Method for generating typical power system fault scenarios for extreme events

Abstract

This invention discloses a method for generating a set of typical fault scenarios for power systems oriented towards extreme events. The method includes: obtaining an original fault scenario matrix; establishing a DC power flow model based on the network parameters of the power system to be evaluated and calculating the line power transfer distribution factor; calculating the bus injection weights based on the load and generation scale of each bus; calculating and normalizing the line load-generation weighted sensitivity weights; constructing a single-line difference function that simultaneously characterizes the difference between whether a line is faulty and the difference in the time of fault occurrence; weighting the single-line difference function according to the line load-generation weighted sensitivity weights to obtain the scenario distance; and clustering the original fault scenario matrix based on the scenario distance using a specified clustering algorithm to obtain a reduced set of typical fault scenarios and outputting it. This invention aims to achieve the generation of a set of typical fault scenarios that takes into account electrical sensitivity, fault time information, and balanced coverage constraints.

Description

Technical Field

[0001] This invention belongs to the field of power system extreme disaster risk assessment technology, specifically relating to a method for generating a set of typical power system fault scenarios for extreme events. Background Technology

[0002] Under the disaster effects of extreme events such as typhoons, tsunamis, earthquakes and snowstorms, power systems may experience multi-line and multi-time-period associated faults. In order to obtain the system load loss, power supply reliability and vulnerability indicators, it is usually necessary to perform optimal power flow (OPF) calculations on a large number of random fault scenarios. As the number of scenarios reaches tens of thousands or even higher, solving for each scenario will lead to excessive computational burden. Existing scenario reduction methods mostly use clustering or sampling, but they usually have the following shortcomings: (1) Treating each line equally and ignoring the dominant role of "electrically critical lines" in power flow distribution, congestion and load shedding results, resulting in insufficient representativeness of the reduced scenarios for optimal power flow output; (2) The cluster capacity obtained by clustering varies greatly, resulting in unbalanced weights of typical fault scenarios, which in turn leads to deviations in expected load loss or tail risk assessment, and it is difficult to uniformly allocate batch optimal power flow calculation resources; (3) Insufficient use of information on "fault occurrence time", only describing whether there is a fault and ignoring the impact of fault timing on the system operating status. Summary of the Invention

[0003] The technical problem to be solved by this invention is to provide a method for generating a set of typical fault scenarios for power systems in response to the above-mentioned problems in the prior art. This invention aims to generate a set of typical fault scenarios that takes into account electrical sensitivity, fault time information and balance coverage constraints. This ensures that the typical fault scenarios can highlight the differences in faults of critical lines that are sensitive to optimal power flow output, while also ensuring that the number of original scenarios covered by each typical fault scenario is basically consistent, so as to be used for subsequent calculation of optimal power flow load loss and vulnerability assessment.

[0004] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows:

[0005] A method for generating a typical fault scenario set for power systems oriented towards extreme events includes the following steps: S1: Obtaining the original fault scenario matrix The original fault scenario matrix Each row in the matrix corresponds to a fault scenario, and each column corresponds to a line. Matrix elements represent the fault occurrence time or non-fault status of the line in its corresponding fault scenario. S2: Establish a DC power flow model based on the network parameters of the power system to be evaluated and calculate the line power transfer distribution factor; calculate the bus injection weight based on the load and generation scale of each bus; S3: Calculate and normalize the line load-generation weighted sensitivity weight based on the line power transfer distribution factor and the bus injection weight; S4: Construct a single-line difference function that simultaneously represents the difference between whether a line is faulty and the difference in the fault occurrence time; obtain the scenario distance by weighting the single-line difference function according to the line load-generation weighted sensitivity weight; S5: Apply the original fault scenario matrix... Clustering is performed using a specified clustering algorithm based on scene distance to obtain a set of typical fault scenes after scene reduction, and then output.

[0006] Optionally, the calculation of the line power transfer distribution factor in step S2 includes: specifying a balanced bus for the DC power flow model consisting of lines and buses, and the remaining buses being unbalanced buses; applying a unit active power injection to the unbalanced buses and applying a unit active power extraction from the balanced buses, and calculating the line power transfer distribution factor according to the following formula:

[0007] ;

[0008] in, For the first Line 1 to the 1st The power transfer distribution factor of the busbar For the first Changes in active power flow along each line. For the first Changes in active power injection of each busbar.

