Power system risk assessment method in extreme weather based on deep feature mining

Abstract

The present invention relates to the technical field of power system risk assessment, and provides a power system risk assessment method in extreme weather based on deep feature mining. The risk assessment method comprises: performing regional division on the whole power system on the basis of the geographic location; on the basis of the historical wind speed data, constructing a correlation model of the strong wind extreme weather in a plurality of regions of the power system; constructing a strong wind scenario sample set of each region on the basis of the correlation model; acquiring the power transmission line fault probability of the corresponding region under each strong wind scenario in the strong wind scenario sample set; randomly allocating a power system operation condition to each strong wind scenario of each region, and acquiring an operation risk value of the power system under the corresponding operation condition; constructing and training a risk assessment model; and using the risk assessment model to perform power system operation risk assessment. The present invention can implement real-time and accurate assessment of the power system operation risk, and can provide an auxiliary decision-making suggestion for the power system.

Description

Power system risk assessment method considering deep feature mining under extreme weather conditions Technical Field

[0001] The present invention belongs to the technical field of power system risk assessment, and specifically relates to a power system operation risk assessment method considering deep feature mining under strong wind extreme weather conditions. Background Art

[0002] As global climate change intensifies, the frequent occurrence of extreme weather events poses significant challenges to the safe and stable operation of power systems. Extreme weather conditions, such as strong winds, can damage power facilities, such as broken power lines and collapsed towers, impacting power supply and system stability. Therefore, accurately assessing the operational risks of power systems under strong wind and extreme weather conditions is crucial.

[0003] Existing technologies have already developed methods for analyzing and assessing the stability and security of power systems using Monte Carlo methods, big data analysis, and machine learning. These methods involve risk assessment models, risk classification, risk warnings, and decision-making recommendations. Furthermore, with the development of intelligent technologies, some existing technologies employ methods such as neural networks and deep learning to intelligently assess power system operational risks, attempting to more accurately and rapidly capture and respond to risks. However, most existing technologies fail to consider the geographical distribution of strong winds and extreme weather events and their impact on the power system, resulting in inaccurate risk assessments. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to provide a power system operation risk assessment method, assessment device, electronic equipment and storage medium that considers deep feature mining under strong wind extreme weather conditions in response to the shortcomings of the existing technology.

[0005] In order to solve the above technical problems, the technical solution adopted by the present invention is:

[0006] A power system operation risk assessment method under extreme weather conditions that considers deep feature mining includes:

[0007] Divide the entire power system into regions based on geographical location;

[0008] Combined with historical wind speed data, a correlation model for strong wind extreme weather in multiple regions of the power system was constructed;

[0009] Constructing a strong wind scene sample set for each region according to the correlation model;

[0010] Obtaining the transmission line failure probability of the corresponding area under each strong wind scenario in the strong wind scenario sample set;

[0011] A power system operating condition is randomly assigned to each strong wind scenario in each region, and the operating risk value of the power system under the corresponding operating condition is obtained;

[0012] Build and train risk assessment models;

[0013] The risk assessment model is used to conduct power system operation risk assessment.

[0014] Furthermore, it also includes: obtaining auxiliary decision-making suggestions based on the assessment results of power system operation risks.

[0015] Further, methods for building and training risk assessment models include:

[0016] Obtain the operating data of the allocated power system under each strong wind scenario in each region, as well as the corresponding operating risk value, and construct a training sample set based on the obtained operating data, operating risk value, and corresponding wind speed value;

[0017] Build risk assessment models;

[0018] The risk assessment model is trained using the training sample set, wherein the operation data and wind speed value are input data, and the operation risk value is output data.

[0019] Furthermore, the method for constructing a risk assessment model includes: using an extreme learning machine (ELM) to construct a risk assessment model, and taking the weighted sum of a Linear kernel function, a Poly polynomial kernel function, an Rbf radial basis function kernel function, and a Wav wavelet kernel function as the final kernel function.

[0020] Furthermore, the final kernel function Expressed as:

[0021] ,

[0022] ,

[0023] ,

[0024] ,

[0025] ,

[0026] in, is a linear kernel function, is a polynomial kernel function, is the radial basis function kernel function, is the wavelet kernel function, is a positive scaling factor, is the bias, is the order of the polynomial, is a wavelet function that satisfies certain conditions, are the weight values ​​of the four kernel functions respectively.

[0027] Furthermore, the method for obtaining the operating risk value of the power system under corresponding operating conditions includes:

[0028] Get voltage limit indicators separately , current overload index , tidal transfer index and load loss rate index ;

[0029] Obtaining power system operation risk value ,in are the importance coefficients of each indicator.

[0030] Furthermore, the method for dividing the entire power system into regions includes: using the fuzzy C-means clustering algorithm to divide the power system into regions, and defining the objective function of the algorithm as ;in Indicates the number of clusters, that is, the number of regions to be divided; represents the fuzzy matrix, which indicates the membership degree of each node to each region; The cluster center matrix contains the center point of each region; is the number of nodes; is a node Belong to the region The degree of membership; is the fuzziness parameter; It is a region The center point of .

