Power grid safety margin online monitoring method and system for multi-level vulnerability assessment

Abstract

The invention discloses a power grid safety margin on-line monitoring method and system for multi-level vulnerability assessment, and relates to the technical field of power system safety margin on-line monitoring, and the method comprises the steps: building a multi-level power grid topological structure model based on 110kV and 35kV transformer substations and line sections, and calculating the node importance and the inter-level coupling degree; designing a static safety margin and a dynamic response characteristic index, and constructing a comprehensive vulnerability evaluation system; training an LSTM deep learning model based on the historical operation data; acquiring power grid operation data in real time, and performing standardization processing; evaluating the current safety margin by using the trained LSTM deep learning model; and judging whether early warning needs to be triggered or not according to an early warning grading mechanism, and outputting an evaluation result and early warning information. According to the invention, real-time monitoring and accurate prediction of the power grid safety margin are realized, weak links are effectively identified, the safety and reliability of the power grid are improved, an early warning grading mechanism ensures quick response and risk management, and stable operation and power supply safety of the power grid are guaranteed.

Description

Technical Field

[0001] The present invention relates to the technical field of online monitoring of power system safety margin, and in particular to a method and system for online monitoring of power grid safety margin for multi-level vulnerability assessment. Background Art

[0002] In the field of power system security monitoring, relevant technologies have made significant progress, especially in grid topology modeling, safety margin assessment and online monitoring methods. Traditional grid security assessment methods mostly rely on static analysis to evaluate the stability of the grid by simulating various operating scenarios. However, these methods often ignore the dynamic characteristics of the grid and the potential information in long-term operating data, resulting in the assessment results that may not be accurate and comprehensive.

[0003] Traditional grid safety margin assessment methods are mostly based on a single level, lacking consideration of the mutual impact between equipment at different voltage levels; and existing assessment methods mainly rely on static indicators, which are difficult to reflect the dynamic characteristics of the grid; traditional methods fail to effectively construct a multi-level grid topology model, and thus cannot accurately calculate node importance and inter-level coupling, which is particularly important in complex grid systems.

[0004] The existing safety margin assessment system often lacks consideration of dynamic response characteristics and cannot fully reflect the vulnerability of the power grid when facing emergencies.

[0005] In addition, although deep learning technology has been applied in many fields, its application in power grid safety margin monitoring is still insufficient, especially in using LSTM models to capture long-term dependencies in power grid operation data. Therefore, existing technologies are difficult to achieve real-time, accurate monitoring and early warning of power grid safety margins, which to some extent limits the safe operation and risk management of power grids. Summary of the invention

[0006] In view of the above-mentioned existing problems, the present invention aims to improve the safe operation level of the power grid by constructing a multi-level power grid topology model, designing an evaluation index system that combines static and dynamic evaluation, using a deep learning algorithm to achieve real-time evaluation of the safety margin, and establishing a multi-level collaborative hierarchical early warning mechanism.

[0007] In order to solve the above technical problems, a multi-level vulnerability assessment online monitoring method for power grid security margin is proposed, including:

[0008] A multi-level power grid topology model is constructed based on 110kV and 35kV substations and line sections, and the node importance and inter-level coupling degree are calculated; static safety margin and dynamic response characteristic indicators are designed to build a comprehensive vulnerability assessment system; the LSTM deep learning model is trained based on historical operation data; the power grid operation data is collected in real time and standardized; the current safety margin is evaluated using the trained LSTM deep learning model; the need to trigger an early warning is determined based on the early warning classification mechanism, and the evaluation results and early warning information are output.

[0009] As a preferred solution of the online monitoring method of the power grid safety margin for multi-level vulnerability assessment of the present invention, wherein: the construction of the multi-level power grid topology structure model includes considering power grid levels of different voltage levels and the interaction between power grid levels, and the construction of the multi-level power grid topology structure model includes a state space and an action space;

[0010] The state space includes new energy output, load demand, energy storage state and voltage level, as follows:

[0011] s t =[P re (t),P l (t),E(t),V(t)]

[0012] Among them, s t is the system state at time t, P re (t) is the new energy output at time t, P l (t) is the load demand at time t, E(t) is the remaining energy storage capacity at time t, and V(t) is the voltage level at time t;

[0013] The action space includes the output of conventional units and the charging and discharging power of the energy storage system, as follows:

[0014] a t =[P g (t),P s (t)]

[0015] Among them, a t is the control action at time t, P g (t) is the output of the conventional unit at time t, P s (t) is the energy storage output at time t.

