Power insurance and supply toughness dynamic evaluation method considering extreme weather

Abstract

The invention relates to a dynamic evaluation method for power supply insurance toughness considering extreme weather, and belongs to the technical field of power system supply insurance. Comprising the following steps: based on historical extreme weather data, simulating an operation process of a power system under an extreme weather condition by utilizing a constructed power system operation model under an extreme weather background, and obtaining operation process simulation data of each time period; constructing a three-level evaluation index architecture of the power supply and insurance toughness; based on the operation process simulation data and the corresponding power system construction data, calculating to obtain a corresponding index value of each three-level evaluation index in each time period; based on the corresponding index values and the three-level evaluation index architecture, constructing a power supply insurance dynamic Bayesian evaluation network; and dynamically evaluating the power insurance and supply toughness based on the power insurance and supply dynamic Bayesian evaluation network. According to the method, the evaluation index indicating the operation state of the power system is subjected to probability conduction calculation to obtain the toughness evaluation value of the power system, and the toughness of the power system is accurately described and objectively evaluated.

Description

Technical Field

[0001] The present invention relates to the technical field of power system supply security, and in particular to a method for dynamically evaluating the resilience of power supply security considering extreme weather conditions. Background Art

[0002] As modern society becomes increasingly reliant on the power system, the security and stability of power supply are paramount. The frequent occurrence of extreme weather events such as typhoons, ice storms, and floods makes it difficult for traditional power systems to cope with the complex and volatile risk environment. As the scale of the power grid grows, the security of large-scale grids becomes crucial. Furthermore, with the rapid development of new energy sources such as wind power and photovoltaic power generation, the uncertain impact of extreme weather events on the power system is increasing. Therefore, accurate power supply resilience assessments can provide a basis for power supply strategies, ensuring the operation of critical infrastructure and maintaining basic living order during extreme weather events. This is crucial for guiding emergency response and recovery, and ensuring energy security.

[0003] Currently, domestic and international scholars studying power supply resilience assessment primarily rely on comprehensive evaluation methods such as fuzzy evaluation, analytic hierarchy process (AHP), principal component analysis (PCA), and Topsis. These methods screen power system indicators such as power capacity, grid line robustness, load demand, and energy storage capacity, and then calculate the system's power supply resilience value by subjectively or objectively weighting these indicators. However, due to the limitations of comprehensive evaluation methods, existing assessment methods primarily focus on static indicators and lack a dynamic characterization of power system performance under extreme weather events, making it difficult to accurately assess the dynamic changes in power supply resilience. Furthermore, because power supply risk is influenced by multiple factors, the temporal and spatial evolution of risk is complex and diverse, and the topological structures and connectivity between different network nodes are diverse, the transmission paths of risk events between different power supply nodes are highly uncertain. Existing assessment methods struggle to accurately characterize the impact of uncertainty on the power system. Summary of the Invention

[0004] In view of the above analysis, the present invention aims to disclose a dynamic assessment method for the resilience of power supply taking into account extreme weather conditions, use the power system operation model under the background of historical extreme weather to simulate the operation process of the power system, calculate the three-level assessment indicators of power supply resilience, construct a dynamic Bayesian assessment network for power supply, and use the dynamic Bayesian assessment network for power supply to achieve a refined dynamic resilience assessment of power supply for the entire spatiotemporal process of risk evolution under extreme weather conditions.

[0005] The present invention provides a dynamic assessment method for power supply resilience considering extreme weather conditions, which specifically includes the following steps:

[0006] Based on historical extreme weather data, the power system operation model constructed under extreme weather background is used to simulate the operation process of the power system under extreme weather conditions, and the operation process simulation data of each time period is obtained;

[0007] Construct a three-level evaluation indicator framework for power supply resilience; calculate the corresponding indicator values ​​of each three-level evaluation indicator in each time period based on operation process simulation data and corresponding power system construction data;

[0008] Constructing a dynamic Bayesian evaluation network for power supply security based on the corresponding indicator values ​​and the three-level evaluation indicator architecture;

[0009] The resilience of power supply is dynamically evaluated based on the power supply dynamic Bayesian evaluation network.

[0010] Furthermore, the construction of a dynamic Bayesian evaluation network for power supply security based on the corresponding indicator values ​​and the three-level evaluation indicator architecture includes:

[0011] A Bayesian network is constructed using the evaluation indicators and power supply resilience in the three-level evaluation indicator framework as nodes; the upper-level indicator is the parent node of the lower-level indicator; and power supply resilience is the upper-level indicator of the first-level indicator.

[0012] Constructing a dynamic Bayesian network based on the corresponding indicator values ​​of each three-level evaluation indicator in each time period and the Bayesian network;

[0013] The dynamic Bayesian network is optimized to obtain a dynamic Bayesian evaluation network for power supply security.

[0014] Furthermore, the constructing of a dynamic Bayesian network based on the corresponding indicator values ​​of each of the three-level evaluation indicators in each time period and the Bayesian network includes:

[0015] a. Based on the corresponding indicator values ​​of each third-level indicator in each time period, the weight of each third-level indicator relative to the second-level indicator and the node value of each third-level indicator in each time period are calculated;

[0016] b. Determine the probability transfer calculation formula of the third-level indicator node to the second-level indicator node based on the weight of each third-level indicator relative to the second-level indicator;

[0017] c. calculating the secondary indicator node value for each period based on the probability transfer calculation formula of the third-level indicator node to the secondary indicator node and the value of each third-level indicator node for each period;

[0018] d. Calculate the weight of each secondary indicator relative to the primary indicator based on the secondary indicator node value of each time period;

[0019] e. Based on the methods in steps b, c, and d, determine the probability transfer calculation formula for the secondary indicator node to the primary indicator node, the primary indicator node value for each time period, the weight of the primary indicator for power supply resilience, the probability transfer formula for the primary indicator node to the power supply resilience node, and the power supply resilience node value for each time period;

[0020] f. Determine the joint probability distribution function of any node in the dynamic Bayesian network based on the probability transfer calculation formula and the indicator node value of each time period;

[0021] g. Dynamic Bayesian network learning is performed based on the expectation maximization algorithm to obtain a constructed dynamic Bayesian network.