[0009] Optionally, the functional expression for calculating the bus injection weight based on the load and generation capacity of each bus in step S2 is as follows:

[0010] ;

[0011] in, For the first Busbar weight injection; and The first and Busbar load of the busbar and They are respectively and The absolute value, and The first and The power generation capacity of the busbar and They are respectively and The absolute value, This indicates all buses in the power system to be evaluated. and Summing is performed, and the sum of the bus injection weights of all buses in the power system to be evaluated is 1.

[0012] Optionally, the normalized function expression for calculating the line load-generation weighted sensitivity weight based on the line power transfer distribution factor and the bus injection weight in step S3 is as follows:

[0013] ;

[0014] ;

[0015] in, For the first The load-generation weighted sensitivity of each line. For the first Busbar weight injection; For the first Line 1 to the 1st The absolute value of the line power transfer distribution factor of each busbar. For the first The normalized load-generation weighted sensitivity weights for each line. For the first Load-generation weighted sensitivity of each line.

[0016] Optionally, the function expression of the single-line difference function constructed in step S4, which simultaneously characterizes the difference between whether a line is faulty and the difference between the time of fault occurrence, is as follows:

[0017] ;

[0018] in, For the scene With scene On the line The single-line difference function value, For the scene Downline Fault identifiers for the corresponding time period For the scene Downline The fault flag for the corresponding time period is set to 0, indicating that no fault has occurred. Original fault scenario matrix The number of time periods in the time period.

[0019] Optionally, the expression for calculating the scene distance in step S4 is:

[0020] ;

[0021] in, Let be the scene distance from fault scene s to fault scene i. For the first The normalized load-generation weighted sensitivity weights for each line. For the scene With scene On the line The single-line difference function value.

[0022] Optionally, step S5 includes:

[0023] S5.1: Obtain the number of given typical failure scenarios and capacity Capacity in typical fault scenarios satisfy:

[0024] ;

[0025] in, Original fault scenario matrix The number of fault scenarios in;

[0026] S5.2: Constructing a matrix of original fault scenarios The objective function for clustering based on scene distance;

[0027] S5.3: Under the premise of ensuring that the coverage of each cluster is consistent, the k-medoids representative point selection algorithm is used to select points based on the objective function of the original fault scenario matrix. Clustering is performed to obtain a set of typical fault scenarios after scenario reduction and then output.

[0028] Optionally, in step S5.2, the function expression of the objective function is:

[0029] ;

[0030] in, for The set of representative points of each cluster. For the k-th cluster, Let be any fault scenario s in the k-th cluster and the representative point of the k-th cluster. Scene distance.

[0031] Optionally, step S5.3 includes:

[0032] S5.3.1: Distance-based initialization is used, including: from the original fault scenario matrix A fault scenario is randomly selected as the first representative point. For any candidate scenario Calculate the nearest distance to the selected representative point set using the scene distance function. :

[0033] ;

[0034] in, Candidate scenarios To and representative point The scene distance is determined by the nearest distance to the set of selected representative points. square The next representative point is selected from the candidate scenes with a proportional probability, and this process is repeated until a result is obtained. The initial set of representative points is obtained by selecting 1 set of distinct representative points. ,in, ~ These are the initial representative point sets. K representative points in the middle, Original fault scenario matrix The number of fault scenarios in;

[0035] S5.3.2: In the current fixed The set of representative points of each cluster Under the condition of each candidate scenario As a node with a supply of 1, each representative point Capacity as a typical fault scenario Nodes in the process, to select from candidate scenarios To the representative point Scene distance Assuming the cost of an edge, a minimum cost allocation model with capacity constraints is established. By solving this model, an allocation scheme that satisfies the capacity constraints and minimizes the total cost is obtained. ;

[0036] The decision variables used in the minimum cost allocation model with capacity constraints are:

[0037] ;