[0016] As a preferred solution of the online monitoring method of the power grid safety margin for multi-level vulnerability assessment of the present invention, wherein: the calculation of node importance and inter-level coupling degree includes comprehensively evaluating the node importance index through the height, betweenness centrality and clustering coefficient of the node, and setting the power grid G=(V, E) based on the constructed multi-level power grid topology model, wherein V is a node set and E is an edge set;

[0017] Define the level set L = {l1, l2, l3}, corresponding to 110kV, 35kV and key section levels respectively, and the node importance index is as follows:

[0018] I v =w1D v +w2B v +w3C v

[0019] Among them, I v is the importance index of node v, D v is the node degree, which indicates the number of edges directly connected to node v; B v is the betweenness centrality, which indicates the frequency of a node as an intermediary of the shortest path between the remaining nodes in the network; C v is the clustering coefficient, which indicates the degree of interconnection between neighboring nodes of a node; w1, w2, and w3 are weight coefficients, which are used to adjust the influence of different indicators in importance evaluation;

[0020] The inter-level coupling degree is obtained by calculating the association strength between nodes at different levels:

[0021]

[0022] Among them, C ij is the coupling degree between levels i and j, V i 、V j are the node sets of level i and j respectively, α kl is the association strength between nodes k and l, indicating the degree of mutual influence between the two nodes.

[0023] As a preferred solution of the online monitoring method of power grid safety margin for multi-level vulnerability assessment of the present invention, wherein: the comprehensive vulnerability assessment system includes a static safety margin index and a dynamic response characteristic index;

[0024] The static safety margin index includes the safety margin of the evaluation system on different devices and nodes, specifically:

[0025]

[0026] Among them, S s is the static safety margin, P max is the rated capacity of the equipment, and P is the actual load;

[0027] The dynamic response characteristic indicators include response time, stability time and reliability, specifically:

[0028] S d =w1T resp+w2T stab +w3T rel

[0029] Among them, S d is the dynamic response characteristic index, T resp is the response time, i.e., the time required for the system to detect the disturbance and start responding; T stab is the stabilization time, that is, the time required for the system to return to a stable state after a disturbance; T rel It is a reliability index that reflects the ability of the system to maintain normal operation in the face of disturbances;

[0030] The static safety margin index and the dynamic response characteristic index are combined to obtain a comprehensive vulnerability index to comprehensively evaluate the vulnerability of the power system, specifically:

[0031] V=λS s +(1-λ)S d

[0032] Among them, V is the comprehensive vulnerability index and λ is the weight coefficient, which is used to balance the importance of static safety margin and dynamic response characteristics in comprehensive vulnerability assessment.

[0033] As a preferred solution of the online monitoring method of the power grid security margin for multi-level vulnerability assessment described in the present invention, the LSTM deep learning model includes capturing the long-term dependencies in the power system operation data through the LSTM deep learning model, and updating the hidden layer state, memory unit state and output respectively:

[0034] h t =σ(W h *[h t-1 ,x t ]+b h )

[0035] c t =f t ⊙c t-1 +i t ⊙g t

[0036] y t =o t ⊙tanh(c t )

[0037] Among them, h t The hidden state at time t contains all the information before time t, c t is the state of the memory cell at time t, allowing the LSTM network to remember long-term dependencies; x t The input sequence at time t contains the grid operation data at time t, W his the weight matrix, which is used to determine how the hidden state and input data affect the new hidden state and memory unit state; b h is the bias term, providing a benchmark for updating the hidden layer state and memory unit state, f t 、i t , o t They are the forget gate, input gate, and output gate, which control the updating of the memory cell content and the generation of the output; ⊙ represents element-wise multiplication, allowing the model to consider the importance of different information when updating the memory cell state.