[0022] Furthermore, the dynamic evaluation of power supply resilience based on the power supply dynamic Bayesian evaluation network includes:

[0023] Based on given extreme weather scenario data, the power system operation process is simulated to obtain corresponding operation process simulation data; or the actual operation process data of the power system is obtained based on the network detection equipment of the power system;

[0024] Based on the corresponding operation process simulation data and the corresponding power system construction data, or based on the actual operation process data of the power system and the corresponding power system construction data, respectively calculate the corresponding indicator value of each third-level indicator in each time period;

[0025] The node value of the third-level indicator in each period is calculated based on the corresponding indicator value of each third-level indicator in each period;

[0026] The values ​​of the three-level indicator nodes in each time period are substituted into the dynamic Bayesian evaluation network for power supply security to dynamically evaluate the power supply resilience under given extreme weather scenarios or during the actual operation of the power system.

[0027] Furthermore, the operation process of the power system under extreme weather conditions is simulated by using the constructed power system operation model under extreme weather conditions based on historical extreme weather data, including:

[0028] Construct the power system operation model by constructing the power grid line icing model, photovoltaic panel icing model, wind turbine blade icing model, wind turbine output model, photovoltaic unit output model, thermal power unit output model, transmission line fault model, power load model and operation and dispatch model under extreme weather conditions;

[0029] Based on the data of each period in the historical extreme weather data, the power system operation model is used to simulate the operation process simulation data of the power system in each period.

[0030] Furthermore, the calculation method of the corresponding indicator value of each of the three-level evaluation indicators in each time period includes:

[0031] The total installed capacity is calculated based on the rated capacity of each unit;

[0032] The proportion of coal-fired power generation capacity is calculated based on the installed capacity of coal-fired power generation;

[0033] The new energy confidence output is calculated based on the new energy maximum output;

[0034] Calculate the capacity margin based on the maximum load;

[0035] The grid load rate is calculated based on the grid load;

[0036] Calculate line redundancy based on the number of backup lines;

[0037] The energy storage installed capacity is calculated based on the capacity of each energy storage unit;

[0038] The line failure rate is calculated based on the total length of the faulty line;

[0039] Obtain the load loss rate based on the power lost during the fault period;

[0040] The load loss rate is calculated based on the fault duration;

[0041] Power diversity is calculated based on the installed capacity ratio of each type of power supply;

[0042] The load recovery rate is calculated based on the restored load and the lost load;

[0043] The load recovery rate is calculated based on the load recovery time.

[0044] Furthermore, the operation scheduling model includes a scheduling objective function with the goal of minimizing the load supply and demand deviation and power generation cost, a power deviation constraint, a communication time constraint and a generator power balance constraint.

[0045] Furthermore, the scheduling objective function with the goal of minimizing the load supply and demand deviation and the power generation cost is expressed as:

[0046]

[0047] Where ΔP(t) is the load supply and demand deviation; C(t) is the power generation cost; α is the cost weight; P G (t), P L (t) are the generator output power and load demand respectively; a, b, and c are the secondary cost coefficient, primary cost coefficient, and constant cost coefficient of the generated power respectively.

[0048] Furthermore, the generator power balance constraint is expressed as:

[0049]

[0050] Among them, P G (t) represents the power generated by the generator; are the minimum and maximum power generation of the generator respectively; u(t-τ) is the control instruction of the power system network layer, which indicates the power adjustment amount. The larger τ is, the more delayed the system control is. When τ=0, the system responds in real time and the power supply and demand are balanced in real time.

[0051] Furthermore, the transmission line fault model includes:

[0052]

[0053] Among them, λ F (y) represents the transmission line fault, y is the ice thickness, d is the design ice thickness of the transmission line, A is the attenuation coefficient, and τ is the damping coefficient.

[0054] The present invention can achieve at least one of the following beneficial effects:

[0055] By constructing a three-level evaluation indicator framework for power supply resilience, the indicator values ​​of each three-level indicator reflecting the power grid operation status over time are calculated based on objective data during the operation of the power grid system. Then, through the three-level indicator values ​​and probability calculations, the low-level indicators representing the power grid operation status are passed layer by layer to calculate the power supply resilience value, which can accurately and objectively evaluate the power supply resilience.

[0056] By constructing a dynamic Bayesian assessment network for power supply security, focusing on the dynamic changes of various indicators under extreme weather conditions as the grid operates, and using a dynamic Bayesian network to calculate the dynamic evolution of the risk transfer process, we can characterize the impact of uncertainty in the risk process on the power system and accurately assess the dynamic changes in the resilience of power supply security. This helps to improve the ability to identify, assess, and respond to potential risks from a system-wide perspective, thereby ensuring the safety and adaptability of the power system under extreme events and improving the reliability and long-term stability of power supply security.

[0057] By using extreme weather scenarios to simulate the dynamic changes in the operating status of the power grid and adopting the dynamic Bayesian evaluation network for power supply security to dynamically evaluate indicators at all levels and the dynamic changes in power supply resilience, it is helpful to analyze the resilience of each network node in the power system, achieve targeted strengthening, and improve the safety and stability of the power system.

[0058] Other features and advantages of the present invention will be described in the following description, and some advantages may become apparent from the description or be understood through practice of the present invention. The objectives and other advantages of the present invention can be realized and obtained through the contents particularly pointed out in the description, claims and drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] The accompanying drawings are only for the purpose of illustrating particular embodiments and are not to be considered limiting of the present invention. Like reference symbols denote like parts throughout the drawings.

[0060] Figure 1 Flow chart of the method of the present invention;

[0061] Figure 2 、 Figure 3 To simulate the changing process of power system resilience under given extreme weather scenario data based on GeNIe software. DETAILED DESCRIPTION

[0062] The preferred embodiments of the present invention will be described in detail below in conjunction with the accompanying drawings, wherein the accompanying drawings constitute a part of this application and are used together with the embodiments of the present invention to illustrate the principles of the present invention, and are not used to limit the scope of the present invention.

[0063] One embodiment of the present invention discloses a method for dynamically evaluating the resilience of power supply considering extreme weather conditions, specifically comprising steps S1-S4 (e.g. Figure 1 It should be noted that the extreme weather in the present invention is extreme cold wave weather.

[0064] Step S1: Based on historical extreme weather data, the power system operation model under extreme weather background is used to simulate the operation process of the power system under extreme weather conditions to obtain operation process simulation data for each time period.

[0065] S11. Construct the power system operation model by constructing the power grid line icing model, photovoltaic panel icing model, wind turbine blade icing model, wind turbine output model, photovoltaic unit output model, thermal power unit output model, transmission line fault model, power load model and operation scheduling model under extreme weather conditions.

