Smart perception-based power grid data management method and system

Abstract

The application discloses a power grid data management method and system based on intelligent sensing, and belongs to the technical field of data analysis. The system comprises an intelligent sensing module, a power consumption analysis module, a simulation calculation module and a visual early warning module. The intelligent sensing module is used for collecting GIS maps and meteorological parameters in a power grid coverage area, and configuration information and power consumption data of objects. The power consumption analysis module analyzes power consumption habits of the objects according to the configuration information and the power consumption data, and divides abnormal areas on the GIS maps. Cumulus entities are set for the abnormal areas, and the size and behavior of the cumulus entities are predicted, so that cumulus areas are set. The simulation calculation module adopts a physical model combined with environmental parameters to simulate and analyze the moving track of the cumulus entities in the cumulus areas. Facilities and equipment crossed by the moving track on the GIS maps are analyzed, and the danger indexes of the cumulus entities in each cumulus area are calculated. The visual early warning module early warns the cumulus areas in order of the danger indexes from large to small, and displays real-time pictures of the cumulus areas.

Description

Technical Field

[0001] This invention relates to the field of data analysis technology, specifically to a power grid data management method and system based on intelligent sensing. Background Technology

[0002] With the deepening of global energy structure transformation and sustainable development strategies, photovoltaic power generation, as a clean and renewable energy form, has achieved unprecedented development worldwide. However, the grid connection of large-scale photovoltaic power sources, especially the rapid growth of distributed photovoltaic power, has brought unprecedented challenges to the planning, operation, and management of traditional power grids.

[0003] Currently, with the continuous increase in grid-connected photovoltaic capacity, especially the increasing grid risks caused by cumulus cloud obstruction, the rapid migration of cumulus clouds on a minute-scale timescale leads to a sharp drop in solar radiation intensity, causing abrupt changes in distributed photovoltaic output. This, in turn, forces a sharp increase in grid compensation power, seriously threatening the stability of regional power supply. Traditional grid management models rely on meteorological satellite data and static threshold alarm mechanisms, which are difficult to achieve dynamic coupling analysis of meteorological disasters and power facility operation risks, and have significant limitations. Specifically: 1. Data fragmentation: Meteorological data and grid operation data are processed independently, resulting in the inability to dynamically correlate the cumulus cloud trajectory with the risk status of power facilities. Meteorological warnings can only provide the probability of regional precipitation and cannot quantify the output attenuation of specific photovoltaic arrays. Grid-side anomaly detection mechanisms have difficulty distinguishing between meteorological disturbances and equipment failures, resulting in a high false alarm rate. 2. Limitations of static analysis models: Existing anomaly detection technologies use fixed threshold judgment standards and fail to consider the influence of multi-physical field coupling effects such as temperature and humidity gradients and cumulus cloud movement trajectories on photovoltaic output, resulting in insufficient ability to identify gradual meteorological disturbances. 3. Insufficient response timeliness: Current systems typically initiate response procedures after a fault occurs. However, the sharp drop in power output caused by cumulus cloud cover has a rapid propagation characteristic within minutes. Maintenance personnel rely on manual comparison of satellite cloud images with the power grid topology, resulting in significant response delays and making it difficult to meet the resilience requirements of a high proportion of renewable energy connected to the grid. Summary of the Invention

[0004] The purpose of this invention is to provide a power grid data management method and system based on intelligent sensing, so as to solve the problems mentioned in the background art.

[0005] To address the aforementioned technical problems, this invention provides a power grid data management method based on intelligent sensing, comprising:

[0006] S100: Collect GIS maps and meteorological parameters within the power grid coverage area, as well as the configuration information and electricity consumption data of the objects.

[0007] The power grid coverage area refers to the physical area jointly covered by the power grid's power supply service and the data collection range. Meteorological parameters refer to the distribution of basic physical quantities reflecting the state of the atmosphere in the airspace.

[0008] The "object" refers to an electricity-consuming entity that has dual access to both photovoltaic and power grid, supplementing its own required power resources through self-configured photovoltaic panels combined with dual power grid channels.

[0009] The configuration information includes the location of the object and the performance data of the photovoltaic equipment. Electricity consumption data refers to the photovoltaic power consumption and grid power consumption at different times.