[0038] As a preferred solution of the online monitoring method for the safety margin of a power grid for multi-level vulnerability assessment of the present invention, the current safety margin assessment includes collecting power grid operation data including voltage, current, active power, and reactive power in real time through an online monitoring system, and performing standardized processing on the power grid operation data:

[0039]

[0040] Among them, x′ is the standardized data, μ is the mean, and σ is the standard deviation;

[0041] Based on the trained LSTM deep learning model, the safety margin of the power grid is evaluated based on the standardized power grid operation data:

[0042] M=F LSTM (x')

[0043] Among them, M is the safety margin assessment result, F LSTM It is a trained LSTM deep learning model.

[0044] As a preferred solution of the online monitoring method for power grid safety margin of multi-level vulnerability assessment described in the present invention, the warning classification mechanism includes setting three levels of warning thresholds:

[0045] W={W1,W2,W3|W1>W2>W3}

[0046] Among them, W1, W2, and W are the first, second, and third level warnings respectively. The threshold determines the severity of the safety margin assessment results and takes corresponding warning measures.

[0047] Another object of the present invention is to provide an online monitoring system for power grid safety margin with multi-level vulnerability assessment. The present invention realizes online monitoring of power grid safety margin and timely discovers potential risks in power grid operation; through multi-level vulnerability assessment, the structural characteristics and dynamic response characteristics of the power grid are fully considered to improve the accuracy and comprehensiveness of the assessment; deep learning technology is used to process complex data of power grid operation to enhance the intelligence level and predictive ability of the monitoring system; according to the actual operating status of the power grid, the early warning level is dynamically adjusted to provide decision support for the safe and stable operation of the power grid.

[0048] As a preferred solution of the online monitoring system for power grid safety margin of multi-level vulnerability assessment described in the present invention, it is characterized by comprising a power grid topology building module, a vulnerability assessment system building module, an LSTM deep learning model training module, a data acquisition and preprocessing module, a safety margin assessment module, and an early warning classification mechanism module;

[0049] The grid topology construction module constructs a multi-level grid topology model based on 110kV, 35kV substations and line sections, and calculates node importance and inter-level coupling, providing basic grid structure data and inter-level relationships for subsequent vulnerability assessment;

[0050] The vulnerability assessment system construction module designs static safety margins and dynamic response characteristic indicators, builds a comprehensive vulnerability assessment system, and provides training objectives and evaluation criteria for the LSTM deep learning model;

[0051] The LSTM deep learning model training module trains the LSTM deep learning model based on historical operation data, captures the long-term dependencies in the power grid operation data, and outputs the trained model for use by the early warning classification mechanism module;

[0052] The data acquisition and preprocessing module collects power grid operation data in real time and performs standardized processing to provide real-time data input for the LSTM model;

[0053] The safety margin evaluation module receives the preprocessed data, uses the trained LSTM model to evaluate the safety margin of the current power grid, and outputs the safety margin evaluation result;

[0054] The warning classification mechanism module receives the safety margin assessment result, determines whether to trigger a warning according to the warning classification threshold, outputs the assessment result and warning information, and executes warning measures.

[0055] A computer device comprises a memory and a processor, wherein the memory stores a computer program, and is characterized in that when the processor executes the computer program, the steps of the online monitoring method of power grid safety margin for multi-level vulnerability assessment are implemented.

[0056] A computer-readable storage medium having a computer program stored thereon, characterized in that when the computer program is executed by a processor, the steps of the online monitoring method for power grid security margin of multi-level vulnerability assessment are implemented.

[0057] Beneficial effects of the present invention: By constructing a multi-level power grid topology model and calculating the node importance and the coupling degree between levels, the present invention realizes a detailed description and quantitative evaluation of the interdependence between key nodes and levels of the power grid, which not only helps to identify the weak links in the power grid, but also can take preventive measures in advance to improve the overall safety and reliability of the power grid. At the same time, by designing static safety margin and dynamic response characteristic indicators, and combining the LSTM deep learning model to analyze the power grid operation data, the present invention realizes real-time monitoring and accurate prediction of the power grid safety margin, thereby providing a scientific basis for power grid dispatching and operation and maintenance. The comprehensive use of these methods ensures that the power grid can respond in a timely manner when facing various operating conditions, effectively prevents potential risks, and ensures the stable operation and power supply safety of the power grid.