[0038] in, Candidate scenarios Downline Fault identification within the corresponding time period; the objective function of the minimum cost allocation model with capacity constraints is expressed as follows:

[0039] ;

[0040] in, For the candidate scene set, From scene s to representative point The scene distance; the functional expression of the constraint condition used in the minimum cost allocation model with capacity constraints is:

[0041] ;

[0042] ;

[0043] ;

[0044] S5.3.3: Update the representative points for each cluster:

[0045] ;

[0046] in," "For update operations, Let be the scene distance from fault scene s to fault scene j;

[0047] S5.3.4: Determine whether the objective function has converged or the representative point no longer changes. If not, jump to step S5.3.2 and continue iterating; otherwise, convert the final set of representative points obtained from the iteration. corresponding A representative fault scenario is output as the typical fault scenario set.

[0048] Optionally, in step S5.3.4, the representative point set obtained from the final iteration is... corresponding When a representative fault scenario is output as the obtained typical fault scenario set, the function expression of the obtained typical fault scenario set is:

[0049] ;

[0050] in, For the obtained typical fault scenario set, This indicates the origin of the original fault scenario matrix. Extracting from the middle The set of representative points of each cluster. Original fault scenario matrix Number of time periods The number of typical failure scenarios. This refers to the number of lines.

[0051] Optionally, in step S5.3.4, the representative point set obtained from the final iteration is... When outputting the corresponding K representative fault scenarios as the typical fault scenario set, the output also includes a binary backoff matrix and an allocation scheme that satisfies the capacity constraint and minimizes the total cost. The functional expression for the binary devolution matrix is:

[0052] ;

[0053] in, For the first The first typical fault scenario The binary decommissioning status of the line constitutes the first line in the binary decommissioning matrix. OK Column elements; Indicates the line In the Under typical fault scenarios, it is in operation. Indicates the line In the Under typical fault scenarios, it is in a state of being returned to service; Original fault scenario matrix The Middle Line number The elements of the column represent the original fault scenario of the corresponding line. The fault occurrence period identifier; a value of 0 indicates no fault, and a value of 1 indicates a fault occurring during a specified period. This indicates that a fault occurred within the corresponding time period.

[0054] Compared with existing technologies, the present invention can achieve the following beneficial effects: (1) The present invention constructs line sensitivity weights by using the line power transfer distribution factor (PTDF) and load-generation scale weights, making typical scenarios closer to power flow redistribution and congestion mechanisms, thereby improving the representativeness of the optimal power flow load loss results; (2) The present invention ensures that the coverage of each typical scenario is basically consistent through capacity constraints, realizing equal weighted statistics and balanced calculation resource allocation, and reducing reduction deviations; (3) The present invention can characterize the impact of fault timing on the operational feasible region under extreme events by introducing differences in fault occurrence time periods; (4) The present invention significantly reduces the number of optimal power flow solutions while maintaining evaluation accuracy, thereby improving the efficiency of vulnerability assessment. Therefore, the present invention can generate a set of typical fault scenarios that takes into account electrical sensitivity, fault time period information and balanced coverage constraints, enabling typical fault scenarios to highlight the differences in critical line faults that are sensitive to optimal power flow output, while ensuring that the number of original scenarios covered by each typical fault scenario is basically consistent, thus being used for subsequent optimal power flow load loss calculation and vulnerability assessment. Attached Figure Description

[0055] Figure 1 This is a schematic diagram of the basic process of the method in an embodiment of the present invention. Detailed Implementation

[0056] To enable those skilled in the art to better understand the technical solutions of the present invention, the technical solutions of the present invention will be further described in detail below with reference to the accompanying drawings in the embodiments of the present invention.

