A power system scheduling optimization method and system under abnormal conditions of a transformer

Abstract

The present invention relates to a power system scheduling optimization method and system under abnormal conditions of transformers. The method includes the following steps: collecting high-dimensional data of each transformer in real time to construct a high-dimensional data matrix; performing eigen-spectrum analysis on the high-dimensional data matrix using random matrix theory to identify abnormal time periods; constructing an augmented matrix during the abnormal time periods to perform eigen-spectrum analysis on the transformers to locate abnormal transformers; based on the abnormal time periods and the located abnormal transformers, constructing a scheduling optimization model with the objectives of minimizing operating costs and carbon emissions and solving it to complete the scheduling optimization process under abnormal conditions. Compared with the prior art, the present invention has the advantages of improving the accuracy of transformer abnormalities and realizing the stable operation of the power system, etc.

Description

Technical Field

[0001] The present invention relates to the technical field of smart grids, and particularly to a method and system for optimizing the power system scheduling under abnormal conditions of transformers. Background Art

[0002] With the continuous increase in the scale and complexity of power systems, transformers, as an important part of the power grid, their abnormal detection has become an important task to ensure the stability and reliability of power systems. Traditional methods for transformer abnormal detection usually rely on threshold setting, simple statistical analysis, or empirical rules. However, these methods often prove inadequate when faced with large-scale, high-dimensional data and are difficult to handle the changing operating states and complex fault modes in the system. In modern power systems, the operating environment of transformers involves multiple variables, such as load fluctuations, temperature changes, oil quality changes, etc. These factors make the data of transformers highly dynamic, non-linear, and multi-dimensional. Therefore, traditional methods are prone to false alarms and missed detections and are difficult to accurately and quickly detect transformer abnormalities.

[0003] Random Matrix Theory (RMT), as an effective high-dimensional data analysis tool, has been widely applied in the abnormal detection of complex systems in recent years. By analyzing the characteristic spectrum changes of transformer operating data, RMT can identify abnormal fluctuations in the system state, especially showing its unique advantages in a large-scale data environment. However, the direct application of random matrix theory usually can only detect abnormal time periods and cannot accurately locate the specific components or positions where abnormalities occur in the transformer. This makes it difficult for methods relying solely on random matrices to provide clear location information for transformer fault diagnosis in a complex power grid environment. To solve this problem, the augmented matrix method has been introduced into the abnormal detection of transformers. The augmented matrix method can effectively refine the abnormal detection results by jointly analyzing the specific transformer data of the transformer and the overall system data, and then improve the recognition accuracy of specific fault positions. Through this method, the abnormal location of the transformer can become more accurate and has better adaptability to non-linear features in the data, thus improving the accuracy and timeliness of abnormal detection. However, existing methods for identifying transformer abnormalities based on random matrices and augmented matrices still lack a systematic integration scheme in practical applications. Especially in power systems, an efficient and reliable high-dimensional abnormal recognition system has not been established to comprehensively address the challenges in transformer abnormal detection. Summary of the Invention

[0004] The purpose of the present invention is to provide a method and system for optimizing the power system scheduling under abnormal conditions of transformers to improve the operating stability of the power system under abnormal conditions of transformers.

[0005] The object of the present invention can be achieved by the following technical solutions:

[0006] A power system dispatching optimization method under abnormal conditions of a transformer, characterized by comprising the following steps:

[0007] Collect high-dimensional data of each transformer in real time and construct a high-dimensional data matrix;

[0008] Use random matrix theory to perform eigen-spectrum analysis on the high-dimensional data matrix to identify abnormal time periods;

[0009] Construct an augmented matrix during the abnormal time period to perform eigen-spectrum analysis on the transformer and locate the abnormal transformer;

[0010] Based on the abnormal time period and the located abnormal transformer, construct a dispatching optimization model with the goal of minimizing operating costs and carbon emissions and solve it to complete the dispatching optimization process under abnormal conditions.