[0031] A power system operation risk assessment device considering deep feature mining under extreme weather conditions, comprising an acquisition unit, a training unit, and an assessment unit;

[0032] The acquisition unit is used to obtain the network topology of the power system and divide the entire power system into regions based on geographical locations; to construct a correlation model for strong wind extreme weather in multiple regions of the power system in combination with historical wind speed data; to construct a strong wind scene sample set for each region based on the correlation model; to obtain the transmission line failure probability of the corresponding region under each strong wind scene in the strong wind scene sample set; and to obtain the operating data and operating risk value of the power system under the corresponding operating conditions after randomly assigning a power system operating condition to each strong wind scene in each region.

[0033] The training unit is used to construct a training sample set and a risk assessment model, and is used to train the risk assessment model with the training sample set;

[0034] The evaluation unit is used to perform operation risk evaluation on the power system using the constructed risk evaluation model.

[0035] An electronic device includes a processor and a memory; the memory is used to store executable instructions, and the processor is used to execute the instructions to implement the operation risk assessment method.

[0036] A computer-readable storage medium stores instructions, which, when executed, implement the operation risk assessment method.

[0037] This invention uses an extreme learning machine (ELM) to construct a risk assessment model. The performance of the ELM depends primarily on the choice of kernel function. Common kernel functions include the linear kernel (Linear), the polynomial kernel (Poly), the radial basis function (RBF), and the wavelet kernel (Wav). The linear kernel is one of the simplest kernel functions. It maps input samples based on linear relationships and performs classification by calculating linear combinations of input samples and weights. The linear kernel is generally suitable for linearly separable problems, but its expressive power for complex nonlinear problems is relatively weak, and therefore may not achieve good performance in some cases. The poly kernel is a polynomial kernel function that handles nonlinear problems by mapping input samples to a high-dimensional space. The poly kernel uses a polynomial to fit the data, which can capture more complex relationships between features. The poly kernel performs well for nonlinear problems, but may be prone to overfitting when handling overly complex problems. The RBF kernel has strong expressive power for nonlinear problems and can handle more complex data distributions. The Wav kernel is a kernel function based on the wavelet transform. Wavelet transform is a mathematical tool for signal processing that can extract the time-frequency characteristics of signals. The mathematical representation of the Wav kernel function is usually complex and depends on the selected wavelet basis. The Wav kernel function can handle nonlinear and non-stationary signals well.

[0038] Based on the characteristics of various kernel functions, this paper uses a weighted sum of the Linear, Poly, Rbf, and Wav kernel functions as the kernel function for the ELM. This improved ELM is called the H-ELM. Figures 6-11 illustrate the performance differences between the hybrid kernel function used in this paper and other kernel functions. Data comparisons confirm that the hybrid kernel function used in this paper has significant advantages.

[0039] Figure 6 shows the accuracy of the power system risk assessment results under extreme weather conditions using the Linear kernel function. First, most of the data is concentrated in the range of 0.05 to 0.2, which means that the accuracy of most samples is at a relatively medium level. However, extreme values ​​can also be observed in the data, such as values ​​higher than 0.5 or close to 0, which means that certain specific samples have very high or very low accuracy when using the Linear kernel function. In particular, some values ​​are very close to 0, such as 8.709×10 -6 The Linear kernel provides moderate to good accuracy for most samples, but some values ​​exceed 0.5, such as 0.5169 and 0.7032, indicating that the method may not be ideal in some cases. Overall, the Linear kernel provides moderate to good accuracy for most samples, but there may be discrepancies on specific samples.

[0040] Figure 7 shows the accuracy of the power system risk assessment results under extreme weather conditions when the Poly kernel function is used. From the data, it can be seen that the range of accuracy is quite wide, with the minimum value close to zero, specifically 2.6138×10 -6 , indicating that some models' predictions are quite accurate. Meanwhile, the maximum value reaches 0.07782, indicating that the model's prediction accuracy needs to be improved in certain scenarios. The data is mostly clustered below 0.01, indicating that the model can provide relatively accurate predictions most of the time. However, some values ​​exceed 0.05, which may indicate that the model's predictions are insufficient under certain specific conditions or scenarios. Therefore, in most cases, the Poly kernel function can provide satisfactory prediction accuracy for power system risk assessment, but there are also cases where the prediction deviation is large in certain specific situations.

[0041] Figure 8 shows the accuracy of the power system risk assessment results under extreme weather conditions using the Rbf kernel function. The data shows that the accuracy values ​​are primarily concentrated in a lower range, meaning that in most cases, the model using the Rbf kernel function provides fairly high prediction accuracy. For example, some values ​​are even below 0.001, representing very high accuracy. However, a few values ​​exceed 0.01 and even approach 0.03. These higher values ​​may indicate that the model may have some limitations or deficiencies in certain situations or conditions. Therefore, the Rbf kernel function exhibits relatively high accuracy, but also has cases of high error.

[0042] Figure 9 shows the accuracy of the power system risk assessment results under extreme weather conditions using the Wav kernel function. The vast majority of these data are less than 0.01, which means that the model's predictions are relatively accurate in most cases. There are also a few extremely small values, such as 2.999×10 -6 These values ​​indicate that the model achieves very high accuracy in some cases. However, there are also some large values, such as 0.01047, indicating that the model's predictions are suboptimal in some cases. Overall, the data distribution shows a variety of predictions, with both very good results and relatively poor results. The majority of the accuracy values ​​for the Wav kernel function are concentrated in the lower range, so the Wav kernel function can provide relatively good classification accuracy for power system safety accident levels in most cases.

