Power grid information safety integrated operation monitoring method

Abstract

The invention relates to the technical field of electric power, in particular to a power grid information safety integrated operation monitoring method, which comprises the following steps of: acquiring electrical parameter information of power distribution equipment in power grid information, forming a power grid topological structure by taking the power distribution equipment as a power grid node, and selecting a current interaction area; performing double-layer mapping on each power grid node, acquiring an abnormal event of each power grid node from an uploading period of the power grid information within a preset time period, and defining a time interaction form of the power grid information according to an interaction frequency and an interaction type of the abnormal event; constructing a node change matrix; utilizing a conduction path of the node change matrix under chain propagation to obtain an interactive update sequence; obtaining an interaction target of each power grid node from the interaction update sequence, and determining a current compensation strategy according to the interaction target of each power grid node; performing interactive evaluation on each power grid node included in the current compensation strategy, and adjusting a switch corresponding to each power grid node; and the stability and efficiency of power grid operation are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of electric power, in particular to a power grid information security integrated operation monitoring method. BACKGROUND

[0002] Traditional distribution network dynamic adjustment is a process of modifying the distribution network to meet the optimization of one or more aspects based on the analysis of various aspects of the existing distribution network. For example, it can include: dynamic load, load forecasting, system network loss, voltage level, power supply reliability, etc. In the case of distribution network operation monitoring, the lack of correlation between the electrical parameters of the distribution equipment and the physical topology leads to insufficient positioning accuracy of abnormal situations, forming a data island; at the same time, it is necessary to track the dynamic changes of each electrical parameter to prevent overcompensation or undercompensation.

[0003] For example, Chinese Patent Publication No. CN115663807A discloses a strong distribution network method based on predicted load and distributed power output, relating to the technical field of electric power. It includes: when the load value of any node in the current power grid network structure exceeds the preset load value, the load of the future time period of the power grid is predicted based on historical data to obtain the load data of all nodes in the future time period; the load clustering algorithm is used to cluster the load data of all nodes in the current time period and the load data of all nodes in the future time period to obtain the load data of k cluster centers; a dynamic structure model of the power grid is constructed according to the load data of the k cluster centers; the dynamic structure model is solved using a search algorithm to obtain the optimal solution corresponding to each cluster center and adjust the current power grid network structure according to the optimal solution.

[0004] For example, Chinese Patent Publication No. CN117639042A discloses a loss reduction power supply control method and device for a medium and low voltage distribution network, electronic equipment and medium. A specific embodiment of the method includes: obtaining initial distribution network power value, distribution network output device information and distribution network energy storage device information; generating a random power supply power planning information set; based on the initial distribution network power value and the distribution network energy storage device information, each random power supply power planning information in the random power supply power planning information set is initialized to generate initial power supply power planning information, and an initial power supply power planning information set is obtained; based on the distribution network output device information, the initial power supply power planning information set is planned to obtain target power supply power planning information; based on the target power supply power planning information, the associated distribution network energy storage system is controlled to supply power.

[0005] The prior art illustrates the way of identifying the predicted load output and the iteration number of power grid planning and scheduling by predicting the clustering between the loads, but in the power grid topology structure, if there is a corresponding anomaly on the power grid node corresponding to the topology structure, it is difficult to identify the working condition of the power distribution equipment represented by each power grid node through the load clustering and power planning of the power grid topology structure, so that when the power grid scheduling is abnormal, the abnormal condition of the power grid node connected to the related power distribution equipment cannot be found in time, which reduces the positioning accuracy when the anomaly is located, and causes loss of power scheduling. SUMMARY

[0006] In order to solve the above technical problems, the technical scheme adopted by the present application is: a power grid information security integrated operation monitoring method, comprising: S1, obtaining the electric parameter information of the power distribution equipment in the power grid information, according to the connection relationship of the power distribution equipment, the power distribution equipment is taken as the power grid node to form the power grid topology structure, and the distance between each power grid node and the mobilization area on the topology structure when the current interacts, and the current interaction area is selected.

[0007] S2, double-layer mapping is performed on the physical level and the information level of each power grid node in the current interaction area, the electric parameter information of each power grid node in the current interaction area is feature extracted according to the upload period of the power grid information in a preset time period, the abnormal events of each power grid node are obtained, and the time interaction form of the power grid information is defined according to the interaction frequency and the interaction type of the abnormal events.

[0008] S3, based on the time interaction form of the power grid information, a node change matrix is constructed, and a conduction path under chain propagation is used to obtain an interaction update sequence of the current interaction area.

[0009] S4, the interaction target of each power grid node is obtained from the interaction update sequence, and the current compensation strategy under different current states is determined according to the interaction target of each power grid node.

[0010] S5, each power grid node contained in the current compensation strategy is interactively evaluated, and the switch corresponding to each power grid node is adjusted.