[0058] Furthermore, the present invention collects power grid operation data in real time and performs standardized processing, uses a trained LSTM model to evaluate the current safety margin, and outputs evaluation results and warning information according to the early warning classification mechanism, thereby realizing hierarchical monitoring and early warning of the power grid security status. This mechanism not only improves the response speed to power grid security incidents, but also enables power grid managers to take corresponding early warning measures according to the severity of the risk, thereby significantly improving the risk management level of the power grid, ensuring that the power grid can take measures quickly and effectively when facing an emergency, and maintaining the stability of the power grid and the continuity of power supply. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative work, among which:

[0060] Figure 1 An overall flow chart of a method for online monitoring of power grid safety margin with multi-level vulnerability assessment provided by an embodiment of the present invention.

[0061] Figure 2 A system solution module diagram of a power grid safety margin online monitoring system for multi-level vulnerability assessment provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0062] In order to make the above-mentioned purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in the art without creative work should fall within the scope of protection of the present invention.

[0063] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0064] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor is it an embodiment that is mutually exclusive with other embodiments, either individually or selectively.

[0065] The present invention is described in detail with reference to schematic diagrams. When describing the embodiments of the present invention, for the sake of convenience, the cross-sectional diagrams showing the device structure will not be partially enlarged according to the general scale, and the schematic diagrams are only examples, which should not limit the scope of protection of the present invention. In addition, in actual production, the three-dimensional dimensions of length, width and depth should be included.

[0066] At the same time, in the description of the present invention, it should be noted that the directions or positional relationships indicated by the terms "upper, lower, inner and outer" are based on the directions or positional relationships shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific direction, be constructed and operated in a specific direction, and therefore cannot be understood as limiting the present invention. In addition, the terms "first, second or third" are only used for descriptive purposes and cannot be understood as indicating or implying relative importance.

[0067] In the present invention, unless otherwise clearly specified and limited, the terms "install, connect, connect" should be understood in a broad sense, for example: it can be a fixed connection, a detachable connection or an integral connection; it can also be a mechanical connection, an electrical connection or a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0068] Example 1, reference Figure 1 , which is the first embodiment of the present invention, and provides a multi-level vulnerability assessment power grid security margin online monitoring method, comprising:

[0069] The present invention builds a multi-level vulnerability assessment system based on 110kV, 35kV substations and line sections, and realizes online monitoring and early warning of power grid safety margin through real-time data acquisition and deep learning models. Specifically, it includes key technologies such as multi-level power grid topology modeling, vulnerability assessment index design, deep learning model development and online monitoring system implementation.

[0070] S1: Construct a multi-level grid topology model based on 110kV and 35kV substations and line sections, and calculate the node importance and inter-level coupling.

[0071] In the power system, the modeling of the grid topology is a very complex process, which needs to consider the grid levels of different voltage levels and the interactions between them. Based on complex network theory, a multi-level grid topology model can be constructed, which can describe the interdependence and influence between different levels.

[0072] Furthermore, considering grid levels with different voltage levels and the interactions between grid levels, a multi-level grid topology model is constructed, including state space and action space: the state space is a representation of the environment, which contains all the necessary information to decide what action to take; the action space is the set of all possible actions.

[0073] The state space includes new energy output, load demand, energy storage state and voltage level, as follows:

[0074] s t =[P re (t),P l (t),E(t),V(t)]

[0075] Among them, s t is the system state at time t, P re (t) is the new energy output at time t, P l (t) is the load demand at time t, E(t) is the remaining energy storage capacity at time t, and V(t) is the voltage level at time t.

[0076] The action space includes the output of conventional units and the charging and discharging power of the energy storage system, as follows:

[0077] a t =[P g (t),P s (t)]

[0078] Among them, a t is the control action at time t, P g (t) is the output of the conventional unit at time t, P s(t) is the energy storage output at time t.

[0079] It should be noted that the node importance index is a method to evaluate the relative importance of each node in the power grid.

[0080] The node importance index is comprehensively evaluated by the node height, betweenness centrality and clustering coefficient. The multi-level power grid topology model is constructed based on complex network theory. The power grid G ​​= (V, E), where V is the node set and E is the edge set;

[0081] Define the level set L = {l1, l2, l3}, corresponding to 110kV, 35kV and key section levels respectively, and the node importance index is as follows:

[0082] I v =w1D v +w2B v +w3C v

[0083] Among them, I v is the importance index of node v, D v is the node degree, which indicates the number of edges directly connected to node v; B v is the betweenness centrality, which indicates the frequency of a node as an intermediary of the shortest path between the remaining nodes in the network; C v is the clustering coefficient, which indicates the degree of interconnection between neighboring nodes of a node; w1, w2, and w3 are weight coefficients, which are used to adjust the influence of different indicators in importance evaluation.