[0066] Specifically, ice disasters caused by extreme cold waves can cause ice to cover the overhead wires and towers of the power system. As the impact of the ice disaster continues, the thickness of the ice will become increasingly larger, which will not only affect the light reception efficiency of the photovoltaic panels and the speed of the wind turbines, but also affect the wind and solar processing. In severe cases, it may cause accidents such as overhead wire breakage and tower collapse. Therefore, it is necessary to build grid line ice coverage models, photovoltaic panel ice coverage models, and wind turbine blade ice coverage models for ice disaster scenarios.

[0067] Specifically, in the ice disaster scenario, the general model of ice thickness is:

[0068]

[0069] Where h(t) represents the ice thickness (mm), v(t) represents the wind speed (m / s), which is generally 1-20m / s; RH(t) represents the relative humidity (%), which is generally 85-100%; T represents the ambient temperature (°C), and the optimal ice temperature is generally -1°C to -5°C; η represents the ice efficiency coefficient, which is generally 0.01-0.1mm / (m·s·%); λ represents the ablation rate coefficient, which is generally 0.05-0.2mm / (°C·h), and T m represents the critical melting temperature, which is generally 0°C; f(T) represents the temperature influence function, which is expressed as follows:

[0070]

[0071] Where: t b The optimal ice-covering temperature indicates the temperature at which the freezing efficiency is highest. k represents the temperature influence coefficient, which is generally between 0 and 1. When the temperature approaches 0°C, the freezing capacity decays rapidly.

[0072] Furthermore, the grid line icing model is expressed as:

[0073]

[0074] Where: h g Indicates the ice thickness of the power line, v ⊥ (t) = v(t) sinθ, which represents the vertical component of the wind speed along the power grid line, θ is the wind direction angle, η g ,λ g are the power grid line icing efficiency coefficient and ablation coefficient respectively.

[0075] Furthermore, the photovoltaic panel ice coverage model is expressed as:

[0076]

[0077] Where h p represents the thickness of ice covering the photovoltaic panel, η p ,λ p are the icing efficiency coefficient and ablation coefficient of the photovoltaic panel, θ is the inclination angle of the photovoltaic panel (the horizontal plane is 0°), and I(t) represents the solar irradiance.

[0078] Furthermore, the wind turbine blade icing model is expressed as:

[0079]

[0080] Where η w , δ represent the wind turbine blade icing efficiency coefficient and ablation coefficient respectively, v r (t) represents the relative wind speed of the blade, and ω(t) represents the angular velocity of the blade rotation (ice accumulation is fastest when the blade stops).

[0081]

[0082] Where, v l (t) represents the linear speed of the fan blade;

[0083] v l (t) = ω(t)·R = λ·v(t);

[0084] R is the blade radius, and λ represents the tip linear speed ratio of the fan, which is determined by the material of the fan. For lift-type wind turbines, it is generally 3 to 8.

[0085] Specifically, the wind turbine output model is expressed as:

[0086]

[0087] Where, P rated Indicates rated power, ρ is air density (kg / m 3 ), R eff is the effective radius of the fan blade (m), is the effective power generation efficiency coefficient, R eff and Respectively expressed as:

[0088] R eff =R·(1-γH ice );

[0089]

[0090] R represents the original radius of the fan blade (m), C p0 represents the fan efficiency coefficient when there is no icing, γ and β represent the attenuation coefficient of icing on radius and efficiency (1 / m), H ice Indicates the ice thickness of wind turbine blades (m).

[0091] Furthermore, the constraint on wind turbine output is expressed as:

[0092]

[0093] H ice ≤H max ;

[0094] Where, v in is the cut-in wind speed; v out is the cut-out wind speed; v rate is the rated wind speed; H max The maximum ice thickness.

[0095] Specifically, the photovoltaic unit output model is expressed as:

[0096]

[0097] Where, P stc is the rated output (kW) under standard test conditions, S eff (t) represents the actual light intensity (W / m 2 ), S stc Indicates the light intensity under STC (default 1000W / m 2 ), T is the real-time ambient temperature (℃), T stc It represents the temperature under STC (default is 25℃), and γ is the temperature coefficient (1 / ℃, usually -0.005).

[0098] Furthermore, the actual light intensity formula is as follows:

[0099] S eff (t) = S(t)·(1-αH ice (t));

[0100] Where S(t) represents the real-time light intensity (W / m 2 ), α is the light attenuation coefficient caused by ice (1 / mm), H ice (t) represents the thickness of ice covering the photovoltaic panel (mm).

[0101] Furthermore, using the Beta distribution probability density function f(S t ) describes the actual light intensity S(t):

[0102]

[0103]

[0104]

[0105] Where S max is the maximum value of light intensity; α and β are the shape parameters of Beta distribution; μ PV is the average light intensity; σ PV is the standard deviation of light intensity.

[0106] Specifically, thermal power units include coal-fired units and gas-fired units. The output model of thermal power units is expressed as:

[0107]

[0108]

[0109] Where: F coal,t 、F gas,t are coal and gas consumption respectively, P i,t Indicates the real-time output of the unit, a c 、bc 、c c and a g 、b g 、c g are the characteristic parameters of coal-fired and gas-fired units respectively.

[0110] Specifically, the transmission line fault model is expressed as:

[0111]

[0112] Among them, λ F (y) represents the transmission line fault, y is the ice thickness, d is the design ice thickness of the transmission line, A is the attenuation coefficient, and τ is the damping coefficient.

[0113] Specifically, the power load model is expressed as:

[0114] L=L b +L w ;

[0115] Where, L is the total load; L b It is the basic load, reflecting the overall development trend of the load over a long period of time, and has certain stability, periodicity and seasonality; L w It is a climate-sensitive load that reflects the impact of climate factors (such as temperature, humidity, rainfall, wind speed, cloud cover, etc.) on the load.

[0116] It should be noted that when the present invention simulates the operation process of the power system under extreme weather conditions, it directly adopts the load curve corresponding to the historical extreme weather data, including the total load, base load and climate-sensitive load.

[0117] Specifically, the operation scheduling model includes a scheduling objective function with the goal of minimizing the load supply and demand deviation and power generation cost, as well as constraints including power deviation constraints, communication time constraints and generator power balance constraints.