[0011] The present application has the advantages that: firstly, the present application constructs the power grid topology structure based on the connection relationship of the power distribution equipment, integrates the data by combining the value of the electric parameter information on the power grid topology structure, identifies the current interaction area when the current loss of each power grid node is the smallest, determines the running condition of the power distribution equipment represented by each power grid node during power scheduling, and perceives the global state of the overall power grid, so that the abnormal condition of each power grid node can be identified in time when the running state of each power grid node is viewed.

[0012] 2. The present invention extracts various abnormal events when abnormal grid scheduling occurs by associating the topological structure of grid nodes with the physical location of equipment, and maps these abnormal events with the topological structure and physical location of grid nodes. It can timely discover the relative number of abnormal occurrences in different data upload cycles during power dispatch, and combine these numbers with the locations of corresponding grid nodes to identify abnormal grid nodes existing during abnormal detection. At the same time, the threshold of dynamic baseline reference data is generated based on historical data to realize dynamic update of abnormal characteristics, thereby improving the timeliness and accuracy of abnormal detection.

[0013] 3. The present invention uses the interaction frequency, duration and power gradient of each grid node to identify the relative changes of each node, quantify the abnormal propagation path of the grid node when the power scheduling is abnormal, and provide a basis for accurate compensation; and performs current compensation on these grid nodes with abnormal propagation, locates the compensation target based on the minimum value of the current branch interaction type, locates the target value by combining the average value and limit value of the current, voltage and power, and then performs power compensation on the area that needs power compensation to achieve dynamic adaptation and economic optimization of the compensation strategy.

[0014] 4. The present invention combines the interaction frequency and the operating cost to screen the target area, and then closes the switch of the target area and disconnects the compensation device of the non-target area according to the principle of maximum similarity; to control the dispatching mode of each grid node in the power system, to prevent losses caused by excessive or insufficient power dispatching, and to improve the processing speed of the overall system under abnormal power dispatching. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] The present invention will be further described below with reference to the accompanying drawings and examples.

[0016] Figure 1 The present invention is a flow chart of an integrated operation monitoring method for power grid information security.

[0017] Figure 2 The present invention is a flow chart of step S1 of a method for integrated operation monitoring of power grid information security.

[0018] Figure 3 The present invention is a flow chart of step S2 of a method for monitoring the integrated operation of power grid information security.

[0019] Figure 4 The present invention is a flowchart of step S3 of the method for integrated operation monitoring of power grid information security.

[0020] Figure 5 The present invention is a flow chart of step S4 of the method for integrated operation monitoring of power grid information security.

[0021] Figure 6is a flowchart of step S5 of the power grid information security integrated operation monitoring method. DETAILED DESCRIPTION

[0022] Embodiments of the present application are described in detail below. The embodiments described below are exemplary and are intended to explain the present application, and are not to be understood as limiting the present application. If a specific technique or condition is not mentioned in the embodiments, the technique or condition described in the literature in the art or according to the product manual is used.

[0023] Reference Figure 1 A power grid information security integrated operation monitoring method comprises the following steps: S1, acquiring electric parameter information of power distribution equipment in power grid information, and according to the connection relationship of the power distribution equipment, the power distribution equipment is taken as a power grid node to form a power grid topology structure, and a current interaction area is selected according to the distance between each power grid node in the topology structure and the mobilization area during current interaction.

[0024] S2, performing double-layer mapping on the physical level and the information level of each power grid node in the current interaction area, performing feature extraction on the electric parameter information of each power grid node in the current interaction area according to the uploading period of the power grid information in a preset time period, acquiring abnormal events of each power grid node, and defining the time interaction form of the power grid information according to the interaction frequency and the interaction type of the abnormal events.

[0025] S3, constructing a node change matrix based on the time interaction form of the power grid information, and obtaining an interaction update sequence of the current interaction area by using the conduction path of the node change matrix under chain propagation.

[0026] S4, acquiring an interaction target of each power grid node from the interaction update sequence, and determining a current compensation strategy under different current states according to the interaction target of each power grid node.

[0027] S5, performing interaction evaluation on each power grid node contained in the current compensation strategy, and adjusting the corresponding switch of each power grid node.

[0028] The current interaction area represents the process of dynamically adjusting the current distribution between areas according to the change of the power grid load, and the current value and the position adjusted at this time in the position of the power grid topology structure are set as the current interaction area; the power is used to describe the power size during the current interaction, and is used to describe the current change during the current interaction process; the mobilization area is a relative position for quickly scheduling the current to restore power supply or support other areas in an emergency, and this area is the area that needs to be mobilized during the current scheduling and interaction.

[0029] Preferably, the electric parameter information comprises current, voltage and power.

[0030] In step S1, the current interaction form is described mainly for the topology and current values between the devices existing in the power grid information, and whether there are various abnormalities or how the current interaction is trained is determined.

[0031] Preferably, the power information includes the current, power, voltage, device position, communication flow data, etc. of the power distribution devices, the device position of the power distribution device at this time is regarded as each power grid node in the power grid topology to complete the extraction of the current and power, and the mobilization area indicates the position of each power distribution device when the power is allocated, the area indicates the required current and power, and the node that needs to be dispatched, and these contents are summarized as the current interaction area.

