents normal distribution, μ l is the mean value of the load, σ l 2 is the standard deviation; Modeling renewable energy intermittency using Poisson distribution: X p ~P(λ), P represents the Poisson distribution, λ is the expected value of renewable energy power generation per unit time; Use exponential distribution to model equipment failure rate: X age ~E(β), where E represents the exponential distribution and describes the failure time of a device or component, i.e., the time interval between two failures. β is the inverse of the failure rate, which represents the average number of failures per unit time. In step S2, the method for constructing the state space model includes: S21. Using the model established for the uncertainty factors in step S1, select key response indicators that can characterize changes in the system state, and quantify the impact of the uncertainty factors on the system based on the key response indicators; S22. Determine the state variable, select the key response indicator as the state variable, the state variable can directly reflect the impact of uncertainty factors; S23. Establish a state equation to establish the dynamic relationship between state variables. The state equation is expressed as a set of differential equations to describe the changes of state variables over time. The state equation formula is: X(t)=[V t (t),P g (t),I l (t),P l (t),P r (t),λ(t),f(t)]; In the formula, is the derivative of the state vector, which represents the rate of change of the state over time; A is the state matrix, which describes the dynamic relationship between state variables; X(t) is the state variable, which is the minimum information set describing the current state of the system; B is the input matrix, which links the control input to the state change; U(t) is the control input vector, including power input and load changes; ω(t) is the process noise, which represents the internal random disturbance that the model cannot capture; V t (t) represents the voltage amplitude and phase angle of each node in the power system, P g (t) represents the output power of each generator in the system at time t; I l (t) represents the magnitude of the current flowing through each line in the system, including the real part and the imaginary part; P l (t) represents the total load in the system or the load of each node in the time period t; P r (t) represents the predicted power generation from wind and solar renewable energy sources; λ(t) represents the rate at which equipment in the system may fail; f(t) represents the operating frequency of the power system; S24. Establish observation equations to establish the relationship between state variables and observable variables. Observable variables include node voltage, line current, and power. The observation equations are expressed as a set of algebraic equations. The formula for the observation equations is: Z(t)=CX(t)+DU(t)+υ(t); X(t)=[V t (t),P g (t),I l (t),P l (t),P r (t),λ(t),f(t)]; Observation models allow estimating state variables from observable variables, integrating data from different sources, and by taking into account observation noise, observation models deal with the uncertainty in actual measurements; Z(t) is the observation vector, which contains a set of variables measured directly from the power system. C is the observation matrix, a coefficient matrix used to map state variables to the observation space. It defines how state variables affect observable variables. D is the direct transfer matrix, which is used to map control inputs directly to the observation space. The direct transfer matrix describes the direct impact of control inputs on observable variables. υ(t) is the observation noise, which represents random errors and external interference in the measurement process. The observation noise is Gaussian white noise with zero mean and a certain variance. S25. Construct a state space model. Combine the state equation and the observation equation to construct a state space model. The state space model formula is expressed as: is the derivative of the state vector, X(t) is the state variable, A is the state matrix, B is the input matrix, U(t) is the control input vector, ω(t) is the process noise, Z(t) is the observation vector, and υ(t) is the observation noise.

2. The method for quantitatively analyzing uncertainties of a grid-connected entity actively connected to a power distribution system according to claim 1, characterized in that: The key response indicators described in step S21 include: node voltage amplitude and phase angle, which reflect the voltage stability of the power system; generator output power, which indicates the supply capacity of the power supply; line flow, which reflects the condition of power flowing through the system; load fluctuation, which affects the supply and demand balance of the power grid; renewable energy power generation prediction accuracy, which affects power grid scheduling and planning; equipment failure rate, which affects the reliability and maintenance requirements of the system.