[0043] Figure 10 shows the accuracy of the power system risk assessment results under extreme weather conditions using the hybrid kernel function adopted by the present invention. First, most of the data points are below 0.005, and even many are below 0.001, which means that in most cases, the model error is very small. Among them, the lowest error is only a value close to 0, such as 9.463×10 -6 , demonstrating that the hybrid kernel function achieves extremely high accuracy in certain scenarios. This extremely low error value occurs more than once in the data, further demonstrating the stability and accuracy of the method. Furthermore, while there are some relatively high error values, they account for a small proportion of the overall data, and the highest error is only around 0.005, which is not significantly different from the lowest error.

[0044] In summary, when the hybrid kernel function is adopted in the present invention, the accuracy of the power system operation risk assessment results under extreme weather conditions is quite outstanding, and it has the characteristics of high efficiency and stability, and has strong practical value.

[0045] The present invention not only considers the impact of geographical location on the operational risk of the power system under extreme weather conditions, but also combines deep feature mining with H-ELM intelligent technology. By using deep feature mining and combining it with historical wind speed data, the present invention can establish a more accurate multi-region strong wind extreme weather correlation model, thereby more accurately predicting the response of the power system under specific wind speeds. The present invention uses H-ELM for training and evaluation, which can achieve rapid, accurate and real-time assessment of the operational risk of the power system.

[0046] In extreme weather conditions, power systems often face enormous risks and challenges. This invention not only provides accurate assessments of these risks but also assists in decision-making, offering clear recommendations on how to effectively address and manage these risks. This helps ensure timely responses to risks and effective measures, ensuring the stable operation of the power system and enhancing its overall resilience. By using this invention to conduct real-time risk assessments and early warnings of power system operations, the power system can avoid large-scale power outages or damage caused by sudden extreme weather events, thereby saving significant repair and restoration costs.

[0047] Therefore, the technical solution provided by the present invention has obvious improvements in prediction accuracy, real-time response, power system resilience and economic benefits compared with the existing technology.

[0048] BRIEF DESCRIPTION OF THE DRAWINGS

[0049] The present invention will be described in further detail below with reference to the accompanying drawings.

[0050] Figure 1 is a flow chart of Example 1 of the present invention;

[0051] FIG2 is a schematic structural diagram of the H-ELM in Example 1 of the present invention;

[0052] FIG3 is a schematic diagram of node voltage amplitudes in a typical operation scenario in Example 1 of the present invention;

[0053] FIG4 is a schematic diagram of branch active power flow in a typical operation scenario in Example 1 of the present invention;

[0054] FIG5 is a schematic diagram of the active power output of the generator in a typical operation scenario in Example 1 of the present invention;

[0055] FIG6 is a schematic diagram showing the accuracy of risk assessment results when the Linear kernel function is used in Example 1 of the present invention;

[0056] FIG7 is a schematic diagram showing the accuracy of risk assessment results when the Poly kernel function is used in Example 1 of the present invention;

[0057] FIG8 is a schematic diagram showing the accuracy of risk assessment results when the Rbf kernel function is used in Example 1 of the present invention;

[0058] FIG9 is a schematic diagram showing the accuracy of risk assessment results when the Wav kernel function is used in Example 1 of the present invention;

[0059] FIG10 is a schematic diagram showing the accuracy of risk assessment results when a hybrid kernel function is used in Example 1 of the present invention;

[0060] Figure 11 is a schematic structural diagram of embodiment 2 of the present invention;

[0061] Figure 12: Schematic diagram of the structure of embodiment 3 of the present invention.

[0062] DETAILED DESCRIPTION

[0063] In order to better understand the present invention, the content of the present invention is further clearly described below in conjunction with the examples and drawings, but the protection content of the present invention is not limited to the following examples. In the following description, a large number of specific details are given in order to provide a more thorough understanding of the present invention. However, it is obvious to those skilled in the art that the present invention can be implemented without one or more of these details.

[0064] Example 1: Referring to Figures 1 to 10, the purpose of this example is to provide a power system operation risk assessment method that considers deep feature mining under extreme weather conditions. As shown in Figure 1, the operation risk assessment method includes:

[0065] 1. Step S1: Construct a correlation model for strong wind extreme weather in multiple regions of the power system.

[0066] Specifically, step S1 includes:

[0067] S11. Divide the entire power system into regions based on geographical location.

[0068] This step first constructs the network topology of the entire power system, which is expressed as ,in Represents a collection of nodes, Represents the line set. Then calculate the power load concentration of each node , the relevant formula is as follows:

[0069] ,

[0070] in Representation and Node A set of connected nodes, Representation node The electricity load, Representation node of electrical load.

[0071] Then the Fuzzy C-Means (FCM) clustering algorithm is used to divide the power system into regions. The objective function of the FCM algorithm is defined as .in, Indicates the number of clusters, that is, the number of regions to be divided; represents the fuzzy matrix, which indicates the membership degree of each node to each region; The cluster center matrix contains the center point of each region; is the number of nodes; is a node Belong to the region The degree of membership; is the fuzziness parameter; It is a region The center point of .