[0032] As shown in Figure 2 The implementation mode of step S1 includes: S11, constructing at least one hierarchical connection relationship of the power distribution device according to the position of the power distribution device in the open state of the power distribution device, the hierarchical connection relationship indicates that the power distribution devices are located in any one of the station control management layer, the network communication layer and the field device layer, and is used to indicate the position of the power distribution device in the physical structure and the information space.

[0033] S12, generating the power grid topology of the power distribution device by using the hierarchical connection relationship of the power distribution device, and integrating the data of each power grid node according to the electrical parameter information collected on the power grid topology.

[0034] S13, monitoring the electrical parameter information of each power grid node, and dividing the mobilization area according to the fluctuation proportion of the electrical parameter information of each power grid node and the distance between each power grid node.

[0035] S14, setting the area radius as the center of the scheduling area, and outputting each power grid node included in the area radius as the current interaction area when the current loss of each power grid node included in the area radius is the minimum.

[0036] At this time, the electrical parameter information collected on the multiple power distribution devices is integrated, and then the distance relationship between the normally running nodes and the pre-defined scheduling area is integrated to find out which power grid position corresponding to the power distribution device is selected to perform power scheduling when there is a part of the area that needs current scheduling, so as to maintain or restore the normal operation of other areas.

[0037] Preferably, each power grid node of the power grid topology represents a variety of power distribution devices, such as power stations, substations, power distribution cabinets, power users, etc. The power user represents a unit that consumes power, such as a residence, a factory, etc., which is the final load point of the power system.

[0038] Preferably, the implementation of dividing the scheduling area is that the distance of each grid node in the power grid topology is calculated using the admittance matrix, the ratio of the average power difference of each grid node at consecutive time points to the sum of the average power differences of all grid nodes is used as the fluctuation proportion of the electrical parameter information, and the ratio is taken as the weight of each grid node; and each grid node is clustered according to the weight and distance of each grid node, and the largest cluster after clustering is taken as the output scheduling area.

[0039] Generally, the power distribution devices selected as the scheduling area need to be close in relative distance and have small changes in electrical parameters over time, at which time the average power difference of the power at consecutive time points is selected to obtain multiple nodes with stable power, and then clustering is performed to obtain multiple power distribution devices with stable power and similar distances, and the largest cluster after clustering is taken as the scheduling area for subsequent use, which can maintain the safe operation of the current power grid information while maintaining the maximum schedulable power. At this time, the clustering method can use the weighted K-means clustering algorithm to obtain the output scheduling area.

[0040] Preferably, the implementation of selecting the current interaction area further includes: determining the connection path of each grid node and the switch at the adjacent position according to the number of switches connected to each grid node, and screening each grid node in the current interaction area according to the connection path; and making each grid node in the current interaction area correspond to at least one switch to clearly define its topology, thereby indicating the direction of the current flow when power is exchanged.

[0041] Preferably, the radius of the scheduling area is based on the output of the scheduling area, the radius of the cluster corresponding to the scheduling area is the initial radius, and the current loss of the grid node in the scheduling area is measured, which is obtained by multiplying the square value of the current by the line resistance of the grid node; then the radius of the area is continuously adjusted, the adjustment value is 10% of the initial radius, and then iteration is continuously performed, when the average value of the current loss of the grid node in the scheduling area relative to the number of grid nodes is the smallest, the current loss of each grid node is considered to be the smallest; and the grid nodes included in the radius of the area at this time and the position of the grid node in the power grid topology are output as the current interaction area.

[0042] In an embodiment of the present application, in step S2, whether the current power distribution device is normally running in the power grid topology is identified mainly according to the mapping relationship of each grid node selected in the current interaction area in the device entity and the information space, and the frequency of current exchange and the coordination state of the exchange are determined according to the normal running form, to prevent the problem of scheduling abnormality in current scheduling.

[0043] Preferably, the identification content of the physical level is the time sequence anomaly pattern identification of each power grid node, and the identification of the information level is the identification of the parameter conduction anomaly of each power grid node under the guidance of information transmission.

[0044] As shown in Figure 3 The implementation mode of step S2 includes: S21, checking the power grid topology of each power grid node in the current interaction area, associating the power grid topology with the physical position to obtain the mapping relationship of the physical level, and introducing the mapping relationship of the physical level into the uploading period of each power grid node to obtain the mapping relationship of the information level.

[0045] S22, checking the electric parameter information in the uploading period, identifying the characteristic type of the abnormal event corresponding to the electric parameter information, and the characteristic type includes: continuous anomaly, transient anomaly, periodic anomaly and correlation anomaly of the electric parameter information; the continuous anomaly is represented by the current effective value and the voltage harmonic distortion rate; the continuous time is used to indicate whether the values of the voltage, current and other parameters appear continuously beyond the limit, for example, the part of the voltage and current beyond the average value plus three times the standard deviation in the set unit time is identified, and the length of the unit time can be selected as 10s, 30s, 60s and the like according to the demand, or the unit time can be linked with the data uploading period, and the time stamp corresponding to one tenth of the data in the data uploading period is used for setting.