[0011] Further, the step of constructing the high-dimensional data matrix includes:

[0012] Construct the high-dimensional data of each transformer into an initial high-dimensional data matrix , where the initial high-dimensional data matrix is:

[0013] In the formula, represents the high-dimensional data of the i-th transformer operating at the j-th time point, and its dimension is , m is the number of transformers in the substation, n is the number of time sampling points;

[0014] Perform preprocessing on the initial high-dimensional data matrix to obtain the final high-dimensional data matrix.

[0015] Further, the preprocessing step includes:

[0016] Perform normalization processing on the initial high-dimensional data matrix , and the normalization expression is:

[0017] In the formula, and respectively represent the minimum value and the maximum value in all time data of transformer i; is the data value after normalization, satisfying ;

[0018] According to the normalization result, use the moving average method for noise filtering to obtain the final high-dimensional data matrix, where the expression of the moving average method is:

[0019] In the formula, is the data after denoising processing; k is the radius of the sliding window, which determines the degree of smoothing; is the normalized data of transformer i at the p-th time point.

[0020] Further, the step of identifying the abnormal time period includes:

[0021] Construct a covariance matrix: Based on the high-dimensional data matrix, construct a covariance matrix , where the covariance matrix is:

[0022] In the formula, n is the number of time sampling points, is the high-dimensional data matrix;

[0023] Calculate the eigenvalue distribution: Calculate the eigenvalues of the covariance matrix to obtain a set of eigenvalues , where m is the number of transformers, is the k-th eigenvalue of the covariance matrix;

[0024] Calculate the average spectral radius: Based on the eigenvalue distribution, calculate the average spectral radius as a monitoring index of the transformer state, where the calculation expression of the average spectral radius is:

[0025] In the formula, MSR is the average spectral radius;

[0026] Identify the abnormal time period: Judge whether the average spectral radius MSR exceeds the set threshold . If so, it means that the substation is in an abnormal state and is marked as an abnormal time period. If not, it is not marked.

[0027] Further, the step of locating the abnormal transformer includes:

[0028] Construct an augmented matrix: Combine the high-dimensional data of a single transformer with the high-dimensional data matrix to construct an augmented matrix, where the augmented matrix is:

[0029] In the formula, represents the data vector of the i-th transformer at time, with a dimension of , is the high-dimensional data matrix, is the augmented matrix, with a dimension of , n is the number of time sampling points,m is the number of transformers;

[0030] Construct an augmented covariance matrix: Based on the augmented matrix, construct an augmented covariance matrix, where the augmented covariance matrix is:

[0031] In the formula, is the augmented covariance matrix, with a dimension of ;

[0032] Calculate the eigenvalue distribution of the augmented matrix: Calculate the set of eigenvalues of the augmented covariance matrix, where is the k-th eigenvalue of the augmented covariance matrix;

[0033] Calculate the augmented average spectral radius: Based on the eigenvalue distribution of the augmented matrix, calculate the average spectral radius of the augmented covariance matrix, where the calculation expression of the average spectral radius is:

[0034] In the formula, AMSR is the average spectral radius;

[0035] Identify abnormal transformers: Judge whether the average spectral radius AMSR exceeds the set threshold . If so, it is determined that the transformer i is abnormal. If not, it is considered that the transformer i is normal.

[0036] Furthermore, the scheduling optimization model includes an objective function and corresponding constraint conditions. The objective function is:

[0037] In the formula, Z is the total optimization target value; is the unit power cost coefficient of the i-th type of energy; is the scheduling power of the i-th type of energy at time t; is the carbon emission weight coefficient; is the unit power carbon emission coefficient of the i-th type of energy; is the load and voltage change coefficient of the i-th type of transformer, reflecting the impact of transformer abnormalities on scheduling; is the weight coefficient of the abnormal state of the transformer, controlling the priority of transformer abnormalities in the optimization target;

[0038] The constraint conditions include:

[0039] (1) Power limit constraint (2) Load demand balance constraint, (3) Energy storage constraint (4) Transformer safe operation constraint In the formula, is the maximum output power of the j-th type of energy, is the output power of the j-th type of energy at time t, m is the number of energy types, is the total load demand of the system at time t, is the storage state of the energy storage device at time t, is the change in the charge and discharge amount, and are the temperature and voltage of the transformer respectively, and are the maximum values for its safe operation.