[0057] like Figure 1 As shown, this embodiment provides a method for generating a typical fault scenario set of a power system oriented towards extreme events, including the following steps: S1: Obtain the original fault scenario matrix The original fault scenario matrix Each row in the matrix corresponds to a fault scenario, and each column corresponds to a line. Matrix elements represent the fault occurrence time or non-fault status of the line in its corresponding fault scenario. S2: Establish a DC power flow model based on the network parameters of the power system to be evaluated and calculate the line power transfer distribution factor; calculate the bus injection weight based on the load and generation scale of each bus; S3: Calculate and normalize the line load-generation weighted sensitivity weight based on the line power transfer distribution factor and the bus injection weight; S4: Construct a single-line difference function that simultaneously represents the difference between whether a line is faulty and the difference in the fault occurrence time; obtain the scenario distance by weighting the single-line difference function according to the line load-generation weighted sensitivity weight; S5: Apply the original fault scenario matrix... Clustering is performed using a specified clustering algorithm based on scene distance to obtain a set of typical fault scenes after scene reduction, and then output.

[0058] The original fault scenario matrix obtained in step S1 It can be represented as:

[0059] ;

[0060] in, For the number of scenes, The number of lines; where any element For the scene Downline The fault identifier for the corresponding time period is set as follows: a value of 0 indicates no fault has occurred, and a value of 1 indicates a fault has occurred. Original fault scenario matrix The number of time periods in the time period .

[0061] Step S2, calculating the line power transfer distribution factor, includes: specifying a slack bus for the DC power flow model consisting of lines and buses, with the remaining buses being unbalanced buses; for the DC power flow model consisting of lines and buses, constructing the bus admittance matrix and the branch-bus correlation matrix under the DC power flow approximation based on the power system network parameters (buses, generators, branches). Since this is existing technology, its implementation details will not be elaborated here. A unit active power injection is applied to the unbalanced bus and a unit active power extraction is applied to the slack bus, and the line power transfer distribution factor is calculated according to the following formula:

[0062] ;

[0063] in, For the first Line 1 to the 1st The power transfer distribution factor of the busbar For the first Changes in active power flow along each line. For the first The change in active power injection for each bus. As an optional implementation, a 1MW disturbance is injected into each bus and a 1MW disturbance is extracted from the balancing bus, and then the line power transfer distribution factor is calculated. For branch index, For the bus index.

[0064] The functional expression for calculating the bus injection weight based on the load and power generation capacity of each bus in step S2 is as follows:

[0065] ;

[0066] in, For the first Busbar weight injection; and The first and The bus load of a busbar (multiple units on the same busbar can be summed). and They are respectively and The absolute value, and The first and The power generation capacity of a single busbar (multiple units on the same busbar can be summed). and They are respectively and The absolute value, This indicates all buses in the power system to be evaluated. and Summing, and the sum of the bus injection weights of all buses in the power system to be evaluated is 1, that is: .

[0067] In step S3, the function expression for calculating and normalizing the line load-generation weighted sensitivity weight based on the line power transfer distribution factor and the bus injection weight is as follows:

[0068] ;

[0069] ;

[0070] in, For the first The load-generation weighted sensitivity of each line. For the first Busbar weight injection; For the first Line 1 to the 1st The absolute value of the line power transfer distribution factor of each busbar. For the first The normalized load-generation weighted sensitivity weights for each line. For the first Load-generation weighted sensitivity of each line.

[0071] The functional expression of the single-line difference function constructed in step S4, which simultaneously characterizes the difference between whether a line is faulty and the difference between the time of fault occurrence, is as follows:

[0072] ;

[0073] in, For the scene With scene On the line The single-line difference function value, For the scene Downline Fault identifiers for the corresponding time period For the scene Downline The fault flag for the corresponding time period is set to 0, indicating that no fault has occurred. Original fault scenario matrix The number of time periods in the time period.

[0074] The function expression for calculating the scene distance in step S4 is:

[0075] ;

[0076] in, Let be the scene distance from fault scene s to fault scene i. For the first The normalized load-generation weighted sensitivity weights for each line. For the scene With scene On the line The single-line difference function value.

[0077] Step S5 is used to perform representative point selection (equalized coverage of k-medoids) under capacity constraints. In this embodiment, step S5 includes:

[0078] S5.1: Obtain the number of given typical failure scenarios and capacity Capacity in typical fault scenarios satisfy:

[0079] ;

[0080] in, Original fault scenario matrix The number of fault scenarios in;

[0081] S5.2: Constructing a matrix of original fault scenarios The objective function for clustering based on scene distance;

[0082] S5.3: Under the premise of ensuring that the coverage of each cluster is consistent, the k-medoids representative point selection algorithm is used to select points based on the objective function of the original fault scenario matrix. Clustering is performed to obtain a set of typical fault scenarios after scenario reduction and then output.