[0046] The transient anomaly is represented by the voltage sag depth and the current impact peak value, which indicates the instantaneous drop of data, that is, the sudden decrease of the electric parameter information at a certain time point, and the occurrence of this situation is identified as a transient anomaly.

[0047] The periodic anomaly is represented by the power fluctuation frequency spectrum, which indicates whether the current and voltage increase and decrease in a certain period, and the existence of the periodic anomaly is regarded as a periodic anomaly, and the frequency of the anomaly in the preset uploading period in a certain period is identified to indicate the duration of the periodic change.

[0048] The correlation anomaly is represented by the power imbalance degree of multiple nodes, which indicates the case that the power difference between multiple nodes is too large, at this time, the correlation anomaly is mainly used to indicate whether the uploaded data exists parameter conduction anomaly between nodes, and the difference between the nodes. As shown in Table 1, the four characteristic types can be identified by using the corresponding algorithm to determine whether the interaction is abnormal.

[0049] Table 1. Characteristic type of abnormal event

[0050]

[0051] In Table 1, the main identified parameters under multiple feature types are illustrated, and the types of the main identified parameters and the data corresponding to the parameters are illustrated. Then, the power grid information is monitored in the process of power dispatching to prevent dispatching abnormalities or repeated operation of individual power distribution equipment, so that the finally dispatched current cannot meet the demand of power dispatching.

[0052] In S23, the number of occurrences of the feature type of the abnormal event in the preset time period is taken as the interaction frequency, and the node associated with the feature type of the abnormal event is taken as the interaction type. The data corresponding to the interaction frequency and the interaction type are output as time interaction forms.

[0053] The output time interaction form will summarize the data under the same feature type in the form of time sequence according to the power distribution equipment represented by the corresponding power grid node, identify the abnormal position existing in the current power interaction area and the power grid topology, take the power grid topology as the starting point, summarize the abnormal features, identify the main change parameters and specific abnormal patterns, then summarize the occurrence of the interaction type and the interaction frequency, identify the conditional probability of each node under the corresponding abnormal event, identify the conditional probability according to the normal conduction order of the current data, and find out the single device or multiple devices affecting the abnormal power dispatching.

[0054] Preferably, the time interaction form of the power grid information further comprises: outputting the nodes associated with the interaction type according to the conduction order of the abnormal event between the nodes of the power grid, forming dynamic baseline reference data; the dynamic baseline reference data is used to identify abnormal behaviors deviating from the normal mode, and to illustrate the conduction path of the abnormal power dispatching, and according to the path, how to adjust subsequently to reduce the loss of power dispatching is illustrated.

[0055] The parameter threshold of the dynamic baseline reference data is calculated, the parameter threshold is updated according to the uploading period of the power grid information, and the feature type of each abnormal event is extracted using the dynamic baseline reference data; the updated dynamic baseline reference data each time is taken as the output time interaction form.

[0056] The dynamic baseline reference data explained in the foregoing content is updated according to the abnormal events identified in the period in which the input data is located each time the update is performed, and if the upper limit and the lower limit of the value represented by the abnormal events increase, the update is performed, and then the increased part is output to find the data obviously abnormal in the power grid information; meanwhile, the dynamic baseline reference data is obtained according to the conduction sequence, which is used to identify how the abnormality of which device is connected with other devices under the communication and current interaction between the power distribution devices corresponding to each power grid node under the normal operation of the power distribution device, so as to facilitate the extraction of the device that appears abnormal and the device affected by the abnormal device, and then according to the device, the main abnormal problems existing in the current power dispatching can be found, and after the abnormal problems and the power parameter conduction are associated, it is more convenient to control the power distribution devices in multiple regions and groups, and the power grid topology is adjusted in time.

[0057] Preferably, the parameter threshold of the dynamic baseline reference data can be set by the average value of the upper limit value and the lower limit value of the corresponding electric parameter information of the dynamic baseline reference data, and when the parameter threshold is updated, the data of the nodes associated with the interaction type obtained in the above period is updated, and the data corresponding to the value change each time is output, and the time interaction form of the output includes the interaction type, the feature type, the interaction frequency, and the value of the corresponding data of the electric parameter, to determine the data that obviously changes, to determine whether the current power grid dispatching is normally operated.

[0058] In an embodiment of the present application, the node change matrix records the nodes that have state changes at adjacent time points, and this part of data is regarded as the node change matrix; for example, when the node is converted from the normal working state to the power dispatching state under the power grid dispatching instruction, or from the normal working state to the abnormal working state, the nodes that have state changes at this time are extracted, and the conduction path under the chain propagation is identified, to identify how the state of the current node propagates, and how the current, voltage and power and other parameters change after the propagation, to explain how the current circuit controls the adjustment mode of each power distribution device under the reception of the power grid dispatching instruction.