[0084] The inter-level coupling degree is an indicator to measure the degree of mutual dependence between power grids at different levels. The inter-level coupling degree is obtained by calculating the association strength between nodes at different levels:

[0085]

[0086] Among them, C ij is the coupling degree between levels i and j, V i 、V j are the node sets of level i and j respectively, α kl is the association strength between nodes k and l, indicating the degree of mutual influence between the two nodes.

[0087] S2: Design static safety margin and dynamic response characteristic indicators to build a comprehensive vulnerability assessment system.

[0088] Vulnerability assessment is an important aspect in power system planning and operation, which helps to identify the weak links of the system and take corresponding measures to improve the resilience of the system. An effective vulnerability assessment index system should include two aspects: static safety margin and dynamic response characteristics.

[0089] Furthermore, the comprehensive vulnerability assessment system includes static safety margin index and dynamic response characteristic index: static safety margin index is an index to measure the safety and reliability of the power system under normal operating conditions. It reflects the system's ability to withstand load changes; dynamic response characteristic index is an index to measure the power system's response and recovery capabilities when facing disturbances and faults.

[0090] The static safety margin index includes the safety margin of the evaluation system on different devices and nodes, specifically:

[0091]

[0092] Among them, S s is the static safety margin, P max is the rated capacity of the equipment, and P is the actual load.

[0093] The static safety margin index can evaluate the safety margin of the system on different devices and nodes, so as to take corresponding measures to improve the safety of the system.

[0094] The dynamic response characteristic indicators include response time, stability time and reliability, specifically:

[0095] S d =w1T resp +w2T stab +w3T rel

[0096] Among them, S d is the dynamic response characteristic index, T resp is the response time, i.e., the time required for the system to detect the disturbance and start responding; T stab is the stabilization time, that is, the time required for the system to return to a stable state after a disturbance; T rel It is a reliability indicator that reflects the ability of the system to maintain normal operation in the face of disturbances.

[0097] The static safety margin index and the dynamic response characteristic index are combined to obtain a comprehensive vulnerability index to comprehensively evaluate the vulnerability of the power system, specifically:

[0098] V=λS s +(1-λ)S d

[0099] Among them, V is the comprehensive vulnerability index and λ is the weight coefficient, which is used to balance the importance of static safety margin and dynamic response characteristics in comprehensive vulnerability assessment.

[0100] The comprehensive vulnerability index can comprehensively assess the vulnerability of the power system, help identify the weak links of the system, and take corresponding measures to improve the safety and resilience of the system.

[0101] S3: Train the LSTM deep learning model based on historical operation data.

[0102] Furthermore, the LSTM deep learning model is used to capture the long-term dependencies in the power system operation data, and update the hidden layer state, memory unit state and output respectively:

[0103] h t =σ(W h *[h t-1 ,x t ]+b h )

[0104] c t =f t ⊙c t-1 +i t ⊙g t

[0105] y t =o t ⊙tanh(c t )

[0106] Among them, h t The hidden state at time t contains all the information before time t, c t is the state of the memory cell at time t, allowing the LSTM network to remember long-term dependencies; x t The input sequence at time t contains the grid operation data at time t, W h is the weight matrix, which is used to determine how the hidden state and input data affect the new hidden state and memory unit state; b h is the bias term, providing a benchmark for updating the hidden layer state and memory unit state, f t 、i t , o t They are the forget gate, input gate, and output gate, which control the updating of the memory cell content and the generation of the output; ⊙ represents element-wise multiplication, allowing the model to consider the importance of different information when updating the memory cell state.

[0107] S4: Collect grid operation data in real time and perform standardized processing.

[0108] The online monitoring system is an important part of the power system operation. It can monitor the operating status of the power grid in real time and provide safety margin assessment and early warning.

[0109] Furthermore, the online monitoring system collects grid operation data including voltage, current, active power, and reactive power in real time, and standardizes the grid operation data to eliminate the influence of different data dimensions, so that it has zero mean and unit variance, as follows:

[0110]

[0111] Among them, x′ is the standardized data, μ is the mean, and σ is the standard deviation.