[0010] The performance data of photovoltaic equipment includes key parameters that reflect the actual operating status and environmental adaptability of the equipment, such as installed capacity, conversion efficiency, degradation rate, temperature coefficient, and performance ratio.

[0011] Construct a dynamic sensing network covering the entire domain to achieve three-dimensional data fusion of power grid, meteorological, and photovoltaic equipment, providing a multi-source data foundation for anomaly analysis.

[0012] S200. Based on the configuration information and electricity consumption data analysis of the target objects, anomaly zones are delineated on the GIS map. Cumulus cloud entities are defined for the anomaly zones, and their size and behavior are predicted. Anomaly coefficients are calculated to define the cumulus cloud zones. Specifically, this includes:

[0013] S201. Analyze the electricity consumption data for each object and draw electricity consumption line graphs for each. Analyze electricity consumption habits based on the electricity consumption line graphs and compare them with the current time. The electricity consumption data is used to calculate the anomaly coefficient, thereby setting the anomaly status. Specifically, this includes:

[0014] S2011, Based on object The electricity consumption data is used to create an electricity consumption line graph, which is then evenly divided into cycles and time points.

[0015] The fluctuation index at each time point is calculated and combined to set the electricity consumption range within the cycle. Specifically, this includes:

[0016] S2011-1, Obtaining the Object The electricity consumption data is used to calculate the sum of the power consumption of photovoltaic power and the power consumption of the grid at the same time, which is taken as the total power consumption. The change of total power consumption over time is analyzed to draw a line graph of electricity consumption.

[0017] S2011-2, Setting the Loop Duration The electricity consumption line graph is evenly divided into sections with a duration of [duration]. The time period is used as the cycle period. Within each cycle period, the time intervals are evenly distributed... Analyze the total power consumption at each time point. .

[0018] S2011-3, According to the formula: Calculate the fluctuation coefficient at each time point. The time points where the fluctuation coefficient is greater than a threshold are selected. This represents the average total power consumption at all points in time within the cycle.

[0019] S2011-4. Calculate the standard deviation of the volatility coefficient over all cycles at the same time point, and use it as the volatility index. By freely combining two or more consecutive time points within the cycle, a total of [number] time points are established. A combination.

[0020] S2011-5 Calculate the average volatility index across all time points within each combination. Filter out the smallest The corresponding combination The intervals corresponding to the beginning and end of consecutive time points in the combination are taken as the power consumption intervals of the cycle.

[0021] The electricity consumption zone is a fixed time period within a cycle, such as the period from 9 am to 6 pm every day. If the current time is 8 pm, it does not fall within the electricity consumption zone. The electricity consumption zone also needs to be further intersected with the operating time period of the photovoltaic equipment.

[0022] By iterating through all combinations of consecutive time points within the cycle, the combination with the highest volatility stability is selected. The corresponding time period is designated as the electricity consumption zone.

[0023] By avoiding blind spots in fixed time windows, the baseline range with the smallest power changes is accurately identified, providing a high-confidence reference for subsequent anomaly detection.

[0024] S2012, Analysis of meteorological parameters and objects The electricity consumption data is sampled at each time point within the electricity consumption interval of R cycles. The specific time corresponding to each time point is analyzed to obtain the meteorological parameters and photovoltaic power consumption at each time point.

[0025] S2013. Using the photovoltaic power consumption at any given moment as the dependent variable and all meteorological parameters as independent variables, and packaging them into a sample, all samples are input into a linear regression model for training, thereby obtaining the relationship expression between meteorological parameters and photovoltaic power consumption.

[0026] Specifically, it includes:

[0027] S2013-1. Establish a linear regression model and set the intercept. and regression coefficients The independent variables from these samples are used as input values. The output value of each sample The difference between the dependent variable and the differential variable is used as the difference coefficient. The expression is as follows:

[0028] ;

[0029] S2013-2. Adjust the intercept and regression coefficients until the sum of the difference coefficients of all samples is minimized, and obtain the relationship after training is complete.

[0030] S2014, If the current time In object Within the electricity consumption range of the cycle, the current meteorological parameters are substituted into the expression to calculate the predicted photovoltaic power consumption. .