[0059] As shown in Figure 4 The implementation mode of step S3 further includes: S31, calling the power grid dispatching instruction in the power grid information, defining a state propagation set according to the power grid dispatching instruction, and generating the state change probability of each power grid node based on the power grid topology.

[0060] S32, weighting the state change probability according to the conduction path of each power grid node to form a node change matrix.

[0061] S33, identifying the interaction frequency, interaction duration and power change gradient of the node change matrix to form an interaction update sequence.

[0062] Preferably, the power grid scheduling instruction includes, but is not limited to, the control instruction issued by the energy management system of the current power distribution device, such as AGC adjustment, circuit breaker operation; the scheduling log file contains instruction type, target device, execution time, and real-time control signal, such as GOOSE message, MMS service call, etc. content, to explain the specific content of the current power distribution device scheduling.

[0063] Preferably, the state propagation set is defined according to the propagation type of each power grid node, such as the propagation type, trigger condition and propagation characteristics of each power distribution device in the power grid scheduling instruction, to describe the working condition of the power distribution device represented by each power grid node in the power grid. The content of the propagation type will be labeled in advance according to the type of the power distribution device, at this time, it is to verify whether there is a delay or other form of abnormal problem when the corresponding power distribution device is scheduled, and to quantify these problems specifically, as shown in Table 2.

[0064] Table 2. Propagation type classification

[0065]

[0066] Table 2 explains the data content that each power grid node in the state propagation set can represent, and obtains the probability of each power grid node under the corresponding state propagation according to the state propagation set and the power grid topology; whether the corresponding power distribution device of each power grid node triggers the condition to set whether there is a propagation type, and extracts the data under the corresponding characteristics according to the propagation characteristics.

[0067] For example, the state propagation set is mathematically displayed as ; wherein represents the power grid node set, represents the i-th node; represents the time delay of the state propagation to the node i, represents the time delay calculation function, which is calculated from any node j as the starting point to the time delay of the node to the current node i, mainly describes whether the state propagation is in a normal state, to find out whether there is a transmission delay node, if there is, it means that the current, voltage and other electrical parameter information uploaded by the node may have real-time problems; represents the line impedance between nodes i and j, the value range of i and j is based on the number of data in the power grid node set, such as the value range of i and j is 1 to N, and N represents the number of data in the power grid node set; represents the power flow between nodes, with the unit of MW, mainly represents how the power is conducted between nodes, whether a node is input and a node is output, to identify whether the power flow of the power scheduling meets the preset demand; represents the weight of each grid node, reflecting its importance in state propagation.

[0068] For ; wherein, represents the shortest number of hops from node i to j in the topology path, that is, the number of line sections passed through, which indicates how many connected edges need to be passed through from node i to node j; 、 represents a weighting coefficient, usually taking the values of 0.7 and 0.3. At this time, through the form of state propagation, it is identified that there are grid nodes whose current, voltage and power exceed a certain threshold, and the connection of each grid node under the grid topology structure is described in the form of state propagation.

[0069] Then set the state change probability of each grid node in the state propagation , such as ; wherein, represents the degree of node i, that is, the number of edges connected to other nodes of the current grid node i, which is determined based on the connection relationship between each node and other nodes in the grid topology structure, to represent the propagation of each grid node relative to other propagation; represents the attenuation coefficient, which is selected as 0.1, represents an exponential constant.

[0070] At this time, the propagation of each grid node can be obtained, and then these grid nodes are identified according to the defined conduction path, the propagation between the grid nodes on the conduction path is weighted, and the weighted data is used as the node change matrix.

[0071] Preferably, the current interaction intensity between each grid node on the conduction path is used as the path weighting method, and the current interaction intensity is set according to the average current between each grid node, and the normalized current interaction intensity is used as the weighting method of each grid node on the conduction path.

[0072] Preferably, when constructing the node change matrix, the data collected is extracted from the sliding time window, and the interaction frequency, interaction duration and power change gradient in each sliding time window are recorded. At this time, the identified content is identified based on the current state change to determine how the current power distribution equipment should be turned on and off to complete the corresponding power dispatching. The power change gradient represents the power difference of a fixed time interval divided by the value of the time interval within the interaction duration, which is used to represent the change of power in the time interval.

[0073] For example, when forming an interaction update sequence, its implementation method also includes: identifying the input end corresponding to each grid node; if the current grid node is the input end, limiting the power supply of the current grid node according to the average value of the power change gradient under the corresponding interaction frequency; and using the interaction frequency, interaction duration and power change gradient corresponding to the current grid node as the output interaction update sequence.

[0074] If the current grid node is the output end, the node change matrix is ​​screened according to the interaction frequency and interaction duration of the current grid node, and the data with a power change gradient greater than the average power change gradient is used as the output interaction update sequence.