[0083] In step S5.2, the function expression of the objective function is:

[0084] ;

[0085] in, for The set of representative points (medoids) of each cluster. For the k-th cluster, Let be any fault scenario s in the k-th cluster and the representative point of the k-th cluster. Scene distance.

[0086] As an optional implementation, step S5.3 includes: (1) initialization: using distance-distributed initialization or multiple random restarts; (2) allocation: in a fixed Below, to minimize With the objective of each representative receiving cap samples, a minimum-cost allocation model with capacity constraints is constructed to solve for the allocation that satisfies the capacity constraints and minimizes the total cost. (3) Update: For each cluster (3) Update the representative point; (4) Iterate until the objective function converges or the representative point no longer changes. Specifically, step S5.3 includes:

[0087] S5.3.1: Distance-based initialization is used, including: from the original fault scenario matrix A fault scenario is randomly selected as the first representative point. For any candidate scenario Calculate the nearest distance to the selected representative point set using the scene distance function. :

[0088] ;

[0089] in, Candidate scenarios To and representative point The scene distance is determined by the nearest distance to the set of selected representative points. square The next representative point is selected from the candidate scenes with a proportional probability, and this process is repeated until a result is obtained. The initial set of representative points is obtained by selecting 1 set of distinct representative points. ,in, ~ These are the initial representative point sets. K representative points in the middle, Original fault scenario matrix The number of fault scenarios in;

[0090] S5.3.2: In the current fixed The set of representative points of each cluster Under the condition of each candidate scenario As a node with a supply of 1, each representative point Capacity as a typical fault scenario Nodes in the process, to select from candidate scenarios To the representative point Scene distance Assuming the cost of an edge, a minimum cost allocation model with capacity constraints is established. By solving this model, an allocation scheme that satisfies the capacity constraints and minimizes the total cost is obtained. ;

[0091] The decision variables used in the minimum cost allocation model with capacity constraints are:

[0092] ;

[0093] in, Candidate scenarios Downline Fault identification within the corresponding time period; the objective function of the minimum cost allocation model with capacity constraints is expressed as follows:

[0094] ;

[0095] in, For the candidate scene set, From scene s to representative point The scene distance; the functional expression of the constraint condition used in the minimum cost allocation model with capacity constraints is:

[0096] ;

[0097] ;

[0098] ;

[0099] The method for solving the minimum cost allocation model with capacity constraints is an existing well-known method that can be implemented using commercial solvers, so the details of its solution will not be described here.

[0100] S5.3.3: Update the representative points for each cluster:

[0101] ;

[0102] in," "For update operations, Let be the scene distance from fault scene s to fault scene j;

[0103] S5.3.4: Determine whether the objective function has converged or the representative point no longer changes. If not, jump to step S5.3.2 and continue iterating; otherwise, convert the final set of representative points obtained from the iteration. The corresponding K representative fault scenarios are output as the typical fault scenario set.

[0104] In step S5.3.4, the final set of representative points obtained from the iteration is... When the corresponding K representative fault scenarios are output as the obtained typical fault scenario set, the function expression of the obtained typical fault scenario set is:

[0105] ;

[0106] in, For the obtained typical fault scenario set, This indicates the origin of the original fault scenario matrix. Extracting from the middle The set of representative points of each cluster. Original fault scenario matrix Number of time periods The number of typical failure scenarios. This refers to the number of lines.

[0107] In step S5.3.4, the final set of representative points obtained from the iteration is... When outputting the corresponding K representative fault scenarios as the typical fault scenario set, the output also includes a binary backoff matrix and an allocation scheme that satisfies the capacity constraint and minimizes the total cost. , This represents the mapping from the original fault scenario to a typical fault scenario, where the functional expression of the binary regression matrix is:

[0108] ;

[0109] in, For the first The first typical fault scenario The binary decommissioning status of the line constitutes the first line in the binary decommissioning matrix. OK Column elements; Indicates the line In the Under typical fault scenarios, it is in operation. Indicates the line In the Under typical fault scenarios, it is in a state of being returned to service; Original fault scenario matrix The Middle Line number The elements of the column represent the lines under the corresponding original fault scenarios. The fault occurrence period identifier; a value of 0 indicates no fault, and a value of 1 indicates a fault occurring during a specified period. This indicates that a fault occurred within the corresponding time period.