[0072] The FCM algorithm is a conventional algorithm and will not be described here. By running the FCM algorithm, the membership of each node to each area is obtained, and then the node can be assigned to the closest area according to the membership. In this way, the power system is successfully divided into There are several areas, each area contains several nodes and several lines.

[0073] S12. Combine historical wind speed data to construct a correlation model for strong wind extreme weather in multiple regions of the power system.

[0074] The joint distribution of multidimensional random variables is decomposed into a two-dimensional copula function including original variables and conditional variables, and the copula function is used to establish a joint distribution model of strong wind extreme weather in multiple regions. Considering the strong wind extreme weather random variables in multiple regions, the probability density function It can be broken down into:

[0075] ,

[0076] in express The probability density function of ; represents the conditional probability density function, .

[0077] Use C-Vine to decompose the above formula. The decomposition expression is as follows:

[0078] ,

[0079] in, express The marginal cumulative distribution function of ,and represents the conditional cumulative distribution function, .

[0080] 2. Step S2: construct a strong wind scene sample set for each region based on the correlation model.

[0081] Based on the established correlation model of strong wind extreme weather in multiple regions of the power system and combined with the Monte Carlo sampling method, the strong wind extreme weather scenes in each region are extracted. samples, each sample corresponds to a different wind speed. ,generate A uniformly distributed random sample ,in , these samples represent the marginal distribution of extreme wind speeds in their respective regions.

[0082] Using the established vine copula model, each sample in each area All are converted into strong wind scene samples in the area ,The strong wind scene samples in this area together constitute the strong wind scene sample set for this area.

[0083] Specifically, for each , using the inverse transform of Vine Copula, it is converted into a multidimensional random variable The inverse Vine Copula transform can be expressed as: , is the inverse transformation function of Vine Copula; is the sample number of the corresponding area, Then, according to the marginal distribution of each region, Converted to a sample of a strong wind extreme weather scene , which can be expressed as: ,in, It is a region Middle The inverse transformation function of the marginal distribution of a random variable.

[0084] 3. Step S3: Obtain the transmission line failure probability of the corresponding area under each strong wind scenario in the strong wind scenario sample set.

[0085] Strong winds cause horizontal wind loads on overhead lines exceeding their corresponding loads, and vertical wind loads on poles and their connecting conductors exceeding the maximum bearing capacity of the poles. These are the main causes of power line disconnections and tower collapses. Therefore, transmission line failures caused by strong winds mainly include line disconnections and tower collapses.

[0086] The ultimate wind load that the transmission line can withstand is set as , then the critical wind speed for power outage is The ultimate wind load that the tower can withstand is set to , then the critical wind speed of the tower collapse is .in, is the wind pressure unevenness coefficient; is the wind pressure height coefficient; is the wind load adjustment factor; is the body coefficient of the conductor; is the diameter of the power line; is the length of the power line; is the angle between the wind direction and the direction of the transmission line; is the air density; is the drag coefficient of wind load on tower components; It is the effective area of ​​the tower that withstands wind pressure.

[0087] For transmission lines, the predicted wind speed can be considered Less than the critical wind speed of the line When the power outage probability is 0, the predicted wind speed is Greater than When the wind speed and direction are combined, the probability density function based on the wind speed and direction can be used to express the probability of power outage. The probability calculation model of power outage under extreme wind speed is:

[0088] ,

[0089] Similarly, the probability calculation model for power tower collapse under extreme wind speed is:

[0090] ,

[0091] in, for Frequency of wind direction; for Probability density function of wind speed in the direction.

[0092] 4. Step S4: randomly assign a power system operating condition to each strong wind scenario in each area, and obtain the operating risk value of the power system under the corresponding operating condition.

[0093] Specifically, step S4 includes:

[0094] S41. Randomly assign load conditions of the power system to each strong wind scenario in each region.

[0095] Use a random number generator to generate random load values, ensuring that they are within a reasonable range for system load. Specifically, use To represent the load distribution process, where and are the minimum and maximum values ​​of the system load, respectively; Represents a random number uniformly distributed between 0 and 1; is a size of The matrix represents the load distribution of each node in each scenario. The elements in the matrix Indicates the In this scenario The load of each node; is the total number of strong wind extreme weather scenarios; is the number of nodes in the power system.

[0096] S42. Use optimal power flow to calculate the generator output under a given strong wind scenario.

[0097] The goal of the optimal power flow calculation is to minimize power loss and meet the power flow balance and various constraints of the power system, namely ,in, is a node The generator output, is a node and The constraints include load-power relationship, power flow balance, generator output range and node voltage amplitude range.

[0098] The load-power relationship constraint is .in, is a node The generator output, is a node The load power, is a node of distributed power generation.

[0099] The power flow balance constraint condition is .in, is a node and The admittance between is a node The voltage amplitude, is a node The voltage phase angle.

[0100] The generator output range constraint condition is .in, is a node Generator output. and are the minimum and maximum output of the generator respectively.

[0101] The node voltage amplitude range constraint condition is: .in, is a node The voltage amplitude, and are the minimum and maximum node voltage amplitudes, respectively.

[0102] S43. Obtain the power system operation risk value corresponding to the strong wind scenario.