[0075] If the current grid node is an intermediate node, the interaction frequency and power gradient where both exceed the average are taken as the output interaction update sequence. The average value here refers to the average value of the interaction frequency extracted within multiple sliding time windows, while the power gradient is the value greater than the average power gradient. These values ​​with obvious frequent changes are used as the output interaction update sequence.

[0076] In the above description, the grid node at the input end dynamically adjusts the power supply at the input end based on the average value of the power change gradient under the interaction frequency; this is mainly to prevent grid instability: by limiting the power supply, it avoids grid frequency or voltage fluctuations caused by excessive power change gradients, such as the volatility of renewable energy.

[0077] The grid nodes at the output end screen the node change matrix based on the interaction frequency and duration, and select data with a power change gradient greater than the average value. By screening high-frequency, long-duration interaction events, the power distribution at the output end is optimized, invalid transmission is reduced, and priority is given to loads with large power change gradients, thereby improving the economy of the grid. At the same time, in the event of corresponding abnormalities, the abnormal data is quickly identified to ensure the operation of the grid.

[0078] The grid nodes at the intermediate nodes quickly respond to load changes or faults in the grid by selecting high-frequency, high-gradient data; they select the corresponding data of each grid node during transmission to verify whether its transmission loss is normal, which facilitates the subsequent cost verification of the grid scheduling and improves the overall efficiency of the grid scheduling.

[0079] In one embodiment of the present invention, the interaction target of each grid node represents the grid nodes selected in pairs when identifying the relative propagation between grid nodes in the interaction update sequence, and the structure corresponding to the corresponding grid node in the grid topology structure is regarded as its mapping structure interval to identify the area where multiple distribution equipment that interact during current interaction are located, and then query the generated operating cost to quantify the loss under each current interaction mobilization.

[0080] likeFigure 5 As shown, step S4 further comprises: S41, querying the interaction types of each interaction target with the current branch of each interaction target at the time of current interaction, finding the minimum value of each interaction type, and tracking the interaction target with the minimum value of each interaction type; wherein the current branch represents a path in the circuit through which the same current flows, and contains elements such as resistors and capacitors, and is divided into active branches containing power sources and passive branches not containing power sources.

[0081] S42, for any interaction type in which the interaction target is located, identifying the current, voltage and power of the interaction target in each interaction type, and fusing the average value and the limit value of the current, voltage and power to obtain the target value of each interaction target.

[0082] S43, setting a current compensation strategy according to the target value of each interaction target and the position of the interaction target on the current branch.

[0083] Preferably, the interaction type indicates the label set for the data output in the processing in steps S3 and S2 corresponding to the current, voltage and power in the interaction update sequence; then each interaction type is tracked for each power grid node represented by each interaction target, the minimum value of each interaction type is identified, and the minimum value is used to identify the cost generated by the interaction of the current at the minimum value. This cost is convenient for predicting the total lower limit of the current interaction cost, and the purpose is to optimize the circuit performance and reduce the cost; that is, the current of the current power grid node is set to adjust the compensation of the current under the condition that the overall cost of the current power grid node is represented, to reduce the loss generated by the abnormal propagation of the current, voltage and other contents, and to reduce the consumption cost of abnormal scheduling.

[0084] Preferably, the fusion of the average value and the limit value of the current, voltage and power can be obtained by weighted averaging, such as setting the standard value and the limit value with weights of 0.6 and 0.4, or calculating the ratio of the average value and the limit value in the historical data according to the current average value and the limit value to obtain the target value expressed after weighting; then according to the target value of each interaction target and the position of the interaction target on the current branch, the current compensation strategy is retrieved from the database, when the target value and the position of the interaction target on the current branch reach the maximum similarity with the historical data, the corresponding current compensation strategy is output; wherein, the target value of the interaction target is calculated using cosine similarity, and the position of the interaction target on the current branch is calculated using the distance value of the position to calculate the similarity, and finally the current compensation strategy corresponding to the maximum value of the two similarities is selected to adjust the output of the current power grid node, so that the power distribution equipment can be operated in a relatively stable manner under abnormal conditions.

[0085] Preferably, the limit values of current, voltage and power represent the maximum and minimum values existing in the current uploading cycle; the average values of current, voltage and power also refer to the average values calculated in the current uploading cycle.

[0086] Preferably, the current compensation strategy represents the values of current, voltage and power of each power grid node operating under the power grid topology of the current power grid node, and the corresponding logical position of each power grid node, so as to facilitate subsequent identification of the position of the interaction target on the current branch and whether it is close to the power grid node corresponding to the adjustment of current, voltage and power required by the current compensation strategy, so as to facilitate the description of the control mode of the power grid node contained in the interaction target.

[0087] In an embodiment of the present application, as shown in Figure 6 the implementation of step S5 further includes: S51, obtaining the coordinates of each power grid node contained in the current compensation strategy, generating a current compensation area, and detecting the coincidence degree of the current compensation area and the current interaction area.