[0110] To verify the method for generating typical power system fault scenario sets for extreme events in this embodiment, the IEEE 39-bus system is used as an example: Number of lines Number of scenes Original fault scenario matrix Number of time periods Select the number of given typical failure scenarios. Therefore, the capacity cap = 125 for a given typical fault scenario. Following steps S1–S5, the original fault scenario matrix is ​​finally processed. Clustering is performed using a specified clustering algorithm based on scene distance to obtain a set of typical fault scenes after scene reduction, and then output.

[0111] As a comparison with the method in this embodiment, the comparison method used, except that the line weights are taken as uniform weights... Except for the above, all other conditions are completely consistent with the method in this embodiment: similarly , , , , Similarly, k-medoids with capacity constraints are used to generate typical fault sets and mapping relationships.

[0112] To verify the representativeness of the load shedding statistics for typical scenarios, the DC-OPF load shedding model was used to calculate the minimum load shedding amount for each scenario. The statistics of all original scenes are used as a benchmark. Under the condition of balanced coverage, each typical scene has the same weight. Therefore, the statistical error between the mean and quantiles of the load shedding in the typical set can be defined as:

[0113] ;

[0114] ;

[0115] in, The relative error index is the average value. This is the estimated average value of the system performance index calculated based on a typical fault set. This represents the true average value of the system performance metrics calculated based on the full fault set. The relative error index is the 95th percentile value. This represents the 95th percentile value of the system performance index calculated based on a typical fault set. This represents the 95th percentile value of the system performance index calculated based on the full fault set. By all Calculated for each scenario The results were obtained through equally weighted statistics from a set of typical scenarios (80 representative scenarios). Comparison of statistical errors in optimal power flow load shedding (also...). Each cluster covers 125 (units), to the whole Using the original scene's DC-OPF results as a baseline, we obtained: Baseline Mean: benchmark .

[0116] In this embodiment, the statistical error between the mean and quantile of the load shedding in the typical set is:

[0117] ;

[0118] ;

[0119] In the comparison method, the statistical error between the mean and quantile of the load shedding in the typical set is:

[0120] ;

[0121] ;

[0122] The comparison shows that, in the same typical scenarios, the number of Under the same balanced coverage constraint, compared with the unweighted comparison method, the method in this embodiment can reduce the approximation error of the optimal power flow load statistics (mean and P95), thereby improving the reliability and engineering applicability of subsequent vulnerability assessment.

[0123] In summary, the method for generating a weighted typical fault scenario set for power systems oriented towards extreme events in this embodiment first obtains an original fault scenario matrix containing the time periods of line fault occurrence; it then constructs a DC power flow model based on power system network parameters and calculates the branch power transfer distribution factor (PTDF); it determines the injection weights based on bus load and generation scale, thereby obtaining the line load-generation weighted sensitivity weights; under these weights, it constructs a weighted scenario distance that simultaneously represents the differences in the line fault set and the differences in fault time periods; and, under the capacity constraint that each typical scenario covers a basically consistent number of original scenarios, it uses a k-medoids representative point selection algorithm with capacity constraints to generate a typical fault scenario set, and outputs the equal weights of the typical scenarios and the mapping relationship between the original scenarios. This embodiment's method can enhance the characterization of fault modes of critical electrical lines while ensuring the balanced statistical representativeness of typical scenarios, significantly improving the efficiency and accuracy of subsequent optimal power flow calculations and vulnerability analysis.