[0112] S5: Use the trained LSTM deep learning model to evaluate the current safety margin.

[0113] Furthermore, based on the trained LSTM deep learning model, the safety margin of the power grid is evaluated based on the standardized power grid operation data:

[0114] M=F LSTM (x')

[0115] Among them, M is the safety margin assessment result, F LSTM It is a trained LSTM deep learning model.

[0116] S6: Determine whether an early warning needs to be triggered based on the early warning classification mechanism, and output the evaluation results and early warning information. Furthermore, in order to respond promptly to possible safety issues in the power grid, three levels of early warning thresholds are set:

[0117] W={W1,W2,W3|W1>W2>W3}

[0118] Among them, W1, W2, and W are the first, second, and third level warnings respectively. The threshold determines the severity of the safety margin assessment results and takes corresponding warning measures.

[0119] The online monitoring system can evaluate the safety margin of the power grid in real time and provide timely warnings according to the warning classification mechanism, so as to take necessary measures to ensure the safe and stable operation of the power grid.

[0120] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

[0121] Embodiment 2 is the second embodiment of the present invention, which provides a method for online monitoring of power grid safety margin with multi-level vulnerability assessment. In order to verify the beneficial effects of the present invention, scientific demonstration is carried out through experiments.

[0122] Taking a municipal power grid system with three voltage levels (i.e. 110kV, 35kV and key sections) as an example, the multi-level topology structure of the system is analyzed.

[0123] Table 1 Example parameter settings

[0124] Parameter Type Parameter Value Voltage level 110kV、35kV Number of nodes 110kV: 15 nodes; 35kV: 25 nodes <![CDATA[Weight coefficients (w1, w2, w3)]]> 0.4、0.3、0.3 LSTM model parameters Number of hidden layer nodes: 128; time step: 24; number of training rounds: 100

[0125] Table 2 Performance index comparison

[0126] Evaluation Metrics Methods Traditional BP neural network Support Vector Machine Prediction accuracy 95.8％ 88.3％ 86.5％ Computation time (ms) 45 82 95 Error rate 2.3％ 5.7％ 6.8％ Timely warning rate 98.2％ 92.1％ 90.5％

[0127] Table 3 Algorithm advantage analysis

[0128] feature Advantages Multi-level modeling Considering the coupling relationship between different voltage levels is more in line with the actual grid structure Timing characteristics The LSTM model can effectively capture the time series characteristics of power grid operation data Computational efficiency Use optimized data processing methods to significantly improve computing efficiency Early warning mechanism The three-level early warning mechanism can more accurately reflect the system security status

[0129] Table 4 Actual application effect

[0130] Application Scenario Effect evaluation Daily monitoring Accuracy increased by 15% and response time reduced by 40% Fault warning The average advance warning time increased by 10 minutes Emergency treatment Reduce decision support time by 50%

[0131] Embodiment 3, the third embodiment of the present invention, is different from the first two embodiments in that:

[0132] If the functions are implemented in the form of software functional 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, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, and other media that can store program codes.

[0133] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in conjunction with such instruction execution systems, devices or apparatuses. For the purposes of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in conjunction with such instruction execution systems, devices or apparatuses.

[0134] More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic device), a portable computer disk case (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be a paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering or, if necessary, processing in another suitable manner, and then stored in a computer memory.

[0135] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, a plurality of steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0136] Example 3, reference Figure 2 , which is the third embodiment of the present invention, and provides a multi-level vulnerability assessment power grid safety margin online monitoring system, including a power grid topology structure building module 10, a vulnerability assessment system building module 20, an LSTM deep learning model training module 30, a data acquisition and preprocessing module 40, a safety margin assessment module 50, and an early warning classification mechanism module 60;

[0137] The power grid topology construction module 10 constructs a multi-level power grid topology model based on 110kV, 35kV substations and line sections, and calculates the node importance and the coupling degree between levels, providing basic data of the power grid structure and the relationship between levels for subsequent vulnerability assessment;

[0138] Vulnerability assessment system construction module 20 designs static safety margin and dynamic response characteristic indicators, builds a comprehensive vulnerability assessment system, and provides training objectives and evaluation criteria for the LSTM deep learning model;