[0088] S52, judging whether the coincidence degree is less than the coincidence degree threshold value, if less than the coincidence degree threshold value, re-obtaining the current compensation area, otherwise, judging whether the current compensation area is the target area according to the interaction frequency and operation cost of the current compensation area.

[0089] S53, if it is the target area, recording the interaction frequency of the target area, updating the target area with the uploading cycle corresponding to the interaction frequency, and closing the switches of each power grid node in the target area when the similarity value of the target area is maximum in multiple uploading cycles; if it is not the target area, then disconnecting the switches of each power grid node in the current compensation area.

[0090] Preferably, the coincidence degree described above is to compare whether the power grid nodes in the current compensation area and the current interaction area coincide to obtain the coincidence degree, that is, the ratio of the intersection to the union of the current compensation area and the current interaction area; by generating the current compensation area and analyzing the coincidence degree with the interaction area, it is ensured that the compensation measures accurately cover the power grid area that actually needs to be intervened, and resource waste is avoided; the coincidence degree threshold value required here can be 90%, or the average value of the coincidence degree calculated by executing the current compensation strategy in historical data can be used as the coincidence degree threshold value at this time. That is, when the current interaction area appears abnormally, the compensation area is quickly located through the coincidence degree analysis. If the coincidence degree is insufficient, it means that the adjustment method may not cover the current branch that appears abnormally, so the compensation area is re-planned to ensure that the intervention measures directly act on the abnormal source; avoid resource waste caused by blind compensation, and shorten the response time of abnormal processing.

[0091] Preferably, the running cost described later can be described based on the power loss generated by each power grid node in the target area when the current interacts, that is, the current square value is multiplied by the line resistance of the corresponding power distribution equipment between the power grid nodes, which can be directly obtained by measuring the power distribution equipment; as for whether the current compensation area is the target area, according to the interaction frequency interval of the current frequency at the location of the current power grid node, if the interaction frequency of the area is in this interval, and its running cost is less than 5% of the annual income of electricity at this location, or less than the average value of the running cost in its historical data when the current interacts, it is considered that the area is the target area.

[0092] Preferably, the interval of the interaction frequency is adjusted according to the location of the power distribution equipment, such as industrial power grid, commercial power distribution network and residential area power grid, etc. The interval is set as high frequency interval greater than 12 times / hour, medium frequency interval of 6 to 12 times / hour and low frequency interval less than 6 times / hour. The three intervals can also be adjusted according to the average value of the upper and lower limit values of the historical data in the area to describe the current interaction of the power distribution equipment in different positions of the power grid topology. At this time, the target area is screened according to the interaction frequency and the running cost, which reduces the economic cost while ensuring the compensation effect; then the target area is updated according to the data collected in the upload period, and the dynamic balance of the power grid is realized by periodically updating the target area and controlling the switch state, and the fluctuation caused by frequent adjustment is reduced. Finally, the similarity in the upload period is described, that is, the cosine similarity of the current, voltage and power of the target area corresponding to the power grid node in each upload period is calculated, and then the average value is calculated to describe the similarity of the target area in multiple upload periods. If the similarity is the largest at this time, the switch is closed to solidify the compensation strategy, so as to avoid the power grid fluctuation caused by frequent switching. If the target area for current scheduling interaction does not meet at this time, the switch of the non-target area is disconnected to prevent overcompensation or undercompensation, so as to improve the stability of the power grid control.

[0093] Although the embodiments of the present application have been shown and described above, it should be understood that the above embodiments are exemplary and should not be construed as limiting the present application. Those skilled in the art can make changes, modifications, replacements and variations to the above embodiments within the scope of the present application, which are still covered by the protection scope of the present application.

Claims

1. A method for monitoring the integrated operation of power grid information security, characterized in that: include: S1, obtaining electrical parameter information of distribution equipment from the power grid information, and forming a power grid topology structure based on the connection relationship of the distribution equipment as the power grid nodes. The current interaction area is selected based on the distance between each power grid node and the mobilization area in the topological structure during current interaction; S2: Perform double-layer mapping of the physical and information layers of each grid node in the current interaction area. Extract features from the electrical parameter information of each grid node in the current interaction area based on the grid information upload cycle within a preset time period. Acquire abnormal events of each grid node. Define the temporal interaction form of the grid information based on the interaction frequency and interaction type of the abnormal events. S3, constructs a node change matrix based on the time interaction form of power grid information; By using the conduction path of the node change matrix under chain propagation, the interactive update sequence of the current interaction area is obtained; S4, obtaining the interaction target of each grid node from the interaction update sequence, and determining the current compensation strategy under different current states based on the interaction target of each grid node; S5, interactively evaluating each grid node included in the current compensation strategy, and adjusting the switch corresponding to each grid node.