[0124] Those skilled in the art will understand that the technical solutions provided by this invention may take the form of a method, system, or computer program product. Therefore, this invention may take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this invention may take the form of a computer program product embodied on one or more computer-readable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code. This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, produce an implementation of the flowchart. Figure 1 One or more processes and / or boxes Figure 1 The computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to operate in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The functions specified in one or more boxes. These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable apparatus for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0125] The above description is merely a preferred embodiment of the present invention. The scope of protection of the present invention is not limited to the above embodiments. All technical solutions falling within the scope of the present invention's concept are within the scope of protection of the present invention. It should be noted that for those skilled in the art, any improvements and modifications made without departing from the principles of the present invention should also be considered within the scope of protection of the present invention.

Claims

1. A method for generating a set of typical power system fault scenarios oriented towards extreme events, characterized in that, The following steps are included: S1: Obtain the original fault scenario matrix The original fault scenario matrix Each row in the matrix corresponds to a fault scenario, and each column corresponds to a line. The matrix elements represent the fault occurrence time or non-fault status of the line in the corresponding fault scenario. S2: Establish a DC power flow model based on the network parameters of the power system to be evaluated and calculate the line power transfer distribution factor. Calculate the bus injection weight based on the load and generation scale of each bus. S3: Calculate and normalize the line load-generation weighted sensitivity weight based on the line power transfer distribution factor and the bus injection weight. S4: Construct a single-line difference function that simultaneously characterizes the difference between whether a line is faulty and the difference in the time of fault occurrence; obtain the scenario distance by weighting the single-line difference function according to the line load-generation weighted sensitivity weight; S5: Process the original fault scenario matrix... Clustering is performed using a specified clustering algorithm based on scene distance to obtain a set of typical fault scenes after scene reduction and then output. Step S2, calculating the line power transfer distribution factor, includes: designating a balanced bus for the DC power flow model consisting of lines and buses, with the remaining buses being unbalanced buses; applying unit active power injection to the unbalanced buses and extracting unit active power from the balanced buses, and calculating the line power transfer distribution factor according to the following formula: ； in, For the first Line 1 to the 1st The power transfer distribution factor of the busbar For the first Changes in active power flow along each line. For the first The change in active power injection for each bus; the functional expression for calculating the bus injection weight based on the load and generation capacity of each bus in step S2 is as follows: ； in, For the first Busbar weight injection; and The first and Busbar load of the busbar and They are respectively and The absolute value, and The first and The power generation capacity of the busbar and They are respectively and The absolute value, This indicates all buses in the power system to be evaluated. and Summing is performed, and the sum of the bus injection weights of all buses in the power system to be evaluated is 1.

2. The method for generating a set of typical power system fault scenarios oriented towards extreme events according to claim 1, characterized in that, In step S3, the function expression for calculating and normalizing the line load-generation weighted sensitivity weight based on the line power transfer distribution factor and the bus injection weight is as follows: ； ； in, For the first The load-generation weighted sensitivity of each line. For the first Busbar weight injection; For the first Line 1 to the 1st The absolute value of the line power transfer distribution factor of each busbar. For the first The normalized load-generation weighted sensitivity weights for each line. For the first Load-generation weighted sensitivity of each line.

3. The method for generating a set of typical power system fault scenarios oriented towards extreme events according to claim 1, characterized in that, The functional expression of the single-line difference function constructed in step S4, which simultaneously characterizes the difference between whether a line is faulty and the difference between the time of fault occurrence, is as follows: ； in, For the scene With Scene On the line The single-line difference function value, For the scene Downline Fault identifiers for the corresponding time period For the scene Downline The fault flag for the corresponding time period is set to 0, indicating that no fault has occurred. Original fault scenario matrix The number of time periods in the time period.

4. The method for generating a set of typical power system fault scenarios oriented towards extreme events according to claim 1, characterized in that, The function expression for calculating the scene distance in step S4 is: ； in, Let be the scene distance from fault scene s to fault scene i. For the first The normalized load-generation weighted sensitivity weights for each line. For the scene With Scene On the line The single-line difference function value, This refers to the number of lines.

5. The method for generating a set of typical power system fault scenarios oriented towards extreme events according to claim 1, characterized in that, Step S5 includes: S5.1: Obtain the number of given typical failure scenarios and capacity Capacity in typical fault scenarios satisfy: ； in, Original fault scenario matrix The number of fault scenarios in; S5.2: Constructing a matrix of original fault scenarios The objective function for clustering based on scene distance; S5.3: Under the premise of ensuring that the coverage of each cluster is consistent, the k-medoids representative point selection algorithm is used to select points based on the objective function of the original fault scenario matrix. Clustering is performed to obtain a set of typical fault scenarios after scenario reduction and then output.