[0118] The scheduling objective function with the goal of minimizing the load supply and demand deviation and power generation cost is expressed as:

[0119]

[0120] Where ΔP(t) is the load supply and demand deviation; C(t) is the power generation cost; α is the cost weight; P G (t), P L (t) are the generator output power and load demand respectively; a, b, and c are the secondary cost coefficient, primary cost coefficient, and constant cost coefficient of the generated power respectively.

[0121] The generator power balance constraint is expressed as:

[0122]

[0123] Among them, P G (t) represents the power generated by the generator; are the minimum and maximum power generation of the generator respectively; u(t-τ) is the control instruction of the power system network layer, which indicates the power adjustment amount. The larger τ is, the more delayed the system control is. When τ=0, the system responds in real time and the power supply and demand are balanced in real time.

[0124] The power deviation constraint is expressed as:

[0125] ΔP(t)≤Δ(t) max ;

[0126] Where: Δ(t) max Indicates the maximum power deviation that the system can handle, which is generally determined by the system's effective capacity, capacity-load ratio, and other factors.

[0127] The communication time constraint is used to represent the communication time delay of the power system dispatch instruction, which is expressed as:

[0128] τ≤τ max (τ≥0);

[0129] Where: τ represents the time delay, τ max The maximum allowed time delay.

[0130] S12. Based on the data of each time period in the historical extreme weather data, the power system operation model is used to simulate and obtain the operation process simulation data of the power system in each time period.

[0131] The operation process simulation data includes the maximum output of renewable energy in each period, the maximum load of the power system, the total length of the fault line, the power loss during the fault, the load demand power, the restored load, the lost load, the load recovery time, etc.

[0132] Specifically, based on the data of each time period in the historical extreme weather data and the corresponding historical load curves, the power system operation model is used to simulate the operation conditions of each unit during extreme weather (including unit output and start and stop), icing conditions and fault conditions, etc., to obtain the simulation data of the power system operation process in each time period.

[0133] Step S2: Construct a three-level evaluation indicator framework for power supply resilience; calculate the corresponding indicator values ​​of each three-level evaluation indicator in each time period based on the operation process simulation data and the corresponding power system construction data.

[0134] Step S21: Construct a three-level evaluation indicator framework for power supply resilience.

[0135] Specifically, the three-level evaluation indicator framework for power supply resilience includes multiple first-level, second-level, and third-level indicators. The corresponding relationships between each indicator and the indicators are shown in Table 1.

[0136] Table 1 Three-level evaluation indicator framework for power supply resilience

[0137]

[0138]

[0139] Step S22: Calculate the corresponding index values ​​of each third-level evaluation index in each time period based on the operation process simulation data and the corresponding power system construction data.

[0140] The total installed capacity is calculated based on the rated capacity of each unit in the power system construction:

[0141]

[0142] Where: P i (t) is the rated capacity of the i-th generator set in period t, which is the power system construction data.

[0143] The proportion of coal-fired power generation capacity is calculated based on the coal-fired power generation capacity of the power system construction:

[0144]

[0145] Where: P coal (t) is the installed capacity of coal-fired power generation in period t, and is the power system construction data.

[0146] The new energy confidence output is calculated based on the new energy maximum output obtained by simulation;

[0147] P renew (t) = P renew_max (t)×α;

[0148] Where: P renew (t), P renew_max (t) are the actual and maximum output of new energy in period t, P renew_max (t) is the simulation data of the operation process, and α is the confidence level coefficient.

[0149] The capacity margin M(t) is calculated based on the maximum load obtained by simulation:

[0150]

[0151] Where: P max_load (t) is the maximum load during period t, which is the simulation data of the operation process.

[0152] The grid load rate L(t) is calculated based on the grid load of the power system construction:

[0153]

[0154] Where: P actual (t) is the grid load during period t, and is the power system construction data.

[0155] The line redundancy R is calculated based on the number of backup lines in the power system construction. line (t):

[0156]

[0157] Where: N backup (t) is the number of backup lines in period t, N total_line (t) is the total number of lines in period t, all of which are power system construction data.

[0158] The energy storage installed capacity E is calculated based on the capacity of each energy storage unit in the power system construction. storage (t):

[0159]

[0160] Where: E j (t) is the capacity of the j-th energy storage unit in period t, which is the power system construction data.

[0161] The grid line failure rate λ is calculated based on the total length of the fault line obtained by simulation line (t):

[0162]

[0163] Where: N fault_line (t) is the total length of the fault line during period t, is the simulation data of the operation process, L total (t) is the total length of the line in period t, which is the power system construction data.

[0164] The load loss ratio LOLR(t) is obtained based on the power loss during the fault period obtained by simulation:

[0165]

[0166] Where: E lost (t) is the power loss during the fault period t, E demand (t) is the load demand power during period t, which are all simulation data of the operation process.

[0167] The load loss rate R is calculated based on the fault duration obtained by simulation. loss (t):

[0168]

[0169] Where: t fault(t) is the fault duration in period t, which is the simulation data of the operation process.

[0170] The power diversity H(t) is calculated based on the installed capacity ratio of various power sources in the power system construction:

[0171]

[0172] Where: p k (t) is the installed capacity proportion of the kth type of power source in period t, which is the power system construction data.

[0173] The load recovery rate R is calculated based on the restored load and lost load obtained by simulation load (t):

[0174]

[0175] Where: P recovered (t), P lost (t) are the restored load and lost load in period t, which are the simulation data of the operation process.

[0176] The load recovery rate v is calculated based on the load recovery time obtained by simulation recovery (t):

[0177]

[0178] Where: t recovery (t) is the load recovery time in period t, and is the simulation data of the operation process.

[0179] The above is a method for calculating the corresponding index values ​​for each of the three-level evaluation indicators in each time period based on the operation process simulation data and the corresponding power system construction data. It should be noted that because the power system construction data corresponding to each time period of historical extreme weather data may be incomplete, the calculated corresponding index values ​​for each time period are also incomplete. The index values ​​need to be filled with data, so step S2 also includes step S23.

[0180] Step S23: Fill in the missing data in the corresponding indicator values ​​of each time period based on the improved KNN algorithm, which specifically includes S231 to S234.

[0181] Specifically, the corresponding indicator value of each period is used as a data sample, expressed as x i =(x i1 ,x i2 ,...,x iM), where i represents the sample number, M represents the number of corresponding indicator values ​​of the three-level evaluation indicators in each period, and each indicator value in the sample is called a feature. Furthermore, all samples constitute the indicator data set D = {x1, x2, ..., x N}, N is the number of samples.