[0040] Furthermore, it also includes an abnormal alarm step, specifically including:

[0041] Based on the identified abnormal time period and the located abnormal transformer, automatically generate and output an alarm message, where the alarm message includes the abnormal time period, the location of the abnormal transformer, and the degree of abnormality, and the degree of abnormality is determined according to the deviation degree of the distribution of the transformer characteristic values.

[0042] Furthermore, the calculation expression of the deviation degree is:

[0043] In the formula, is the deviation index, M is the number of transformers, is the actual value of the transformer MSR, d is the ideal value of the transformer MSR, is the metric function.

[0044] Furthermore, an optimization algorithm is used for solving, and the optimization algorithm includes one of a linear programming algorithm, a genetic algorithm, an ant colony algorithm, and a particle swarm algorithm.

[0045] The present invention also provides a power system scheduling optimization system under the abnormal state of a transformer, including:

[0046] Matrix construction module: used to collect high-dimensional data of each transformer in real time and construct a high-dimensional data matrix;

[0047] Abnormal time period identification module: used to perform eigen-spectrum analysis on the high-dimensional data matrix by using random matrix theory to identify the abnormal time period;

[0048] Abnormal transformer location module: used to construct an augmented matrix during the abnormal time period to perform eigen-spectrum analysis on the transformer and locate the abnormal transformer;

[0049] Dispatch optimization module: used to construct and solve a dispatch optimization model with the goal of minimizing operating costs and carbon emissions based on the abnormal time period and the located abnormal transformer, and complete the dispatch optimization process in the abnormal state.

[0050] Compared with the prior art, the present invention has the following beneficial effects:

[0051] (1) The present invention constructs a high-dimensional data matrix from the real-time high-dimensional data of each substation, and uses the random matrix eigenvalue spectrum analysis technology and the augmented matrix eigenvalue spectrum analysis technology to capture the abnormal time period and abnormal location of the transformer through eigenvalue analysis in sequence, improving the accuracy of abnormal transformer detection, and improving the stability of the power system operation in the abnormal state of the transformer by performing dispatch optimization on the power system in the abnormal state.

[0052] (2) Compared with simple abnormal recognition methods, the augmented matrix eigenvalue spectrum analysis technology adopted by the present invention can achieve precise positioning of abnormal transformers through the combination of specific transformer data and overall substation data, greatly improving the detection accuracy.

[0053] (3) The dispatch optimization model of the present invention ensures that under the condition of minimizing operating costs and carbon emissions, the dispatch power meets the power and load balance requirements of the system, enabling reasonable dispatch and risk control through real-time dispatch optimization even in the abnormal state of the transformer.

[0054] (4) Compared with traditional single-dimensional data abnormal monitoring, the present invention adopts a multi-dimensional data-driven alarm mechanism, uses high-dimensional data for real-time dynamic monitoring, ensures that the control personnel can quickly and intuitively understand the abnormal state of the system, and improves the power system situation awareness level. Description of the Drawings

[0055] Figure 1 is a schematic flow chart of the method of the present invention;

[0056] Figure 2 is a flow chart for locating abnormal transformers of the present invention. Detailed Embodiments

[0057] The present invention will be described in detail below with reference to the drawings and specific embodiments. This embodiment is implemented on the premise of the technical solution of the present invention, and gives the detailed implementation manner and specific operation process, but the protection scope of the present invention is not limited to the following embodiments.

[0058] Embodiment 1

[0059] This embodiment provides a power system dispatch optimization method under the abnormal state of a transformer, as Figure 1 shown, the method includes the following steps:

[0060] Step 1: Perform high-dimensional data acquisition and preprocessing on each transformer.