6. The method for generating a set of typical power system fault scenarios oriented towards extreme events according to claim 5, characterized in that, In step S5.2, the function expression of the objective function is: ； in, for The set of representative points of each cluster. For the k-th cluster, Let be any fault scenario s in the k-th cluster and the representative point of the k-th cluster. Scene distance, This represents the number of typical failure scenarios.

7. The method for generating a set of typical power system fault scenarios oriented towards extreme events according to claim 5, characterized in that, Step S5.3 includes: S5.3.1: Distance-based initialization is used, including: from the original fault scenario matrix A fault scenario is randomly selected as the first representative point. For any candidate scenario Calculate the nearest distance to the selected representative point set using the scene distance function. : ； in, Candidate scenarios To the representative point The scene distance is determined by the nearest distance to the set of selected representative points. square The next representative point is selected from the candidate scenes with a proportional probability, and this process is repeated until a result is obtained. The initial set of representative points is obtained by selecting 1 set of distinct representative points. ,in, ~ These are the initial representative point sets. K representative points in the middle, Original fault scenario matrix The number of fault scenarios in; S5.3.2: In the current fixed The set of representative points of each cluster Under the condition of each scenario As a node with a supply of 1, each representative point Capacity as a typical fault scenario Nodes in the process, to select from candidate scenarios To the representative point Scene distance The cost of each edge is used to establish a minimum cost allocation model with capacity constraints. By solving the minimum cost allocation model with capacity constraints, an allocation scheme that satisfies the capacity constraints and minimizes the total cost is obtained. ; The decision variables used in the minimum cost allocation model with capacity constraints are: ； in, Candidate scenarios Downline Fault identification within the corresponding time period; the objective function of the minimum cost allocation model with capacity constraints is expressed as follows: ； in, For the candidate scene set, From scene s to representative point The scene distance; the functional expression of the constraint condition used in the minimum cost allocation model with capacity constraints is: ； ； ； S5.3.3: Update the representative points for each cluster: ； in," "For update operations, Let be the scene distance from fault scene s to fault scene j. This represents the k-th cluster; S5.3.4: Determine whether the objective function has converged or the representative point no longer changes. If not, jump to step S5.3.2 and continue iterating; otherwise, convert the final set of representative points obtained from the iteration. corresponding A representative fault scenario is output as the typical fault scenario set.

8. The method for generating a set of typical power system fault scenarios oriented towards extreme events according to claim 7, characterized in that, In step S5.3.4, the final set of representative points obtained from the iteration is... corresponding When a representative fault scenario is output as the obtained typical fault scenario set, the function expression of the obtained typical fault scenario set is: ； in, For the obtained typical fault scenario set, This indicates the origin of the original fault scenario matrix. Extracting from the middle The set of representative points of each cluster. Original fault scenario matrix Number of time periods The number of typical failure scenarios. This refers to the number of lines.

9. The method for generating a set of typical power system fault scenarios oriented towards extreme events according to claim 7, characterized in that, In step S5.3.4, the final set of representative points obtained from the iteration is... corresponding When outputting a representative fault scenario as the typical fault scenario set, the output also includes a binary backoff matrix and an allocation scheme that satisfies the capacity constraint and minimizes the total cost. The functional expression for the binary devolution matrix is: ； in, For the first The first typical fault scenario The binary decommissioning status of the line constitutes the first line in the binary decommissioning matrix. OK Column elements; Indicates the line In the Under typical fault scenarios, it is in operation. Indicates the line In the Under typical fault scenarios, it is in a state of being returned to service; Original fault scenario matrix The Middle Line number The elements of the column represent the lines under the corresponding original fault scenarios. The fault occurrence period identifier; a value of 0 indicates no fault, and a value of 1 indicates a fault occurring during a specified period. This indicates that a fault occurred within the corresponding time period.

Images