[0031] S2015. Based on current electricity consumption data and predicted photovoltaic power consumption... , computational object coefficient of variation Objects with an anomaly coefficient greater than a threshold will be set to an anomaly state. Specifically, this includes:

[0032] S2015-1, Time analysis based on electricity consumption data Photovoltaic power consumption and power consumption of the power grid Time Mapping the electricity consumption curve to the electricity consumption intervals within each cycle, select the closest time point. .

[0033] S2015-2, Obtaining Time Points At corresponding times within each of the R cycle periods, the total power consumption is obtained by summing the power consumption of photovoltaic power and grid power at the same time. Then, the average of these R total power consumptions is calculated. .

[0034] S2015-3, Analysis Time Points Volatility Index Substitute the object into the formula for calculation The coefficient of variation:

[0035] ;

[0036] In the formula, It is a constant greater than 1. This represents the maximum fluctuation index across all time points within the electricity consumption range.

[0037] This term is used to balance the impact of the fluctuation index. When the data at a given point in time fluctuates significantly, a lower base is provided to suppress the impact of power consumption fluctuations on the anomaly coefficient. Conversely, a higher base is provided to amplify the impact of power consumption fluctuations on the anomaly coefficient.

[0038] Used to calculate the gap between the actual power consumption of photovoltaic systems and the predicted value.

[0039] Used to calculate the percentage increase in actual power consumption of the power grid compared to the predicted value.

[0040] By identifying the objects of photovoltaic equipment affected by meteorological interference through the anomaly coefficient, the false alarm bottleneck of traditional threshold alarms is overcome.

[0041] S2016. Draw power consumption line graphs for each object, divide different cycle periods and calculate the anomaly coefficient. If the conditions are met, set the anomaly status.

[0042] Different cycle periods are set for each object, and the specific values ​​are predefined based on the object's production activity time and electricity consumption habits.

[0043] For example, if a factory's production activities typically run from 8 a.m. to 6 p.m. each day, then the cycle time is set to 24 hours, from midnight to midnight each day.

[0044] If a company's production activities typically take place on weekdays, then the cycle should be set to 24 x 7 hours, from Monday to Sunday each week.

[0045] S202. Mark objects with abnormal status on the GIS map. and the surrounding distance Other objects within the system are set to monitoring status, and these other objects are designated as objects. When the object When the monitoring status changes to an abnormal status, the calculation object and Distance between .

[0046] Status setting refers to setting an object in a normal state to an abnormal state, while status change refers to setting an object in a monitored state to an abnormal state.

[0047] S203, Analysis Object Interval between state changes And set the warning duration. .Will Divide by the interval duration Then multiply by the warning duration The warning distance was then obtained. .

[0048] S204, using objects Location is the center of the circle, warning distance A circular region with a fixed radius is established as the anomaly zone, and cumulus entities are set within the anomaly zone. This breaks through the limitation of a fixed radius and enables adaptive modeling of the spatiotemporal propagation of meteorological disturbances.

[0049] By dividing the area into anomalous zones, we can model the spatiotemporal diffusion of meteorological disturbance events and accurately delineate the impact range of cumulus clouds.

[0050] S205, Based on objects within the exception area The state changes and time intervals of cumulus cloud entities are analyzed to determine their size and behavior parameters and calculate anomaly coefficients. Anomaly regions with anomaly coefficients exceeding a threshold are identified as cumulus cloud regions. Specifically, this includes:

[0051] S2051. Obtain the change time of the object status within the abnormal area, analyze the temporal sequence of the change time of different objects, and analyze the scale parameters of the cumulus cloud entity, including its location and coverage area, in conjunction with the location of each object in the GIS map.

[0052] The coverage area adopts the maximum coverage range. When multiple objects change state at the same time, the location distribution of the monitoring state objects corresponding to these objects is analyzed. Combined with the typical shape of cumulus clouds, a maximum coverage area is planned that covers both the location of the object with the changed state and is within the range of the location distribution of the monitoring state objects.

[0053] S2052. Calculate the distance between different object locations and the time interval of the change, and analyze the behavioral parameters of the cumulus cloud entity, including the direction of movement and the speed of movement.