[0061] For the substations in the power system, high-dimensional data including oil temperature, carbon dioxide, carbon monoxide and other gas concentrations are collected in real time from each transformer in the substation. These data form a big data matrix for subsequent feature spectrum analysis. The steps for high-dimensional data acquisition and preprocessing based on the operation of the substation are as follows:

[0062] Step 1.1: Data acquisition. Assume that the data matrix of the substation collected is denoted as , and its dimension is , where: m represents the number of transformers in the substation; n represents the number of time sampling points.

[0063] In the data matrix , each row represents the operation status data of a transformer at different time points, and each column represents the data of different transformers at the same time point. The form of the data matrix is as follows:

[0064]

[0065] Among them, represents the operation data of the i-th transformer at the j-th time point.

[0066] Step 1.2: Data normalization. The original data matrix is normalized to eliminate the differences caused by different dimensions and ensure that the data is analyzed on the same scale. The normalization process is achieved through the following formula:

[0067]

[0068] Among them: is the original data value of the i-th transformer at the j-th time point; and represent the minimum and maximum values of all time data of transformer i respectively; is the normalized data value, satisfying .

[0069] The normalization process can limit all data within the same range, making subsequent analysis more stable and reliable.

[0070] Step 1.3: Data denoising. To improve the stability and accuracy of the data, a denoising method is used to filter the noise from the normalized data matrix . Assume that the noise in the data conforms to a Gaussian distribution. The moving average method can be used to smooth the data and remove the noise interference caused by short-term fluctuations. The formula is as follows:

[0071]

[0072] Wherein: is the data after denoising processing; k is the radius of the sliding window, which determines the degree of smoothing; represents the normalized data of transformer i at the p-th time point.

[0073] The matrix after denoising processing is denoted as , wherein is the high-dimensional denoised data, which is used for further feature spectrum analysis to ensure the reliability of the analysis results.

[0074] Step 2: Based on the feature spectrum analysis of the random matrix, preliminarily identify the abnormal time period of the transformer.

[0075] After completing the data preprocessing, use the random matrix theory to perform feature spectrum analysis on the high-dimensional data matrix, so as to identify the abnormal time period of the transformer. In this step, the feature spectrum index (such as the average spectral radius) of the data matrix is calculated to judge whether there is an abnormality in the transformer.

[0076] Step 2.1: Construct the covariance matrix. Let the denoised data matrix after processing be , and its dimension is , where m is the number of transformers and n is the number of time sampling points. First, construct the covariance matrix , and the calculation formula is as follows:

[0077]

[0078] Wherein, has a dimension of , which reflects the correlation relationship between transformers.

[0079] Step 2.2: Calculate the eigenvalue distribution. By finding the eigenvalues of the covariance matrix , a set of eigenvalues is obtained. These eigenvalues reflect different dimensions of the transformer state. The distribution of eigenvalues can reveal the change of the transformer operation state, and the change of the feature spectrum may indicate the occurrence of transformer abnormality.

[0080] Step 2.3: Calculate the average spectral radius. To further quantify the distribution of eigenvalues, calculate the average spectral radius (Mean Spectral Radius, MSR) as a monitoring index of the transformer state. The calculation formula of the average spectral radius is as follows:

[0081] Wherein, is the k-th eigenvalue of the covariance matrix , and m is the number of transformers.

[0082] Step 2.4: Abnormal time period identification. When the mean spectral radius (MSR) exceeds a certain set threshold , the transformer is considered to be in an abnormal state, and this time period is marked as an abnormal time period. This threshold can be set through statistical analysis of historical data or experiments to ensure the sensitivity and accuracy of detection.

[0083] Step 2.5: Dynamic monitoring and updating. In this step, the MSR value is calculated and monitored in real time during the operation of the substation to achieve dynamic abnormal detection. If the MSR exceeds the threshold , then this time period is automatically marked as abnormal and subsequent abnormal location analysis is triggered.