[0103] The indicators related to the power system operation risk value include voltage over-limit index, power flow overload index, power flow transfer index and load loss rate index.

[0104] The voltage over-limit indicator is defined as .in, is the number of buses; Indicates busbar The voltage exceeds the limit, that is:

[0105] ,

[0106] in, is the bus voltage; and They are the upper and lower limits of the bus voltage respectively.

[0107] The power flow overload index is defined as .in, Indicates the set of lines with power flow overload; Indicates the current transmission power of the line; Indicates the rated transmission power of the line.

[0108] The power transfer index reflects the total power transmission distance of the system before and after the fault. First, define the total power transmission distance ,in, A collection of transmission lines. For transmission lines The length of the line represents the power The transmission distance on the line is expressed by the line electrical distance (line impedance value), that is, ,in, is the line resistance; is the line reactance. For the line The amount of active power transmitted, i.e. ,in, Node voltage; Line The conductance and susceptance of For nodes 、 The voltage phase difference.

[0109] In summary, the power transfer degree is defined as .in, and are the total power transmission distances of the system before and after the fault, respectively. The larger the change, the more there is no alternative active power transmission channel near the fault line. The active power has to migrate over a large range in the system, that is, the more vulnerable the system is.

[0110] The load loss rate index is defined as .in, Indicates the set of nodes that need to be load cut; For the Importance coefficient of each load cutting node; For the The load cutting amount of each load cutting node.

[0111] Finally, the above four indicators are combined to obtain the power system operation risk value: ,in, is the importance coefficient of the corresponding indicator.

[0112] Preferably, .

[0113] 5. Step S5: Build and train a risk assessment model.

[0114] Specifically, step S5 includes:

[0115] S51. Construct a training sample set.

[0116] Step S4 is looped multiple times, obtaining the operational data and corresponding operational risk values ​​for the power system assigned to each high wind scenario in each region during each execution. A training sample set is constructed using the acquired operational data, operational risk values, and corresponding wind speed values ​​until the number of samples in the training sample set meets the training requirements. The operational data for the power system includes the active and reactive loads, voltage, and phase of the busbars; the active and reactive output and operating status of the generators; the active and reactive power at both ends of the line; the fault probability and fault location. The fault probability and fault location are determined by the transmission line fault probability acquired in step S3.

[0117] S52. Build a risk assessment model.

[0118] This step uses the Extreme Learning Machine (ELM) to build a risk assessment model, and uses the weighted sum of the linear kernel function (Linear), the polynomial kernel function (Poly), the radial basis function kernel function (Rbf), and the wavelet kernel function (Wav) as the ELM kernel function. In this application document, this improved ELM is called H-ELM. H-ELM kernel function Expressed as:

[0119] ,

[0120] ,

[0121] ,

[0122] ,

[0123] ,

[0124] in, is a linear kernel function, is a polynomial kernel function, is the radial basis function kernel function, is the wavelet kernel function, is a positive scaling factor, is the bias, is the order of the polynomial, is a wavelet function that satisfies certain conditions, are the weight values ​​of the four kernel functions respectively.

[0125] S53. Train the risk assessment model using the training sample set.

[0126] The wind speed values ​​and power system operation data in the training sample set are used as the input data for H-ELM training, and the corresponding power system operation risk values ​​are used as the output data for H-ELM training.

[0127] Randomly initialize the hidden layer weights and biases, randomly initialize the weights for each neuron in the H-ELM hidden layer, and initialize the bias of each hidden neuron. Convert the input data to the hidden layer output through the weighted kernel function: . is the weight of each kernel function, is the input data.

[0128] Using the Moore-Penrose pseudo-inverse method, the weights from the hidden layer to the output layer are solved, which can be expressed as ,in, is the pseudo-inverse of the hidden layer output, is the target output.

[0129] 6. Step S6: Use the constructed risk assessment model to perform power system operation risk assessment.

[0130] collection The wind speed value of strong wind extreme weather in each region, as well as the active and reactive load, voltage and phase of the power system bus, the active and reactive output and operating status of the generator, the active and reactive power at both ends of the line, the fault probability and the fault location are input into the H-ELM risk assessment model. The output result of H-ELM is the operational risk value of the power system in the corresponding region. Then, according to the corresponding relationship between risk value and risk level, the corresponding regional power system operation risk level is determined. The corresponding relationship between risk value and risk level is as follows:

[0131]

[0132] 7. Step S7: Obtain auxiliary decision-making suggestions based on the evaluation results of the power system operation risk.

[0133] Risk value obtained based on the power system operation risk assessment results , conduct risk warning and regulation, and obtain auxiliary decision-making suggestions. Specifically:

[0134] like , the power system is in a safe or low-risk operating state, and the auxiliary decision recommendation is to enter the power generation rescheduling process. The generator rescheduling strategy is the first response to power system risks. By adjusting the generator operation, it can output more power to meet the load demand. While ensuring that the voltage tolerance and the transmission power of each branch are met, the optimal output power of the generator and the power flow distribution of the power system are calculated to achieve reliable power supply at the lowest cost. Objective function , the constraints are:

[0135] ,

[0136] in, Represents the active power output by the power generation redispatch, represents reactive power, Represents the number of generators working simultaneously in the system, is the overload indicator, is a risk indicator, represents the cost weight of power re-regulation, Represents the weights of overload index and risk index, Represents the active power generated by the load, is the reactive power generated by the load, represents the active power from i to j, represents the corresponding reactive power, and Represent the active and reactive power output of the power generation, Represents the node voltage at point i, where max indicates the maximum value and min indicates the minimum value.