[0054] S2053. Analyze the distance between each normal-state object and the location of the cumulus entity within the abnormal area. Obtain the installed capacity of photovoltaic equipment in each normal state object. Substitute the values ​​into the formula to calculate the anomaly coefficient for each anomaly region. :

[0055] ;

[0056] In the formula, This represents the number of normal objects within the abnormal region. and The first The current photovoltaic power consumption and grid power consumption of a normal state object.

[0057] Anomaly coefficients are used to screen high-threat areas characterized by high-density photovoltaic clusters, close-range cumulus cloud coverage, and strong power replenishment needs. The anomaly coefficient calculation quantifies the potential threat posed by cumulus clouds to the power grid, filtering out low-risk anomaly areas.

[0058] S300 uses a physical model combined with environmental parameters to simulate and analyze the movement trajectory of cumulus entities within a cumulus cloud region. It analyzes the facilities and equipment traversed by the movement trajectory on the GIS map and calculates the hazard index of cumulus entities within each cumulus cloud region. Specifically, this includes:

[0059] S301. Using the scale and behavior parameters of cumulus entities within the cumulus region, as well as environmental parameters, as initial data, an atmospheric hydrodynamic model is driven by temperature and humidity gradient field and wind speed and direction data to simulate the movement and evolution of cumulus entities.

[0060] S302. Combining topographic elevation and surface type spatial data from GIS maps, calculate the correction effect of topographic lifting and surface thermal disturbance on the movement of cumulus entities.

[0061] S303. By introducing the characteristics of the environmental pressure gradient distribution, we analyze its dynamic impact on the development pattern and movement speed of cumulus cloud entities at vertical height, thereby deducing the movement trend of cumulus cloud entities.

[0062] S304. Generate a time-varying spatial location sequence of cumulus entities on the GIS map based on numerical iteration, and output the warning duration. The movement trajectory within.

[0063] S305. Analyze the photovoltaic power stations and objects traversed by the movement trajectories of each cumulus entity on the GIS map, designating them as nodes. Calculate the hazard index of the corresponding cumulus entity based on the installed capacity of each node. The formula is as follows:

[0064] ;

[0065] After deformation, we obtain:

[0066] ;

[0067] In the formula, This represents the number of nodes traversed by the movement trajectory. For the first in the movement trajectory The installed capacity of each node, For the first in the movement trajectory The trajectory distance between each node and the cumulus cloud entity.

[0068] Calculate the sum of the hazard indices of all cumulus entities in the cumulus cloud region, and then sort and display all cumulus cloud regions in descending order of the sum of their hazard indices.

[0069] The S400 and data center issue warnings for cumulus cloud areas in descending order of risk index and display real-time images of the cumulus cloud areas.

[0070] By overlaying cumulus cloud trajectories, real-time monitoring images, and GIS layers, a spatiotemporally coupled emergency decision-making view is generated, improving the emergency response speed of power grid planning and dispatch.

[0071] The present invention also provides a power grid data management system based on intelligent sensing, including an intelligent sensing module, an electricity consumption analysis module, a simulation calculation module, and a visualization early warning module.

[0072] The intelligent sensing module is used to collect GIS maps and meteorological parameters within the power grid coverage area, as well as the configuration information and electricity consumption data of objects.

[0073] Collect multi-dimensional data of the power grid coverage area, including GIS geospatial information, meteorological parameters, location coordinates and performance parameters of photovoltaic objects, and real-time electricity consumption data.

[0074] A dynamic sensing network is constructed to provide a raw data pool of spatiotemporal correlations for subsequent analysis, supporting multi-source correlation modeling between electricity consumption behavior and meteorological disturbances.

[0075] The power consumption analysis module divides abnormal zones on a GIS map based on configuration information and the power consumption habits of the data analysis subjects. It then defines cumulus cloud entities for these abnormal zones, predicts their size and behavior, and calculates anomaly coefficients to define the cumulus cloud areas.

[0076] Based on the electricity consumption curve fluctuation index, abnormal state objects are identified, photovoltaic power generation is predicted through a linear regression model, and abnormal areas are divided by combining the spatiotemporal correlation of the objects.

[0077] Further analysis was conducted on the scale and behavior of cumulus entities within the anomalous area, and anomaly coefficients were calculated based on installed capacity and power consumption to screen high-threat cumulus cloud areas.