[0084] Through the analysis of the eigen-spectrum of the random matrix, this part can quickly identify abnormal time periods in high-dimensional data, providing a basis for the next step of abnormal transformer location.

[0085] Step 3: Augmented matrix construction and abnormal transformer location.

[0086] After identifying the abnormal time period, in order to further determine the specific transformer with abnormalities in the substation, the present invention adopts the augmented matrix method. By performing eigen-spectrum analysis on the data of a specific transformer, the precise location of the abnormal transformer is achieved. As Figure 2 shown, this step includes:

[0087] Step 3.1: Construct an augmented matrix. After detecting the abnormal time period, select the transformers in the system that may have abnormalities and construct an augmented matrix , and combine the data of this transformer with the overall data matrix. Let the data vector of the abnormal transformer be , and the form of the augmented matrix is:

[0088]

[0089] Where: is the processed high-dimensional denoised data matrix with a dimension of ; represents the data vector of the i-th transformer at time t with a dimension of ; the dimension of the augmented matrix is .

[0090] Step 3.2: Calculate the augmented covariance matrix. Through the augmented matrix , construct the augmented covariance matrix for eigen-spectrum analysis. The formula is as follows:

[0091]

[0092] Augmented covariance matrix has a dimension of , which contains the correlation information of the overall data and the target transformer data.

[0093] Step 3.3: Augmented matrix eigenvalue spectrum analysis, calculate the augmented covariance matrix eigenvalue set of . These eigenvalues reflect the performance of the target transformer in the overall state of the system. When the target transformer is abnormal, the distribution of its eigenvalues will shift significantly.

[0094]

[0095] Step 3.4: Augmented mean spectral radius calculation, calculate the augmented mean spectral radius (AMSR) of the augmented matrix for quantitative analysis of the state of the target transformer. The formula is as follows:

[0096]

[0097] where is the k-th eigenvalue of the augmented covariance matrix .

[0098] Step 3.5: Abnormal transformer identification. If the augmented mean spectral radius AMSR exceeds the set threshold of the augmented matrix , it is determined that the target transformer i is abnormal. This threshold can be set based on historical data or experiments to ensure the sensitivity and accuracy of transformer anomaly detection.

[0099] Through the augmented matrix eigenvalue spectrum analysis, this part can further identify the specific abnormal transformers during the abnormal time period, achieve precise positioning of power system anomalies, and provide a reliable basis for the system's abnormal response and dispatching decision-making.

[0100] Step 4: Abnormal alarm and decision support.

[0101] After detecting the abnormal time period and locating the abnormal transformer, the present invention generates alarm information and provides corresponding dispatching decision support for the power system. This step combines the abnormal detection results with the dispatching strategy to ensure reasonable dispatching and risk control under abnormal conditions.

[0102] Step 4.1: Abnormal alarm generation. Once the abnormal transformer and the abnormal time period are identified, the system automatically generates alarm information to notify the relevant control personnel of the abnormal situation. The alarm information includes the abnormal time period, the location of the abnormal transformer, and the degree of abnormality, etc., providing intuitive real-time feedback.

[0103] Step 4.2: Scheduling Optimization Objective. Under abnormal conditions, to ensure the stable operation of the power system and minimize carbon emissions, the present invention proposes an optimization objective function that combines operating costs and carbon emissions. The form of this objective function is as follows:

[0104]

[0105] Where: Z is the total optimization objective value of the system; is the unit power cost coefficient of the j-th type of energy; is the scheduled power of the j-th type of energy at time t; is the carbon emission weight coefficient, which controls the priority of carbon emissions; is the unit power carbon emission coefficient of the j-th type of energy.

[0106] Step 4.3: Constraint Conditions. Scheduling optimization needs to meet the following constraint conditions to ensure the physical feasibility and safety of the system, especially considering the operating limitations of transformers.

[0107] 1) Power Limit Constraint:

[0108]

[0109] Where, is the maximum output power of the j-th type of energy, ensuring that the scheduled power is within the allowable range.