[0139] The LSTM deep learning model training module 30 trains the LSTM deep learning model based on the historical operation data, captures the long-term dependencies in the power grid operation data, and outputs the trained model for use by the early warning classification mechanism module 60;

[0140] The data acquisition and preprocessing module 40 collects power grid operation data in real time and performs standardized processing to provide real-time data input for the LSTM model;

[0141] The safety margin evaluation module 50 receives the preprocessed data, uses the trained LSTM model to evaluate the safety margin of the current power grid, and outputs the safety margin evaluation result;

[0142] The early warning classification mechanism module 60 receives the safety margin assessment result, determines whether to trigger an early warning according to the early warning classification threshold, outputs the assessment result and early warning information, and executes early warning measures.

[0143] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A multi-level vulnerability assessment online monitoring method for power grid security margin, characterized by: include, A multi-level grid topology model is constructed based on 110kV and 35kV substations and line sections, and node importance and inter-level coupling are calculated; Design static safety margin and dynamic response characteristic indicators, and build a comprehensive vulnerability assessment system; Train LSTM deep learning model based on historical operation data; Collect power grid operation data in real time and perform standardized processing; Use the trained LSTM deep learning model to evaluate the current safety margin; Determine whether an early warning needs to be triggered based on the early warning classification mechanism, and output the evaluation results and early warning information.

2. The online monitoring method for power grid safety margin of multi-level vulnerability assessment according to claim 1, characterized in that: The constructing of the multi-level grid topology structure model includes considering grid levels of different voltage levels and the interaction between grid levels, and constructing the multi-level grid topology structure model includes a state space and an action space; The state space includes new energy output, load demand, energy storage state and voltage level, as follows: s t =[P re (t),P l (t),E(t),V(t)] Among them, s t is the system state at time t, P re (t) is the new energy output at time t, P l (t) is the load demand at time t, E(t) is the remaining energy storage capacity at time t, and V(t) is the voltage level at time t; The action space includes the output of conventional units and the charging and discharging power of the energy storage system, as follows: a t =[P g (t),P s (t)] Among them, a t is the control action at time t, P g (t) is the output of the conventional unit at time t, P s (t) is the energy storage output at time t.

3. The online monitoring method for power grid safety margin of multi-level vulnerability assessment according to claim 2, characterized in that: The calculation of node importance and inter-level coupling includes comprehensively evaluating the node importance index through the height, betweenness centrality and clustering coefficient of the node, and setting the power grid G=(V, E) based on the constructed multi-level power grid topology model, where V is a node set and E is an edge set; Define the level set L = {l1, l2, l3}, corresponding to 110kV, 35kV and key section levels respectively, and the node importance index is as follows: I v =w1D v +w2B v +w3C v Among them, I v is the importance index of node v, D v is the node degree, which indicates the number of edges directly connected to node v; B v is the betweenness centrality, which indicates the frequency of a node as an intermediary of the shortest path between the remaining nodes in the network; C v is the clustering coefficient, which indicates the degree of interconnection between neighboring nodes of a node; w1, w2, and w3 are weight coefficients, which are used to adjust the influence of different indicators in importance evaluation; The inter-level coupling degree is obtained by calculating the association strength between nodes at different levels: Among them, C ij is the coupling degree between levels i and j, V i 、V j are the node sets of level i and j respectively, α kl is the association strength between nodes k and l, indicating the degree of mutual influence between the two nodes.

4. The method for online monitoring of power grid safety margin for multi-level vulnerability assessment according to claim 3, characterized in that: The comprehensive vulnerability assessment system includes static safety margin indicators and dynamic response characteristic indicators; The static safety margin index includes the safety margin of the evaluation system on different devices and nodes, specifically: Among them, S s is the static safety margin, P max is the rated capacity of the equipment, and P is the actual load; The dynamic response characteristic indicators include response time, stability time and reliability, specifically: <h2 style=";text-align:left;direction:ltr">S<h2 style=";text-align:left;direction:ltr"> d <h2 style=";text-align:left;direction:ltr"> =w1T<h2 style=";text-align:left;direction:ltr"> resp <h2 style=";text-align:left;direction:ltr"> +w2T<h2 style=";text-align:left;direction:ltr"> stab <h2 style=";text-align:left;direction:ltr"> +w3T<h2 style=";text-align:left;direction:ltr"> rel Among them, S d is the dynamic response characteristic index, T resp is the response time, i.e., the time required for the system to detect the disturbance and start responding; T stab is the stabilization time, that is, the time required for the system to return to a stable state after a disturbance; T rel It is a reliability index that reflects the ability of the system to maintain normal operation in the face of disturbances; The static safety margin index and the dynamic response characteristic index are combined to obtain a comprehensive vulnerability index to comprehensively evaluate the vulnerability of the power system, specifically: V=λS s +(1-λ)S d Among them, V is the comprehensive vulnerability index and λ is the weight coefficient, which is used to balance the importance of static safety margin and dynamic response characteristics in comprehensive vulnerability assessment.