2. A method for monitoring the integrated operation of power grid information security according to claim 1, characterized in that: The implementation of step S1 includes: When the power distribution equipment is in an on state, constructing a hierarchical connection relationship of at least one power distribution equipment according to the location of the power distribution equipment; The hierarchical connection relationship of the power distribution equipment is used to generate the power grid topology of the power distribution equipment, and the data of each power grid node is integrated according to the electrical parameter information collected on the power grid topology; Monitor the electrical parameter information of each grid node and divide the mobilization area according to the fluctuation ratio of the electrical parameter information of each grid node and the distance between the grid nodes; With the dispatching area as the center, the area radius is set. When the current loss of each grid node contained in the area radius is minimized, each grid node contained in the area radius is output as a current interaction area.

3. A method for monitoring the integrated operation of power grid information security according to claim 1, characterized in that: Electrical parameter information includes current, voltage and power.

4. A method for monitoring the integrated operation of power grid information security according to claim 1, characterized in that: The implementation method of selecting the current interaction area also includes: determining the connection path between each grid node and the switches at adjacent positions based on the number of switches connected to each grid node, and screening each grid node in the current interaction area according to the connection path; so that each grid node in the current interaction area corresponds to at least one switch.

5. A method for monitoring the integrated operation of power grid information security according to claim 1, characterized in that: The implementation of step S2 includes: Check the grid topology of each grid node in the current interaction area, associate the grid topology with the physical location to obtain a mapping relationship at the physical level, introduce the mapping relationship at the physical level into the upload cycle of each grid node, and obtain a mapping relationship at the information level; Verify the electrical parameter information within the upload cycle and identify the characteristic types of abnormal events corresponding to the electrical parameter information. The characteristic types include: continuous abnormality, transient abnormality, periodic abnormality, and associated abnormality of the electrical parameter information; The number of occurrences of the characteristic type of the abnormal event in the preset time period is used as the interaction frequency, and the node associated with the characteristic type of the abnormal event is used as the interaction type; the data corresponding to the interaction frequency and interaction type are output in the form of time interaction.

6. A method for monitoring the integrated operation of power grid information security according to claim 5, characterized in that: The time interaction form of defining power grid information also includes: According to the transmission order of abnormal events between various power grid nodes, the nodes associated with the interaction type are output to form dynamic baseline reference data; Calculate the parameter threshold of the dynamic baseline reference data, update the parameter threshold according to the upload cycle of the power grid information, and use the dynamic baseline reference data to extract the characteristic type of each abnormal event; each updated dynamic baseline reference data is used as the time interactive form of output.

7. A method for monitoring the integrated operation of power grid information security according to claim 1, characterized in that: The implementation of step S3 further includes: Retrieve the grid dispatch instructions from the grid information, define the state propagation set according to the grid dispatch instructions, and generate the state change probability of each grid node based on the grid topology structure; The state change probability is weighted according to the conduction path of each grid node to form a node change matrix; Identify the interaction frequency, interaction duration, and power change gradient of the node change matrix and construct an interaction update sequence.

8. A method for monitoring the integrated operation of power grid information security according to claim 7, characterized in that: The implementation of forming an interactive update sequence also includes: Identify the input terminals corresponding to each grid node. If the current grid node is an input terminal, limit the power supply of the current grid node based on the average value of the power change gradient under the corresponding interaction frequency. Use the interaction frequency, interaction duration, and power change gradient corresponding to the current grid node as the output interaction update sequence. If the current grid node is the output end, the node change matrix is ​​screened according to the interaction frequency and interaction duration of the current grid node, and the data with a power change gradient greater than the average power change gradient is used as the output interaction update sequence; If the current grid node is an intermediate node, the part where the interaction frequency and power change gradient exceed the average value is obtained as the output interaction update sequence.

9. A method for monitoring the integrated operation of power grid information security according to claim 1, characterized in that: Step S4 further includes: The interaction types of each interaction target are queried based on the current branches of each interaction target during current interaction, the minimum value of each interaction type is found, and the interaction target is tracked based on the minimum value of each interaction type; For any interaction type of the interactive target, identify the current, voltage, and power of the interactive target under each interaction type, and fuse the average value of the interactive current, voltage, and power with the limit value to obtain the target value of each interactive target; The current compensation strategy is set according to the target value of each interactive target and the position of the interactive target on the current branch.

10. A method for monitoring the integrated operation of power grid information security according to claim 1, characterized in that: The implementation of step S5 further includes: Obtain the coordinates of each grid node included in the current compensation strategy, generate a current compensation area, and detect the overlap between the current compensation area and the current interaction area; Determine whether the overlap exceeds the overlap threshold. If so, reacquire the current compensation area. Otherwise, determine whether the current compensation area is the target area based on the interaction frequency and operating cost of the current compensation area. If it is the target area, the interaction frequency of the target area is recorded, and the target area is updated with the upload cycle corresponding to the interaction frequency. When the similarity value of the target area is the largest within multiple upload cycles, the switches of each grid node in the target area are closed; if it is not the target area, the switches of each grid node in the current compensation area are disconnected.

Images