[0110] 2) Load Demand Balance Constraint:

[0111]

[0112] Where, is the total load demand of the system at time t, ensuring that the system's load demand is met at each time point.

[0113] 3) Energy Storage Constraint. For energy storage devices, the dynamic balance of energy storage needs to be satisfied. The formula is as follows:

[0114]

[0115] Where, is the storage state of the energy storage device at time t, is the change in the charge and discharge amount.

[0116] 4) Transformer Safe Operation Constraint. For transformer equipment, the safety constraints during its operation need to be satisfied. For example, parameters such as temperature, oil pressure, and voltage must be within the safe range. The formula is as follows:

[0117]

[0118] Where, and are the temperature and voltage of the transformer respectively, and are the maximum values for its safe operation.

[0119] Step 4.4: Solve the optimal scheduling strategy. Based on the above objective function and constraint conditions, use an optimization algorithm (such as linear programming algorithm or genetic algorithm or ant colony algorithm or particle swarm algorithm) to solve and obtain the optimal scheduling power distribution , and the formula is as follows:

[0120]

[0121] The optimal scheduling strategy realizes the reasonable allocation of resources under abnormal conditions, meets the load demand while minimizing carbon emissions and operating costs.

[0122] Step 4.5: Dynamic feedback and adjustment. While the system executes the scheduling decision, it monitors the data changes of various evaluation indicators in real time, and adjusts the scheduling strategy when new abnormal situations occur, forming a closed-loop feedback to ensure the dynamic adaptability and intelligence of the scheduling plan.

[0123] The following uses a specific embodiment to illustrate the technical solution of the present invention.

[0124] (1) Obtain the high-dimensional operation data of each transformer in the substation, including the oil temperature, carbon dioxide, carbon monoxide and other gas concentrations, form the original data matrix, and normalize the data to ensure the consistency of the data under the same dimension;

[0125] Based on the random matrix eigenvalue spectrum analysis, construct the covariance matrix of the data and calculate the eigenvalue set. Monitor the transformer status by calculating the average spectral radius. When this value exceeds the set threshold, mark this time period as an abnormal time period;

[0126] (3) During the abnormal time period, select the transformer that may be abnormal, construct an augmented matrix, combine the data of this transformer with the overall data, and calculate the average spectral radius of the augmented matrix. When the augmented spectral radius exceeds the set threshold, determine that this transformer is an abnormal transformer to achieve accurate positioning;

[0127] (4) Generate alarm information based on the abnormal detection results and provide scheduling suggestions. Through the scheduling optimization objective, ensure that under the condition of minimizing the operating cost and carbon emissions, the scheduling power meets the power and load balance requirements of the system, and finally obtain the optimal scheduling plan.

[0128] Embodiment 2

[0129] This embodiment provides a power system scheduling optimization system under the abnormal state of the transformer, including:

[0130] Matrix construction module: used to collect high-dimensional data of each transformer in real time and construct a high-dimensional data matrix;

[0131] Abnormal time period identification module: used to perform eigen-spectrum analysis on the high-dimensional data matrix by using random matrix theory to identify abnormal time periods;

[0132] Abnormal transformer positioning module: used to construct an augmented matrix during the abnormal time period to perform eigen-spectrum analysis on the transformer and locate the abnormal transformer;

[0133] Dispatch optimization module: used to construct and solve a dispatch optimization model with the goal of minimizing operating cost and carbon emissions based on the abnormal time period and the located abnormal transformer, and complete the dispatch optimization process in the abnormal state.

[0134] The rest is the same as in Embodiment 1.

[0135] If the above functions are implemented in the form of software function units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage medium includes: USB flash drives, mobile hard disks, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), magnetic disks, or optical discs, etc., which can store program codes.

[0136] Those skilled in the art should understand that the embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes. The solutions in the embodiments of the present invention can be implemented in various computer languages. For example, object-oriented programming languages such as Java and interpreted scripting languages such as JavaScript.