5. The method for online monitoring of power grid safety margin with multi-level vulnerability assessment according to claim 4, characterized in that: The LSTM deep learning model includes capturing the long-term dependencies in the power system operation data through the LSTM deep learning model, and updating the hidden layer state, memory unit state and output respectively: h t =σ(W h *[h t-1 ,x t ]+b h ) c t =f t ⊙c t-1 +i t ⊙g t y t =o t ⊙tanh(c t ) Among them, h t The hidden state at time t contains all the information before time t, c t is the state of the memory cell at time t, allowing the LSTM network to remember long-term dependencies; x t The input sequence at time t contains the grid operation data at time t, W h is the weight matrix, which is used to determine how the hidden state and input data affect the new hidden state and memory unit state; b h is the bias term, providing a benchmark for updating the hidden layer state and memory unit state, f t 、i t , o t They are the forget gate, input gate, and output gate, which control the updating of the memory cell content and the generation of the output; ⊙ represents element-wise multiplication, allowing the model to consider the importance of different information when updating the memory cell state.

6. The method for online monitoring of power grid safety margin for multi-level vulnerability assessment according to claim 5, characterized in that: The evaluation of the current safety margin includes collecting grid operation data including voltage, current, active power, and reactive power in real time through an online monitoring system, and performing standardized processing on the grid operation data: Among them, x′ is the standardized data, μ is the mean, and σ is the standard deviation; Based on the trained LSTM deep learning model, the safety margin of the power grid is evaluated based on the standardized power grid operation data: M=F LSTM (x') Among them, M is the safety margin assessment result, F LSTM It is a trained LSTM deep learning model.

7. The method for online monitoring of power grid safety margin for multi-level vulnerability assessment according to claim 6, characterized in that: The warning classification mechanism includes setting three levels of warning thresholds: W={W1,W2,W3|W1>W2>W3} Among them, W1, W2, and W are the first, second, and third level warnings respectively. The threshold determines the severity of the safety margin assessment results and takes corresponding warning measures.

8. A system for online monitoring of power grid safety margin using the multi-level vulnerability assessment method according to any one of claims 1 to 7, characterized in that: It includes power grid topology construction module, vulnerability assessment system construction module, LSTM deep learning model training module, data acquisition and preprocessing module, safety margin assessment module, and early warning classification mechanism module; The grid topology construction module constructs a multi-level grid topology model based on 110kV, 35kV substations and line sections, and calculates node importance and inter-level coupling, providing basic grid structure data and inter-level relationships for subsequent vulnerability assessment; The vulnerability assessment system construction module designs static safety margins and dynamic response characteristic indicators, builds a comprehensive vulnerability assessment system, and provides training objectives and evaluation criteria for the LSTM deep learning model; The LSTM deep learning model training module trains the LSTM deep learning model based on historical operation data, captures the long-term dependencies in the power grid operation data, and outputs the trained model for use by the early warning classification mechanism module; The data acquisition and preprocessing module collects power grid operation data in real time and performs standardized processing to provide real-time data input for the LSTM model; The safety margin evaluation module receives the preprocessed data, uses the trained LSTM model to evaluate the safety margin of the current power grid, and outputs the safety margin evaluation result; The warning classification mechanism module receives the safety margin assessment result, determines whether to trigger a warning according to the warning classification threshold, outputs the assessment result and warning information, and executes warning measures.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method for online monitoring of power grid security margin for multi-level vulnerability assessment according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method for online monitoring of power grid security margin for multi-level vulnerability assessment according to any one of claims 1 to 7 are implemented.

Images