[0078] It enables intelligent mapping of abnormal power consumption to potential meteorological hazards, identifies high-threat areas, reduces the cost of manual monitoring, and improves the accuracy of early warnings.

[0079] The simulation module uses a physical model combined with environmental parameters to simulate and analyze the movement trajectory of cumulus entities within cumulus cloud areas. It analyzes the facilities and equipment traversed by the movement trajectory on the GIS map and calculates the hazard index of cumulus entities within each cumulus cloud area.

[0080] By coupling atmospheric hydrodynamics models with GIS topographic data, the cumulus cloud movement trajectory is simulated by driving the temperature and humidity gradient field, and the evolution path is dynamically corrected by superimposing the topographic lifting effect and pressure gradient.

[0081] The danger index is calculated by traversing the power grid facility nodes traversed by the trajectory, based on the installed capacity and trajectory distance.

[0082] Quantitatively predict the dynamic threat of cumulus clouds to power grid facilities, output spatiotemporally accurate trajectory risk sequences, and support priority-driven emergency decision-making.

[0083] The visualization warning module issues warnings for cumulus cloud areas in descending order of hazard index and displays real-time images of the cumulus cloud areas.

[0084] Sort the data in descending order by the sum of the danger index of the cumulus cloud area, retrieve real-time monitoring images and overlay them with GIS geographic layers to generate a spatiotemporally correlated visual early warning panel.

[0085] This enables a clear presentation and rapid location of risk situations, improving the response efficiency and accuracy of operations and maintenance personnel in dealing with high-threat cumulus cloud areas.

[0086] Compared with the prior art, the beneficial effects achieved by the present invention are:

[0087] Dynamic coupling perception: By synchronously collecting power consumption data from dual channels of the photovoltaic power grid, meteorological parameters, and GIS spatial data, a multi-dimensional perception network of "power-meteorology-geography" is constructed. This breaks through the limitations of the separation between meteorological satellite and power grid monitoring data in traditional technologies, and realizes real-time correlation mapping between cumulus cloud cover and power consumption fluctuations.

[0088] Adaptive anomaly detection: The fluctuation index combination optimization algorithm is used to lock the benchmark power consumption range, and the impact of weather on photovoltaic power output is quantified by combining a linear regression model. By integrating the three dimensions of fluctuation deviation, power generation gap and power replenishment anomaly through the anomaly coefficient, the system can accurately distinguish between weather interference and equipment failure, thus solving the problem of high false alarm rate of traditional fixed threshold alarms.

[0089] Spatiotemporal propagation modeling: Based on the spatiotemporal relationship of object state changes, the early warning distance is dynamically calculated, and an adaptive model of cumulus cloud diffusion radius is constructed; by coupling distance-installed capacity-power weight through anomaly coefficient, the high-threat cumulus cloud area is accurately delineated, replacing the static area monitoring mechanism.

[0090] Physics-driven simulation: Couples atmospheric hydrodynamics model with GIS topographic lifting effect, introduces pressure gradient to dynamically correct cumulus cloud movement trajectory; quantifies cumulus cloud movement risk through hazard index, supports priority protection of critical facilities, and overcomes the shortcomings of insufficient prediction accuracy of statistical model.

[0091] Collaborative decision-making response: Triggering visual early warnings by sorting according to the sum of the danger index, automatically overlaying cumulus cloud trajectories, real-time monitoring images and GIS layers to generate a spatiotemporally coupled decision view, upgrading the traditional manual comparison response mode to an active defense closed loop. Attached Figure Description

[0092] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings:

[0093] Figure 1This is a flowchart illustrating the power grid data management method based on intelligent sensing according to the present invention.

[0094] Figure 2 This is a schematic diagram of the power grid data management system based on intelligent sensing according to the present invention. Detailed Implementation

[0095] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0096] Please see Figure 1 This invention provides a power grid data management method based on intelligent sensing, comprising:

[0097] S100: Collect GIS maps and meteorological parameters within the power grid coverage area, as well as the configuration information and electricity consumption data of the objects.

[0098] The power grid coverage area refers to the physical area jointly covered by the power grid's power supply service and the data collection range. Meteorological parameters refer to the distribution of basic physical quantities reflecting the state of the atmosphere in the airspace.