[0182] S231: feature standardization processing.

[0183] Specifically, the following formula is used to calculate each feature x in the sample ij Perform standardization to eliminate the dimension effect and obtain the standardized eigenvalue x′ ij :

[0184]

[0185] Among them, μ j , σ j are the mean and standard deviation of feature j, respectively.

[0186] S232. Calculate the dynamic weighted distance between samples.

[0187] Use the following formula to calculate the two samples x i and x l Dynamic weighted distance between:

[0188]

[0189] Among them, m represents the feature number in the sample, M i For sample x i The set of non-missing features, w m is the weight of feature m, which is determined by information gain.

[0190] S233. Calculate the dynamic K value in the improved KNN algorithm.

[0191] Specifically, for each sample x i , calculate its local density ρ to the nearest neighbor i :

[0192]

[0193] Among them, K base is the basic nearest neighbor number, d(·) is the distance function, and the higher the density, the denser the sample is.

[0194] Furthermore, based on the local density ρ i Dynamically adjust the K value:

[0195]

[0196] Among them, K min , Kmax is the preset minimum and maximum K value, ρ min , ρ max are the minimum and maximum values ​​of the local density in the indicator dataset.

[0197] S234, missing value filling.

[0198] Specifically, for x i The missing feature j, select K nearest neighbors N K (i) Calculate the weighted mean, i.e. x′ ij , based on x′ ij Calculate x i The eigenvalues ​​of missing feature j are:

[0199]

[0200] Through steps S231 to S234 above, the problem of incomplete corresponding indicator values ​​in each time period is solved.

[0201] Furthermore, in calculating the dynamic weighted distance between samples in step S232, it is also necessary to determine the weight of feature m. Therefore, step S232 also includes step S232-a for determining the weight of feature m.

[0202] Step S232-a: Determine the weight of feature m (ie, the mth indicator in each sample) based on information gain.

[0203] Specifically, based on the indicator data set D, samples with no missing feature values ​​are selected to form the indicator data set D'. Based on D', the weight of each feature is determined using information gain.

[0204] Specifically, the weight of each feature in the sample refers to the weight of each third-level indicator relative to the second-level indicator.

[0205] Furthermore, the formula for information gain is:

[0206] IG(Y,Xm)=H(Y)-H(Y|Xm);

[0207] Where IG is information gain, H(Y) is the entropy (for classification tasks) or variance (for regression tasks) of the target variable Y, and H(Y|Xm) is the conditional entropy or conditional variance of Y given feature Xm. Here, the target variable represents a secondary indicator. Given features correspond to each of the third-level indicators.

[0208] Furthermore, since the index values ​​at all levels in this application are continuous variables, the regression task is used for example for calculation, H(Y)=Var(Y), in It represents the expected value of the conditional variance of Y under the condition that the feature Xm takes the value x.

[0209] Further, calculate the normalized weight where w m is the weight of feature m, that is, the weight of each third-level indicator relative to the second-level indicator.

[0210] Step S3: constructing a dynamic Bayesian evaluation network for power supply security based on the corresponding indicator values ​​and the three-level evaluation indicator architecture, specifically including S31 to S33.

[0211] S31. Construct a Bayesian network using the evaluation indicators and power supply resilience in the three-level evaluation indicator architecture as nodes; the upper-level indicator is the parent node of the lower-level indicator; and the power supply resilience is the upper-level indicator of the first-level indicator.

[0212] S32: Constructing a dynamic Bayesian network based on the corresponding indicator values ​​of each three-level evaluation indicator in each time period and the Bayesian network, including steps a to g.

[0213] a. Calculate the weight of each third-level indicator relative to the secondary indicator and the node value of each third-level indicator in each period based on the corresponding indicator value of each third-level indicator in each period;

[0214] b. Determine the probability transfer calculation formula of the third-level indicator node to the second-level indicator node based on the weight of each third-level indicator relative to the second-level indicator;

[0215] c. calculating the secondary indicator node value for each period based on the probability transfer calculation formula of the third-level indicator node to the secondary indicator node and the value of each third-level indicator node for each period;

[0216] d. Calculate the weight of each secondary indicator relative to the primary indicator based on the secondary indicator node value of each time period;

[0217] e. Based on the methods in steps b, c, and d, determine the probability transfer calculation formula for the secondary indicator node to the primary indicator node, the primary indicator node value for each time period, the weight of the primary indicator for power supply resilience, the probability transfer formula for the primary indicator node to the power supply resilience node, and the power supply resilience node value for each time period;

[0218] Specifically, the probability transfer calculation formula of the secondary indicator node to the primary indicator node is determined based on the weight of each secondary indicator relative to the primary indicator;

[0219] The first-level indicator node value of each time period is calculated based on the probability transfer calculation formula of the second-level indicator node to the first-level indicator node and the values ​​of each second-level indicator node in each time period;

[0220] The weight of each first-level indicator relative to the resilience of power supply is calculated based on the first-level indicator node value in each time period;

[0221] Based on the weight of each first-level indicator relative to the resilience of power supply, the probability transfer calculation formula of the first-level indicator node to the resilience of power supply is determined;

[0222] The power supply resilience node value for each period is calculated based on the probability transfer calculation formula of the first-level indicator node for power supply resilience and the first-level indicator node value for each period;

[0223] f. Determine the joint probability distribution function of any node in the dynamic Bayesian network based on the probability transfer calculation formula and the indicator node value of each time period;

[0224] g. Dynamic Bayesian network learning is performed based on the expectation maximization algorithm to obtain a constructed dynamic Bayesian network.

[0225] Specifically, in step a, the weight of each third-level indicator relative to the second-level indicator has been calculated in step S232-a.

[0226] Furthermore, the corresponding indicator values ​​of each three-level evaluation indicator in each time period are normalized to obtain the node values ​​of each three-level indicator in each time period.

[0227] Specifically, the total installed capacity, the proportion of coal-fired power installed capacity, the installed capacity of energy storage, the grid line failure rate, the load loss rate, the load loss rate, and the power source diversity are considered as negative indicators, and the normalized formula is expressed as follows: Among them, y max is the maximum value among all indicators.

[0228] Considering the new energy confidence output, load recovery rate, load recovery speed, and grid load rate as negative indicators, the normalized formula is expressed as:

[0229]

[0230] Where M = max{|y1-y best |,|y2-y best |,...,|y n -y best |};y best The optimal value specified for each indicator.