[0137] like When the power system is in a medium-risk or higher-risk operating state, the auxiliary decision recommendation is to enter the load reduction process. When the output power of the generator set cannot meet the risk control requirements no matter how it is dispatched, it is necessary to reduce the power system load power based on the rebalancing of the power generation to achieve a balance between the supply and demand of the power system under the risk state. The objective function, and the constraints are:

[0138] ,

[0139] in, represents the reduction of active power, represents the reduction of reactive power, represents the number of load-loss nodes, is the overload indicator, is a risk indicator, and Represent the weights of the corresponding risk indicators, Represents the active power generated by the load, is the reactive power generated by the load, represents the active power from i to j, represents the corresponding reactive power, and Represent the active and reactive power output of the power generation, Represents the node voltage at point i, where the footnote "max" indicates its maximum value and "min" indicates its minimum value. When a risk occurs, power system dispatchers can dynamically adjust the weights of various risk indicators based on actual conditions, thereby achieving risk control under different risk conditions.

[0140] The execution of this embodiment is verified below. First, the Matpower toolkit is used to construct a power system model to perform optimal power flow calculations to generate data samples. The IEEE-39 system is used for simulation experiments. The constructed power system contains 24 nodes, 38 lines, and 33 generators. In order to obtain more negative samples, the rated power of all lines is changed to 81% of the original to weaken the system grid structure. The active output of the generator is changed to 95% of the original, and the maximum active power of the generator is changed to 90% of the original, which makes it easier to cause over-limit, overload, and other situations. The Monte Carlo method is used to obtain 10,000 normal operation scenario samples.

[0141] The entire power system was then divided into four regions, A, B, C, and D, and wind speed data was obtained. This wind speed data was derived from actual power system measurements and applied to regions A, B, C, and D. Assuming a base wind speed of 20 m / s, the wind speeds from regions A to D were normalized. The normalized wind speed range of 0-1 was divided into 20 intervals, each with a length of 0.05. The frequency and probability density of the wind speeds from region A to D within each interval were calculated, as shown in Figure 2. Based on this, a nonparametric kernel density estimation method was used to obtain the probability density function of the wind speed from region A to D.

[0142] The probability density function of the joint distribution was decomposed using the C-vine method, with region A as the root node. Maximum likelihood estimation, based on nonparametric kernel density estimation, was used to obtain the probability density function of the correlation coefficient between the two regions. The correlation coefficients for different copula types are shown in the table below.

[0143]

[0144] As shown in the table below, after comparing the Euclidean distances between the empirical copula function and the evaluated copula functions (including normal copula, t copula, gumbel copula, clayton copula, and frank copula), it can be seen that in the two different regions, the Euclidean distance between the empirical copula function and the gumbel copula is the smallest, so the gumbel copula is selected as the optimal copula function.

[0145]

[0146] Using the C-vine copula model constructed above, combined with the Monte Carlo sampling method, 100 high wind scenarios were sampled for each region. The probability of transmission line failures (line breaks and tower collapses) under different high wind scenarios was then calculated for each region. The high wind scenarios were randomly combined with normal power system operation scenarios to generate 50,000 normal power system operation scenarios under high wind conditions. Figures 4-6 show the node voltage amplitudes, branch active power flows, and generator active power output for four typical operation scenarios. Using the Monte Carlo sampling method, fault random sampling was performed on these 50,000 random combinations, with nine fault types sampled for each random combination, totaling 500,000 data points. Each data sample included wind speed, busbar active and reactive load, voltage, and phase, generator active and reactive output, and operating status, and active and reactive power and operating status at both ends of the line. Using these 500,000 data points, the power system operation risk value was calculated and used as the output data of the H-ELM.

[0147] In addition, the present invention collected a total of 500,000 data points for model training and tested the model accuracy when using different kernel functions to train the ELM model using 520 data points. Figure 6 is a schematic diagram of the accuracy of the risk assessment results when using the Linear kernel function, Figure 7 is a schematic diagram of the accuracy of the risk assessment results when using the Poly kernel function, Figure 8 is a schematic diagram of the accuracy of the risk assessment results when using the Rbf kernel function, Figure 9 is a schematic diagram of the accuracy of the risk assessment results when using the Wav kernel function, and Figure 10 is a schematic diagram of the accuracy of the risk assessment results when using the hybrid kernel function. From the test results, it can be seen that when the present invention uses the hybrid kernel function as the kernel function of the ELM, the accuracy of the power system safety accident level classification is quite excellent.

[0148] Example 2: Referring to FIG11 , the purpose of this example is to provide a power system operation risk assessment device that considers deep feature mining under extreme weather conditions, and is used to perform the operation risk assessment method of Example 1. The operation risk assessment device 200 includes: an acquisition unit 201 , a training unit 202 , and an assessment unit 203 .