[0099] The "object" refers to an electricity-consuming entity that has dual access to both photovoltaic and power grid, supplementing its own required power resources through self-configured photovoltaic panels combined with dual power grid channels.

[0100] The configuration information includes the location of the object and the performance data of the photovoltaic equipment. Electricity consumption data refers to the photovoltaic power consumption and grid power consumption at different times.

[0101] The performance data of photovoltaic equipment includes key parameters that reflect the actual operating status and environmental adaptability of the equipment, such as installed capacity, conversion efficiency, degradation rate, temperature coefficient, and performance ratio.

[0102] Construct a dynamic sensing network covering the entire domain to achieve three-dimensional data fusion of power grid, meteorological, and photovoltaic equipment, providing a multi-source data foundation for anomaly analysis.

[0103] S200. Based on the configuration information and electricity consumption data analysis of the target objects, anomaly zones are delineated on the GIS map. Cumulus cloud entities are defined for the anomaly zones, and their size and behavior are predicted. Anomaly coefficients are calculated to define the cumulus cloud zones. Specifically, this includes:

[0104] S201. Analyze the electricity consumption data for each object and draw electricity consumption line graphs for each. Analyze electricity consumption habits based on the electricity consumption line graphs and compare them with the current time. The electricity consumption data is used to calculate the anomaly coefficient, thereby setting the anomaly status. Specifically, this includes:

[0105] S2011, Based on object The electricity consumption data is used to create an electricity consumption line graph, which is then evenly divided into cycles and time points.

[0106] The fluctuation index at each time point is calculated and combined to set the electricity consumption range within the cycle. Specifically, this includes:

[0107] S2011-1, Obtaining the Object The electricity consumption data is used to calculate the sum of the power consumption of photovoltaic power and the power consumption of the grid at the same time, which is taken as the total power consumption. The change of total power consumption over time is analyzed to draw a line graph of electricity consumption.

[0108] S2011-2, Setting the Loop Duration The electricity consumption line graph is evenly divided into sections with a duration of [duration]. The time period is used as the cycle period. Within each cycle period, the time intervals are evenly distributed... Analyze the total power consumption at each time point. .

[0109] S2011-3, According to the formula: Calculate the fluctuation coefficient at each time point. The time points where the fluctuation coefficient is greater than a threshold are selected. This represents the average total power consumption at all points in time within the cycle.

[0110] S2011-4. Calculate the standard deviation of the volatility coefficient over all cycles at the same time point, and use it as the volatility index. By freely combining two or more consecutive time points within the cycle, a total of [number] time points are established. A combination.

[0111] S2011-5 Calculate the average volatility index across all time points within each combination. Filter out the smallest The corresponding combination The intervals corresponding to the beginning and end of consecutive time points in the combination are taken as the power consumption intervals of the cycle.

[0112] The electricity consumption zone is a fixed time period within a cycle, such as the period from 9 am to 6 pm every day. If the current time is 8 pm, it does not fall within the electricity consumption zone. The electricity consumption zone also needs to be further intersected with the operating time period of the photovoltaic equipment.

[0113] By iterating through all combinations of consecutive time points within the cycle, the combination with the highest volatility stability is selected. The time period corresponding to the smallest value is taken as the electricity consumption interval.

[0114] By avoiding blind spots in fixed time windows, the baseline range with the smallest power changes is accurately identified, providing a high-confidence reference for subsequent anomaly detection.

[0115] S2012, Analysis of meteorological parameters and objects The electricity consumption data is sampled at each time point within the electricity consumption interval of R cycles. The specific time corresponding to each time point is analyzed to obtain the meteorological parameters and photovoltaic power consumption at each time point.

[0116] S2013. Using the photovoltaic power consumption at any given moment as the dependent variable and all meteorological parameters as independent variables, and packaging them into a sample, all samples are input into a linear regression model for training, thereby obtaining the relationship expression between meteorological parameters and photovoltaic power consumption.

[0117] Specifically, it includes:

[0118] S2013-1. Establish a linear regression model and set the intercept. and regression coefficients The independent variables from these samples are used as input values. The output value of each sample The difference between the dependent variable and the differential variable is used as the difference coefficient. The expression is as follows:

[0119] ;

[0120] S2013-2. Adjust the intercept and regression coefficients until the sum of the difference coefficients of all samples is minimized, and obtain the relationship after training is complete.