[0231] Considering capacity margin and line redundancy as interval indicators, the normalized formula is expressed as:

[0232]

[0233] Among them, for a set of interval indicator values ​​y1,y2,…,y n ,y max =max{y1,y2,…,y n},y min=min{y1,y2,…,y n}; Propose the optimal interval (a, b), M = max{ay min ,y max -b}.

[0234] From the above, it is found that the node values ​​of each third-level indicator in each period retain the original characteristics and range from 0 to 1.

[0235] Specifically, in step b, the probability transfer calculation formula of the third-level indicator node to the second-level indicator node is expressed as:

[0236]

[0237] Among them, A corresponds to the secondary indicator node; X i Corresponding to the third-level indicator node.

[0238] Specifically, the derivation process of the probability transfer calculation formula (1) is described below:

[0239] Take events A and B as an example, where event A is a finite number of events A1, A2, A3, ..., A n The set of , satisfies the following probability:

[0240]

[0241]

[0242]

[0243] Furthermore, in the Bayesian network, each indicator node can be defined as a complementary event, with event A, its complementary event a, and event B i , its complementary event b i For example, event A is represented as a secondary indicator node, and each event B i Indicates the third-level indicator node, event B i They are independent of each other and their own state probabilities are known, and they meet the following conditions: P(A|X i )=P(a|x i )=1-P(a|x i )=1-P(A|x i )=W i ;W i Represents the transfer probability weight between a single child node (third-level indicator node) and the parent node (second-level indicator node), that is, the weight w of each third-level indicator relative to the second-level indicator m .

[0244] Each event B iConsidered as an independent estimate of event A, the weighted average of all estimates is taken as the final solution, and the total probability formula is obtained:

[0245]

[0246] Furthermore, the conditional probability P(A|B i ) and P(A|b i ) is converted into the conditional probability P(B) in the Bayesian network i |A) and P(b i |A):

[0247] When event B i When it happens,

[0248] When event b i When it occurs:

[0249] Assume i=4, and take events B1, B2, B3, and B4 as examples. Use Bayes’ theorem to calculate the posterior probability:

[0250]

[0251] Among them, X i Corresponding to B i or b i .

[0252] Furthermore, the above posterior probability formula is expanded into a general transfer probability calculation formula for Bayesian network nodes. When i=n, ​​event A and event X i The conduction probability between is expressed as follows:

[0253]

[0254] Specifically, in step c, the secondary indicator node value of each time period is calculated based on the probability transfer calculation formula of the third-level indicator node to the second-level indicator node (i.e., formula (1)) and the values ​​of each third-level indicator node in each time period (step a).

[0255] Specifically, in step d, the weight of each secondary indicator relative to the primary indicator is calculated based on the secondary indicator node value of each time period according to the method of step S232-a.

[0256] Specifically, in step e, repeat methods b, c, and d to obtain the probability transfer calculation formula of the secondary indicator node to the primary indicator node, the primary indicator node value in each time period, the weight of the primary indicator to the power supply resilience, the probability transfer formula of the primary indicator node to the power supply resilience node, and the power supply resilience node value in each time period.

[0257] Specifically, in step f, the Bayesian network for each time period is expressed as BN = (G, θ), where G is a directed acyclic graph of the joint probability distribution on the node X of the Bayesian network, θ represents the parameters of the network, and the joint probability distribution on the node X is defined as: Where Pa(X i ) represents node X i The parent node set that directly affects X i Node.

[0258] Furthermore, the dynamic Bayesian network model extends this expression to random processes with time factors. A dynamic Bayesian network model can be defined as (B0, B → ). Where Bo represents the initial BN, and the prior probability P(X0) of any node can be obtained from the directed acyclic graph. → , represents a graph composed of BNs with more than two time segments.

[0259] Use P(X t ∣X t-1 ) represents the probability of the current state occurring when the state of any variable at the previous moment is known. represents the value of the i-th variable at time t, Represents its parent node, and N represents the existence of N variables. When there are only two time segments, there are:

[0260]

[0261] Furthermore, the joint distribution probability of any node in the dynamic Bayesian network model is calculated:

[0262]

[0263] Specifically, in step g, dynamic Bayesian network learning is performed based on the expectation maximization algorithm (EM algorithm) to obtain a constructed dynamic Bayesian network.

[0264] Specifically, the EM algorithm is an iterative algorithm for maximum likelihood estimation or maximum a posteriori probability estimation of probabilistic models with latent variables. It is used to address the challenges of limited data and chaotic dataset relationships when constructing dynamic Bayesian networks. The basic idea is: if the parameters θ of the distribution that the sample follows are known, the expected value of the latent variable Z can be inferred from the observed training samples (step E). If the value of Z is known, a new value of θ is estimated using maximum likelihood (step M). This process is repeated until the values ​​of Z and θ no longer change. Here, Y represents the observed variable data, Z represents the latent variable data, and θ represents the model parameters.

[0265] Specifically, the iterative steps of the EM algorithm include:

[0266] In step E, the posterior probability of the latent variable is calculated based on the initial value of the parameter or the model parameter of the previous iteration, which is also the expectation of the latent variable. As the first estimate of the latent variable:

[0267]

[0268] Among them, θ (i) is the current estimate of the parameter.

[0269] In the M step, the likelihood function is maximized to obtain the new parameter value:

[0270]

[0271] The EM algorithm improves the likelihood function value of the observed data after each iteration, that is, P(Y|θ (i+1) )≥P(Y|θ (i) ).

[0272] Specifically, when the maximum number of iterations is reached, the value of the network parameter θ is determined to obtain a constructed dynamic Bayesian network.

[0273] S33: Optimize the dynamic Bayesian network to obtain a dynamic Bayesian evaluation network for power supply security. Step S33 is used to optimize the dynamic Bayesian network to reduce the confidence bias in the dynamic Bayesian network generation process and increase the efficiency of the dynamic Bayesian network construction.

[0274] Specifically, the constructed dynamic Bayesian network model is assumed to be an independent causal influence model, meaning that each parent node has a different causal mechanism for its child nodes. The Noisy-Or model is a causal mechanism independent model whose nodes must be binary variables. The Noisy-MAX model can be applied to multi-valued variable models, expanding the state space of the Noisy-Or model variables. By introducing n auxiliary nodes Z into the dynamic Bayesian network, the Bayesian network model with a large number of parent nodes is converted into a causal Noisy-MAX model.