[0149] Acquisition unit 201 is used to obtain the network topology of the power system and divide the entire power system into regions based on geographical location; to construct a correlation model for strong wind extreme weather in multiple regions of the power system based on historical wind speed data; to construct a strong wind scenario sample set for each region based on the correlation model; to obtain the transmission line failure probability of the corresponding region under each strong wind scenario in the strong wind scenario sample set; and to randomly assign a power system operating condition to each strong wind scenario in each region, and then obtain the operating data and operating risk value of the power system under the corresponding operating condition. Specifically, in conjunction with Example 1, acquisition unit 201 is used to execute steps S1-S4.

[0150] The training unit 202 is used to construct a training sample set and a risk assessment model, and to train the risk assessment model with the training sample set. Specifically, in conjunction with Example 1, the training unit 202 is used to execute step S5.

[0151] The evaluation unit 203 is used to use the constructed risk assessment model to perform power system operation risk assessment and obtain auxiliary decision suggestions based on the operation risk assessment results. Specifically, in combination with Example 1, the training unit 202 is used to perform steps S6-S7.

[0152] Regarding the device in this embodiment, the specific manner in which each unit performs operations has been described in detail in Example 1 and will not be elaborated on here.

[0153] Those skilled in the art should understand that for the convenience and simplicity of description, this application only uses the division of the above-mentioned functional modules or units as examples in the device embodiments. In actual applications, the above-mentioned functions can be assigned to different functional modules or units as needed, that is, the internal structure of the device can be divided into different functional modules or units to complete all or part of the functions described above.

[0154] Example 3: Referring to Figure 12, the purpose of this embodiment is to provide an electronic device 300, including at least one processor 301 and one or more memories 302 for storing instructions executable by the processor 301. The processor 301 is configured to execute the instructions in the memory 302 to implement the operational risk assessment method described in Example 1.

[0155] The electronic device 300 further includes a bus 304 , and the processor 301 and the memory 302 are connected to each other via the bus 304 , or are connected to each other in other ways.

[0156] The processor 301 is a central processing unit (CPU), a general-purpose processor, a network processor (NP), a digital signal processor (DSP), a microprocessor, a microcontroller, a programmable logic device (PLD), or any combination thereof. The processor 301 may also be other devices with processing capabilities, such as circuits, devices, or software modules, without limitation. In one example, the processor 301 may include one or more CPUs, such as CPU0 and CPU1 in FIG12 .

[0157] The memory 302 may be a read-only memory (ROM) or other type of static storage device capable of storing static information and / or instructions, a random access memory (RAM) or other type of dynamic storage device capable of storing information and / or instructions, an electrically erasable programmable read-only memory (EEPROM), a compact disc read-only memory (CD-ROM) or other optical disc storage, optical disc storage (including compact discs, laser discs, optical discs, digital versatile discs, Blu-ray discs, etc.), a magnetic disk storage medium, or other magnetic storage device, etc., without limitation.

[0158] It should be noted that the memory 302 can exist independently of the processor 301 or can be integrated with the processor 301. The memory 302 can be used to store instructions, program codes, or some data. The memory 302 can be located inside the electronic device 300 or outside the electronic device 300, without limitation.

[0159] As an optional implementation, electronic device 300 also includes a communication interface 303. Communication interface 303 is a wired interface (or port), such as a fiber distributed data interface (FDDI) or a gigabit Ethernet (GE) interface. Alternatively, communication interface 303 is a wireless interface. Communication interface 303 can be a module, circuit, communication interface, or any device capable of communication. Communication interface 303 is used to communicate with other devices or other communication networks, such as Ethernet, a radio access network (RAN), or a wireless local area network (WLAN).

[0160] As an optional implementation, the electronic device 300 further includes an input device 305 and an output device 306. For example, the input device 305 is a keyboard, a mouse, a microphone, a joystick, and the output device 306 is a display screen, a speaker, and the like.

[0161] It should be noted that the electronic device 300 may be a desktop computer, a portable computer, a network server, a mobile phone, a tablet computer, a wireless terminal, an embedded device, a chip system, or a device having a structure similar to that shown in FIG12 . Furthermore, the component structure shown in FIG12 does not limit the terminal device. In addition to the components shown in FIG12 , the electronic device 300 may include more or fewer components than shown, or combine certain components, or arrange the components differently.

[0162] Embodiment 4: The purpose of this embodiment is to provide a computer-readable storage medium.

[0163] All or part of the processes in the above method embodiments can be completed by computer instructions to instruct relevant hardware. The program can be stored in the computer-readable storage medium. When the program is executed, it can implement the operation risk assessment method described in Example 1.

[0164] The computer-readable storage medium may be an internal storage unit of the electronic device of Example 3, such as a hard disk or memory. The computer-readable storage medium may also be an external storage device of the electronic device, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. Furthermore, the computer-readable storage medium may include both an internal storage unit and an external storage device of the electronic device described above. The computer-readable storage medium is used to store computer programs and other programs and data required by the electronic device. The computer-readable storage medium may also be used to temporarily store data that has been output or is about to be output.

[0165] In conjunction with the several embodiments provided in this application, it should be understood that the provided devices and methods can be implemented in other ways. The device embodiments described above are merely illustrative. For example, the division of the modules or units is merely a logical function division. In actual implementation, other division methods may be used. For example, multiple modules, units, or components may be combined or integrated into another device, or some features may be ignored or not performed.