[0137] The present invention is described with reference to the flowcharts and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It should be understood that each flow and / or block in the flowchart and / or block diagram, and combinations of flows and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing device to produce a machine, such that the instructions executed by the processor of the computer or other programmable data processing device generate means for implementing the functions specified in the flowchart Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0138] These computer program instructions can also be stored in a computer-readable memory 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 memory produce a manufacture including instruction means that implement the functions specified in the flowchart Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0139] These computer program instructions can also be loaded onto a computer or other programmable data processing device, such that a series of operational steps are performed on the computer or other programmable device to produce a computer-implemented process, and thus the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in the flowchart Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0140] Although the preferred embodiments of the present invention have been described, additional changes and modifications can be made by those skilled in the art once they learn of the basic inventive concept. Therefore, the appended claims are intended to be construed to include the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0141] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalent technologies, the present invention is also intended to include these modifications and variations.

Claims

1. A method for optimizing power system dispatching under abnormal transformer conditions, characterized in that: The following steps are involved: Collect high-dimensional data of each transformer in real time and construct a high-dimensional data matrix; Using random matrix theory to perform characteristic spectrum analysis on the high-dimensional data matrix to identify abnormal time periods; Constructing an augmented matrix during the abnormal time period to perform characteristic spectrum analysis on the transformer and locate the abnormal transformer; Based on the abnormal time period and the located abnormal transformer, a scheduling optimization model is constructed and solved with the goal of minimizing operating costs and carbon emissions, completing the scheduling optimization process under abnormal conditions; The step of identifying an abnormal time period comprises: Construct a covariance matrix: Based on the high-dimensional data matrix, construct a covariance matrix C t , where the covariance matrix C t for: Where n is the number of time sampling points, is a high-dimensional data matrix; Calculate the eigenvalue distribution: calculate the eigenvalue of the covariance matrix to obtain a set of eigenvalues ​​{λ1, λ2, ..., λ k , ..., λ m }, where m is the number of transformers, λ k is the kth eigenvalue of the covariance matrix; Calculating the average spectrum radius: Based on the characteristic value distribution, the average spectrum radius is calculated as a monitoring indicator of the transformer state, wherein the calculation expression of the average spectrum radius is: Where MSR is the mean spectral radius; Identify abnormal time period: determine whether the average spectrum radius MSR exceeds the set threshold MSR th ,If so, it means that the substation is in an abnormal state and is marked as an abnormal time period. If not, it is not marked; The step of locating the abnormal transformer comprises: Constructing an augmented matrix: combining the high-dimensional data of a separate transformer with the high-dimensional data matrix to construct an augmented matrix, wherein the augmented matrix is: Where, X i，t represents the data vector of the i-th transformer at time, with a dimension of 1×n, is a high-dimensional data matrix, is an augmented matrix with a dimension of (m+1)×n, where n is the number of time sampling points and m is the number of transformers; Constructing an augmented covariance matrix: Based on the augmented matrix, constructing an augmented covariance matrix, wherein the augmented covariance matrix is: In the formula, is the augmented covariance matrix with dimension (m+1)×(m+1); Calculate the eigenvalue distribution of the augmented matrix: Calculate the eigenvalue set of the augmented covariance matrix in is the kth eigenvalue of the augmented covariance matrix; Calculate the augmented average spectral radius: Calculate the average spectral radius of the augmented covariance matrix based on the eigenvalue distribution of the augmented matrix, wherein the calculation expression of the average spectral radius is: Where, AMSR is the average spectral radius; Identify abnormal transformers: determine whether the average spectrum radius AMSR exceeds the set threshold If yes, it is determined that transformer i has an abnormality, if no, it is considered that transformer i has no abnormality; The scheduling optimization model includes an objective function and corresponding constraints, and the objective function is: Where Z is the total optimization target value; C i is the unit power cost coefficient of the i-th type of energy; P i is the dispatch power of the i-th energy source at time t; i is the carbon emission weight coefficient; E i is the unit power carbon emission coefficient of the i-th type of energy; V i is the load and voltage variation coefficient of the i-th type transformer, reflecting the impact of transformer abnormality on dispatch; μ i is the weight coefficient of the transformer abnormal state, which controls the priority of the transformer abnormality in the optimization target; The constraints include: (1) Power limit constraint 0≤P j，t ≤P max，j (2) Load demand balancing constraints, (3) Energy storage constraint E t+1 =E t +ΔE t (4) Transformer safe operation constraints T i ≤T i max , V i ≤V i max Where P max，j is the maximum output power of the jth energy source, P j,t is the output power of the jth energy at time t, m is the number of energy types, D t is the total load demand of the system at time t, E t is the storage state of the energy storage device at time t, ΔE t is the change of charge and discharge capacity, T i and V i are the temperature and voltage of the transformer, T i max and V i max The maximum value for its safe operation.