[0275] In the Noisy-MAX model, the child node Y must be an ordinal variable, with the smallest value indicating no abnormality and the higher value indicating a serious abnormality. Y The state values ​​are sequentially numbered {0,1,…,n Y -1}. And the parent node X i It does not need to be a sequential variable. The state of the intermediate node Z is consistent with that of the X node, and its state value is determined by the maximum influence of each cause node, that is,

[0276] y=f max (z)=max(z1,z2,…,z n );

[0277] To calculate the conditional probability of the Noisy-MAX model, first calculate P(Y≤y|X). To calculate the conditional probability of the Noisy-MAX model, define the cumulative parameter get

[0278]

[0279] Therefore, each parameter in the conditional probability table can be calculated by the following formula:

[0280]

[0281] Furthermore, considering that the Bayesian network model cannot include all influencing factors and is prone to confidence bias, the leaky probability is increased on the basis of the Noisy-MAX model, and the variable Z is introduced. L Represents a set of factors that are not explicitly included in the model, which is expanded into the Leaky Noisy-MAX model. L By Z L The probability of influencing node Y is recorded as get The final conditional probability calculation formula of the LeakyNoisy-MAX model is:

[0282] P(y|X)=P(Y≤y|X)-P(Y≤y-1|X);

[0283]

[0284]

[0285] It should be noted that by optimizing the dynamic Bayesian network in step S33, a dynamic Bayesian evaluation network for power supply security is obtained, which can realize the transmission probability between the third-level, second-level, and first-level indicator nodes, and the numerical values ​​of the third-level indicators are transmitted layer by layer through the transmission probability to obtain the final power supply resilience node value.

[0286] Step S4: Dynamically evaluate the resilience of power supply security based on the power supply security dynamic Bayesian evaluation network.

[0287] Specifically, based on historical extreme weather data and corresponding power system construction data, we developed a dynamic Bayesian assessment network for power supply security through steps S1-S3. In practical research and application, further research is needed to determine the resilience of power supply security under the most severe extreme weather scenarios.

[0288] For the worst extreme weather scenario, given the corresponding extreme weather scenario data, the dynamic evaluation of power supply resilience based on the power supply dynamic Bayesian evaluation network includes:

[0289] Based on the given extreme weather scenario data, the power system operation process is simulated to obtain the corresponding operation process simulation data;

[0290] Calculating corresponding indicator values ​​of each third-level indicator in each time period based on the corresponding operation process simulation data and corresponding power system construction data;

[0291] The node value of the third-level indicator in each period is calculated based on the corresponding indicator value of each third-level indicator in each period;

[0292] The values ​​of the three-level indicator nodes in each time period are substituted into the dynamic Bayesian evaluation network for power supply security to dynamically evaluate the resilience of power supply security under given extreme weather scenarios.

[0293] like Figure 2 、 Figure 3 Based on GeNIe software, the changing process of power system supply resilience under given extreme weather scenario data is simulated. Figure 2 As shown in the figure, the values ​​of the third-level indicator nodes are passed layer by layer to the second-level indicator nodes and the first-level indicator nodes, and finally the dynamic changes of the power supply resilience value are simulated. The third-level dynamic indicator represents the dynamic changes of the third-level indicator node values. Figure 3 As shown in the figure, during the evolution process, the system resilience values ​​were calculated at five time points: 0.7564, 0.6061, 0.5571, 0.5491, and 0.5435, representing the power supply resilience values ​​under severe weather conditions. A higher power supply resilience value indicates a stronger power system's ability to maintain power supply, reflecting its ability to maintain power supply under severe weather conditions.

[0294] Specifically, in actual production implementation, the physical equipment layer and network system layer of the power system realize real-time monitoring, control and optimization of physical equipment such as power generation equipment, transmission lines, and substations through sensors, actuators and network communication technologies. Through various cutting-edge technologies such as data warehouse technology, satellite remote sensing technology and sensor network technology, various data closely related to the power system, including historical operation data, meteorological data, and equipment status data, are comprehensively and systematically collected.

[0295] In actual production implementation, the dynamic evaluation of power supply resilience based on the power supply dynamic Bayesian evaluation network includes:

[0296] The actual operation process data of the power system is obtained based on the network detection equipment of the power system;

[0297] Based on the actual operation process data of the power system and the corresponding power system construction data, the corresponding index values ​​of each three-level index in each time period are calculated;

[0298] The node value of the third-level indicator in each period is calculated based on the corresponding indicator value of each third-level indicator in each period;

[0299] The values ​​of the three-level indicator nodes in each time period are substituted into the dynamic Bayesian evaluation network for power supply security to dynamically evaluate the power supply resilience during the actual operation of the power system.

[0300] This embodiment discloses a dynamic assessment method for power supply resilience taking into account extreme weather. By constructing a three-level assessment indicator framework for power supply resilience, the indicator values ​​of the three-level indicators reflecting the power grid operation status over time periods are calculated based on objective data during the operation of the power grid system. Then, through the values ​​of the three-level indicators and probability calculations, the low-level indicators representing the power grid operation status are transmitted layer by layer to calculate the power supply resilience value, which can accurately and objectively assess the power supply resilience.

[0301] By constructing a dynamic Bayesian assessment network for power supply security, focusing on the dynamic changes of various indicators under extreme weather conditions as the grid operates, and using a dynamic Bayesian network to calculate the dynamic evolution of the risk transfer process, we can characterize the impact of uncertainty in the risk process on the power system and accurately assess the dynamic changes in the resilience of power supply security. This helps to improve the ability to identify, assess, and respond to potential risks from a system-wide perspective, thereby ensuring the safety and adaptability of the power system under extreme events and improving the reliability and long-term stability of power supply security.

[0302] By using extreme weather scenarios to simulate the dynamic changes in the operating status of the power grid and adopting the dynamic Bayesian evaluation network for power supply security to dynamically evaluate indicators at all levels and the dynamic changes in power supply resilience, it is helpful to analyze the resilience of each network node in the power system, achieve targeted strengthening, and improve the safety and stability of the power system.

[0303] The above description is only a preferred specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions that can be easily thought of by any technician familiar with this technical field within the technical scope disclosed in the present invention should be covered by the scope of protection of the present invention.