[0166] In addition, in the embodiments of the device of the present application, each functional module or unit may be integrated into one unit, each module or unit may exist physically separately, or two or more modules or units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional modules.

[0167] If the above-mentioned integrated unit is implemented in the form of a software functional module and sold or used as an independent product, it can be stored in a readable storage medium. Based on this understanding, the technical solution of the embodiment of the present application, or the part that contributes to the existing technology, or all or part of the technical solution can be embodied in the form of a software product. The software product is stored in a storage medium and includes a number of instructions for enabling a device (which can be a single-chip microcomputer, chip, etc.) or a processor to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: various media that can store program code, such as a USB flash drive, a mobile hard disk, ROM, RAM, a magnetic disk, or an optical disk.

[0168] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention and are not limiting. Other modifications or equivalent substitutions made to the technical solution of the present invention by ordinary technicians in this field should be included in the scope of the claims of the present invention as long as they do not depart from the spirit and scope of the technical solution of the present invention.

Claims

1. A power system operation risk assessment method considering deep feature mining under extreme weather conditions, characterized by: The operational risk assessment method comprises: Divide the entire power system into regions based on geographical location; Combined with historical wind speed data, a correlation model of strong wind extreme weather in multiple regions of the power system is constructed; According to the correlation model, construct a strong wind scene sample set for each area; Obtaining the transmission line failure probability of the corresponding area in each strong wind scenario in the strong wind scenario sample set; A power system operating condition is randomly assigned to each strong wind scenario in each region, and the operating risk value of the power system under the corresponding operating condition is obtained; Build and train risk assessment models; The risk assessment model is used to conduct power system operation risk assessment.

2. The power system operation risk assessment method considering deep feature mining under extreme weather conditions according to claim 1 is characterized by: Also includes: Obtain auxiliary decision-making suggestions based on the assessment results of power system operation risks.

3. The power system operation risk assessment method considering deep feature mining under extreme weather conditions according to claim 1 is characterized by: Methods for building and training risk assessment models include: Obtain the operating data of the power system allocated to each region under each strong wind scenario, as well as the corresponding operating risk value, and construct a training sample set with the obtained operating data, operating risk value and corresponding wind speed value; Construct risk assessment models; The risk assessment model is trained with the training sample set, wherein the operation data and wind speed value are input data, and the operation risk value is output data.

4. The power system operation risk assessment method considering deep feature mining under extreme weather conditions according to claim 3 is characterized by: The method for constructing a risk assessment model includes: using an extreme learning machine (ELM) to construct a risk assessment model, and taking the weighted sum of a Linear kernel function, a Poly kernel function, an Rbf radial basis function kernel function, and a Wav wavelet kernel function as the final kernel function.

5. The power system operation risk assessment method considering deep feature mining under extreme weather conditions according to claim 4 is characterized by: Final kernel function It is expressed as: ， ， ， ， ， in, is a linear kernel function, is a polynomial kernel function, is the radial basis function kernel function, is the wavelet kernel function, is a positive scaling factor, is the bias, is the order of the polynomial, is a wavelet function that satisfies certain conditions, are the weight values ​​of the four kernel functions respectively.

6. The method for assessing the operation risk of a power system under extreme weather conditions according to claim 1 is characterized in that: Methods for obtaining the operating risk value of the power system under corresponding operating conditions include: Get voltage limit indicators separately , Power flow overload index , Flow Transfer Index and load loss rate index ； Obtaining the power system operation risk value ,in are the importance coefficients of the corresponding indicators respectively.

7. The power system operation risk assessment considering deep feature mining under extreme weather conditions according to claim 1 is characterized by: The method of dividing the entire power system into regions includes: using the fuzzy C-means clustering algorithm to divide the power system into regions, and defining the objective function of the algorithm as ;in Indicates the number of clusters, that is, the number of regions to be divided; represents the fuzzy matrix, which indicates the membership degree of each node to each region; The cluster center matrix contains the center point of each region; is the number of nodes; Is a node Belong to the area The degree of membership; is the fuzziness parameter; It is a region The center point of .

8. A power system operation risk assessment device considering deep feature mining under extreme weather conditions, characterized by: It includes acquisition unit, training unit and evaluation unit; The acquisition unit is used to acquire the network topology of the power system and divide the entire power system into regions according to the geographical location; to construct a correlation model of strong wind extreme weather in multiple regions of the power system in combination with the historical wind speed data; and to construct a sample set of strong wind scenes in each region according to the correlation model; Used to obtain the transmission line failure probability of the corresponding area in each strong wind scenario in the strong wind scenario sample set; Used to obtain the operation data and operation risk value of the power system under the corresponding operation condition after randomly assigning one power system operation condition to each strong wind scenario in each area; The training unit is used to construct a training sample set and a risk assessment model, and is used to train the risk assessment model with the training sample set; The evaluation unit is used to evaluate the operation risk of the power system using the constructed risk evaluation model.

9. An electronic device, characterized in that: It comprises a processor and a memory; the memory is used to store executable instructions, and the processor is used to execute the instructions to implement the operation risk assessment method described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that: The readable storage medium stores instructions, and when the instructions are executed, the operation risk assessment method described in any one of claims 1-7 is implemented.

Images