2. The method for optimizing power system dispatching under abnormal transformer state according to claim 1, characterized in that: The step of constructing a high-dimensional data matrix comprises: The high-dimensional data of each transformer is constructed as an initial high-dimensional data matrix X t , where the initial high-dimensional data matrix X t for: In the formula, X i,j Represents the high-dimensional data of the operation of the i-th transformer at the j-th time point, with a dimension of m×n, where m is the number of transformers in the substation and n is the number of time sampling points; The initial high-dimensional data matrix is ​​preprocessed to obtain a final high-dimensional data matrix.

3. The method for optimizing power system dispatching under abnormal transformer state according to claim 2, characterized in that: The pre-processing step comprises: For the initial high-dimensional data matrix X t Normalization is performed, where the normalized expression is: In the formula, min(X i，： ) and max(X i，； ) represent the minimum and maximum values ​​of all time data of transformer i respectively; X i，j′ is the normalized data value, satisfying 0≤X i，j′ ≤1; According to the normalized results, the sliding average method is used to filter the noise and obtain the final high-dimensional data matrix, where the expression of the sliding average method is: In the formula, is the data after denoising; k is the sliding window radius, which determines the degree of smoothing; X i，p′ is the normalized data of transformer i at the pth time point.

4. The method for optimizing power system dispatching under abnormal transformer state according to claim 1, characterized in that: It also includes abnormal alarm steps, including: Based on the identified abnormal time period and the location of the abnormal transformer, alarm information is automatically generated and output, wherein the alarm information includes the abnormal time period, the location of the abnormal transformer and the degree of abnormality, wherein the degree of abnormality is determined according to the degree of deviation of the distribution of the transformer characteristic values.

5. The method for optimizing power system dispatching under abnormal transformer state according to claim 4, characterized in that: The calculation expression of the deviation degree is: In the formula, a m is the deviation index, M is the number of transformers, u m is the actual value of transformer MSR, d is the ideal value of transformer MSR, and h(·) is the measurement function.

6. The method for optimizing power system dispatching under abnormal transformer state according to claim 1, characterized in that: An optimization algorithm is used for solving the problem, and the optimization algorithm includes one of a linear programming algorithm, a genetic algorithm, an ant colony algorithm, and a particle swarm algorithm.

7. A dispatching optimization system based on the power system dispatching optimization method under abnormal transformer state according to any one of claims 1 to 6, characterized in that: include: Matrix construction module: used to collect high-dimensional data of each transformer in real time and construct a high-dimensional data matrix; Abnormal time period identification module: used to perform characteristic spectrum analysis on the high-dimensional data matrix using random matrix theory to identify abnormal time periods; Abnormal transformer positioning module: used to construct an augmented matrix within the abnormal time period to perform characteristic spectrum analysis on the transformer and locate the abnormal transformer; Dispatch optimization module: It is used to build and solve a dispatch optimization model based on the abnormal time period and the located abnormal transformers, with the goal of minimizing operating costs and carbon emissions, and complete the dispatch optimization process under abnormal conditions.

Images