Claims

1. A dynamic assessment method for power supply resilience considering extreme weather conditions, characterized by: The steps include: Based on historical extreme weather data, the power system operation model constructed under extreme weather background is used to simulate the operation process of the power system under extreme weather conditions, and the operation process simulation data of each time period is obtained; Construct a three-level evaluation indicator framework for power supply resilience; Based on the operation process simulation data and the corresponding power system construction data, the corresponding index values ​​of each three-level evaluation index in each time period are calculated; Constructing a dynamic Bayesian evaluation network for power supply security based on the corresponding indicator values ​​and the three-level evaluation indicator architecture; The resilience of power supply is dynamically evaluated based on the power supply dynamic Bayesian evaluation network.

2. The method for dynamic assessment of power supply resilience according to claim 1 is characterized in that: The construction of a dynamic Bayesian evaluation network for power supply security based on the corresponding indicator values ​​and the three-level evaluation indicator architecture includes: A Bayesian network is constructed using the evaluation indicators and power supply resilience in the three-level evaluation indicator framework as nodes; the upper-level indicator is the parent node of the lower-level indicator; and power supply resilience is the upper-level indicator of the first-level indicator. Constructing a dynamic Bayesian network based on the corresponding indicator values ​​of each three-level evaluation indicator in each time period and the Bayesian network; The dynamic Bayesian network is optimized to obtain a dynamic Bayesian evaluation network for power supply security.

3. The method for dynamic assessment of power supply resilience according to claim 2 is characterized in that: The constructing of a dynamic Bayesian network based on the corresponding indicator values ​​of each three-level evaluation indicator in each time period and the Bayesian network includes: a. Based on the corresponding indicator values ​​of each third-level indicator in each time period, the weight of each third-level indicator relative to the second-level indicator and the node value of each third-level indicator in each time period are calculated; b. Determine the probability transfer calculation formula of the third-level indicator node to the second-level indicator node based on the weight of each third-level indicator relative to the second-level indicator; c. calculating the secondary indicator node value for each period based on the probability transfer calculation formula of the third-level indicator node to the secondary indicator node and the value of each third-level indicator node for each period; d. Calculate the weight of each secondary indicator relative to the primary indicator based on the secondary indicator node value of each time period; e. Based on the methods in steps b, c, and d, determine the probability transfer calculation formula for the secondary indicator node to the primary indicator node, the primary indicator node value for each time period, the weight of the primary indicator for power supply resilience, the probability transfer formula for the primary indicator node to the power supply resilience node, and the power supply resilience node value for each time period; f. Determine the joint probability distribution function of any node in the dynamic Bayesian network based on the probability transfer calculation formula and the indicator node value of each time period; g. Performing dynamic Bayesian network learning on the dynamic Bayesian network obtained in step f based on the expectation maximization algorithm to obtain a constructed dynamic Bayesian network.

4. The method for dynamic assessment of power supply resilience according to claim 3 is characterized in that: The dynamically evaluating the resilience of power supply security based on the power supply security dynamic Bayesian evaluation network includes: Based on given extreme weather scenario data, the power system operation process is simulated to obtain corresponding operation process simulation data; or the actual operation process data of the power system is obtained based on the network detection equipment of the power system; Based on the corresponding operation process simulation data and the corresponding power system construction data, or based on the actual operation process data of the power system and the corresponding power system construction data, respectively calculate the corresponding indicator value of each third-level indicator in each time period; The node value of the third-level indicator in each period is calculated based on the corresponding indicator value of each third-level indicator in each period; The values ​​of the three-level indicator nodes in each time period are substituted into the dynamic Bayesian evaluation network for power supply security to dynamically evaluate the power supply resilience under given extreme weather scenarios or during the actual operation of the power system.

5. The method for dynamic assessment of power supply resilience according to claim 1, characterized in that: The process of simulating the operation of the power system under extreme weather conditions using the constructed power system operation model under extreme weather conditions based on historical extreme weather data includes: Constructing at least one of a power grid line icing model, a photovoltaic panel icing model, a wind turbine blade icing model, a wind turbine output model, a photovoltaic unit output model, a thermal power unit output model, a transmission line fault model, a power load model, and an operation and scheduling model under extreme weather conditions to form a power system operation model; Based on the data of each period in the historical extreme weather data, the power system operation model is used to simulate the operation process simulation data of the power system in each period.

6. The method for dynamic assessment of power supply resilience according to claim 4, characterized in that: The calculation method of the corresponding index value of each of the three-level evaluation indicators in each time period includes at least one of the following: The total installed capacity is calculated based on the rated capacity of each unit; The proportion of coal-fired power generation capacity is calculated based on the installed capacity of coal-fired power generation; The new energy confidence output is calculated based on the new energy maximum output; Calculate the capacity margin based on the maximum load; The grid load rate is calculated based on the grid load; Calculate line redundancy based on the number of backup lines; The energy storage installed capacity is calculated based on the capacity of each energy storage unit; The line failure rate is calculated based on the total length of the faulty line; Obtain the load loss rate based on the power lost during the fault period; The load loss rate is calculated based on the fault duration; Power diversity is calculated based on the installed capacity ratio of each type of power supply; The load recovery rate is calculated based on the restored load and the lost load; The load recovery rate is calculated based on the load recovery time.

7. The method for dynamic assessment of power supply resilience according to claim 5, characterized in that: The operation scheduling model includes a scheduling objective function with the goal of minimizing the load supply and demand deviation and the power generation cost, a power deviation constraint, a communication time constraint and a generator power balance constraint.

8. The method for dynamic assessment of power supply resilience according to claim 7 is characterized in that: The scheduling objective function with the goal of minimizing the load supply and demand deviation and power generation cost is expressed as: Where ΔP(t) is the load supply and demand deviation; C(t) is the power generation cost; α is the cost weight; P G (t), P L (t) are the generator output power and load demand respectively; a, b, and c are the secondary cost coefficient, primary cost coefficient, and constant cost coefficient of the generated power respectively.

9. The method for dynamic assessment of power supply resilience according to claim 8, characterized in that: The generator power balance constraint is expressed as: Among them, P G (t) represents the power generated by the generator; are the minimum and maximum power generation of the generator respectively; u(t-τ) is the control instruction of the power system network layer, which indicates the power adjustment amount. The larger τ is, the more delayed the system control is. When τ=0, the system responds in real time and the power supply and demand are balanced in real time.

10. The method for dynamic assessment of power supply resilience according to claim 5, characterized in that: The transmission line fault model includes: Among them, λ F (y) represents the transmission line fault, y is the ice thickness, d is the design ice thickness of the transmission line, A is the attenuation coefficient, and τ is the damping coefficient.

Images