3. The method for quantitatively analyzing uncertainties of a grid-connected entity actively connected to a power distribution system according to claim 1, characterized in that: In step S3, a Monte Carlo simulation is performed based on the state space model established in step S2. The method for estimating the failure probability includes: S31. Sampling uncertainty factors, using the probability model established in step S1, randomly sampling each uncertainty factor, using a pseudo-random number generator to generate sample data according to the established probability model; S32. Set the number of simulations and use the relationship between the confidence interval and the uncertainty factor sample to estimate the required number of simulations and range. The number of simulations is estimated using the following formula: N s Estimated number of simulations; Z α / 2 is the value corresponding to the confidence level of the normal distribution; E m It is the acceptable error range of the simulation results, also known as the half-width of the confidence interval, indicating that it is hoped that the error between the simulation results and the true value does not exceed this range. m The smaller the value, the higher the requirement for simulation accuracy. Accordingly, the number of simulations required, N, is s The larger the N s refers to the dimension of the state vector, that is, the number of state variables; S33 simulates system operation, and for each set of sample data of uncertain factors, inputs the sample data into the state space model established in step S2 to obtain the predicted values ​​of the state variables and observed variables of the power system under the given uncertainty factor sample; S34. Determine whether the system has failed. Based on the preset failure criteria, determine whether the system has failed under the current uncertainty factor sample. The failure criteria judgment formula is: X(t)=[V t (t),P g (t),I l (t),P l (t),P r (t),λ(t),f(t)] S35. Count the number of failures and record the number of system failures; The number of system failures in all simulation samples in step S34 is counted, and the formula is expressed as: N s is the total number of simulations, F i is the failure logic variable of the i-th simulation; S36. Calculate the failure probability. Based on the number of failures and the total number of simulations, calculate the failure probability of the system under the influence of uncertain factors. The calculation formula is: N F is the number of failures, N s is the total number of simulations.

4. The method for quantitatively analyzing uncertainties of a grid-connected entity actively connected to a power distribution system according to claim 3, characterized in that: The failure conditions satisfied by X(t) include: V min ≤V i (t)≤V max ; P g,i,min (t)≤P g,i (t)≤P g,i,max (t); I l,i,max ≥I l,i (t); P l,min ≤P l (t)≤P l,max ; P r,min ≤P r (t)≤P r,max ; λ(t)≤λ max ; f min ≤f(t)≤f max ; V min and V max are the minimum and maximum safety thresholds of the node voltage, V i (t) is the voltage of each node; P g,i,min (t) and P g,i,max (t) are the minimum and maximum output power limits of the ith generator; I l,i,max is the power flow capacity of the i-th line; P l,min and P l,max are the minimum and maximum safety thresholds of the load, P r,min and P r,max are the minimum and maximum safety thresholds for renewable energy generation; λ max is the maximum safety threshold of equipment failure rate; f min and f max are the minimum and maximum safety thresholds for the system frequency.

5. The method for quantitatively analyzing uncertainties of a grid-connected entity actively connected to a power distribution system according to claim 1, characterized in that: In step S4, the Sobol method is used to evaluate the impact of parameter changes on the system, evaluate the impact of parameter changes in the state space model on the system output, and identify the uncertain factors that have the greatest impact on the system. The method includes: S41. Parameter definition: Determine the initial estimated range for each parameter based on historical data and the physical meaning of the model; S42. Perform a sensitivity analysis experiment. Based on the state-space model constructed in step S2, systematically change the value of each parameter of the state-space model, conduct a series of simulation experiments, and record the results. S43. Calculate the sensitivity index. Based on the experimental results of S42, calculate the sensitivity index to evaluate the impact of the parameters on the system. The calculation formula for the sensitivity index is: Among them, K i In the fixed parameter, Var(P F |K i |) is a fixed parameter K i Variance of failure probability at time t; Var(P F ) is the total variance of the failure probability when all parameters are varied; S44. Analyze the results and, based on the size of the sensitivity index, determine which uncertainties have the greatest impact on the system and which parameters have the greatest impact on the probability of system failure, that is, identify sensitive factors.

Images