[0121] S2014, If the current time In object Within the electricity consumption range of the cycle, the current meteorological parameters are substituted into the expression to calculate the predicted photovoltaic power consumption. .

[0122] S2015. Based on current electricity consumption data and predicted photovoltaic power consumption... , computational object coefficient of variation Objects with an anomaly coefficient greater than a threshold will be set to an anomaly state. Specifically, this includes:

[0123] S2015-1, Time analysis based on electricity consumption data Photovoltaic power consumption and power consumption of the power grid Time Mapping the electricity consumption curve to the electricity consumption intervals within each cycle, select the closest time point. .

[0124] S2015-2, Obtaining Time Points At corresponding times within each of the R cycle periods, the total power consumption is obtained by summing the power consumption of photovoltaic power and grid power at the same time. Then, the average of these R total power consumptions is calculated. .

[0125] S2015-3, Analysis Time Points Volatility Index Substitute the object into the formula for calculation The coefficient of variation:

[0126] ;

[0127] In the formula, It is a constant greater than 1. This represents the maximum fluctuation index across all time points within the electricity consumption range.

[0128] This term is used to balance the impact of the fluctuation index. When the data at a given point in time fluctuates significantly, a lower base is provided to suppress the impact of power consumption fluctuations on the anomaly coefficient. Conversely, a higher base is provided to amplify the impact of power consumption fluctuations on the anomaly coefficient.

[0129] Used to calculate the gap between the actual power consumption of photovoltaics and the predicted value (such as the reduction in photovoltaic power generation efficiency caused by cumulus cloud shading).

[0130] Used to calculate the percentage increase in actual power consumption of the power grid compared to the predicted value (e.g., when cumulus cloud cover reduces the efficiency of photovoltaic power generation, the target passively increases the power consumption of the power grid to meet the electricity demand).

[0131] By identifying the objects of photovoltaic equipment affected by meteorological interference through the anomaly coefficient, the false alarm bottleneck of traditional threshold alarms is overcome.

[0132] S2016. Draw power consumption line graphs for each object, divide different cycle periods and calculate the anomaly coefficient. If the conditions are met, set the anomaly status.

[0133] Different cycle periods are set for each object, and the specific values ​​are predefined based on the object's production activity time and electricity consumption habits.

[0134] For example, if a factory's production activities typically run from 8 a.m. to 6 p.m. each day, then the cycle time is set to 24 hours, from midnight to midnight each day.

[0135] If a company's production activities typically take place on weekdays, then the cycle should be set to 24 x 7 hours, from Monday to Sunday each week.

[0136] S202. Mark objects with abnormal status on the GIS map. and the surrounding distance Other objects within the system are set to monitoring status, and these other objects are designated as objects. When the object When the monitoring status changes to an abnormal status, the calculation object and Distance between .

[0137] Status setting refers to setting an object in a normal state to an abnormal state, while status change refers to setting an object in a monitored state to an abnormal state.

[0138] S203, Analysis Object Interval between state changes And set the warning duration. .Will Divide by the interval duration Then multiply by the warning duration The warning distance was then obtained. .

[0139] S204, using objects Location is the center of the circle, warning distance A circular region with a fixed radius is established as the anomaly zone, and cumulus entities are set within the anomaly zone. This breaks through the limitation of a fixed radius and enables adaptive modeling of the spatiotemporal propagation of meteorological disturbances.

[0140] By dividing the area into anomalous zones, we can model the spatiotemporal diffusion of meteorological disturbance events and accurately delineate the impact range of cumulus clouds.

[0141] S205, Based on objects within the exception area The state changes and time intervals of cumulus cloud entities are analyzed to determine their size and behavior parameters and calculate anomaly coefficients. Anomaly regions with anomaly coefficients exceeding a threshold are identified as cumulus cloud regions. Specifically, this includes:

[0142] S2051. Obtain the change time of the object status within the abnormal area, analyze the temporal sequence of the change time of different objects, and analyze the scale parameters of the cumulus cloud entity, including its location and coverage area, in conjunction with the location of each object in the GIS map.

[0143] The coverage area adopts the maximum coverage range. When multiple objects change state at the same time, the location distribution of the monitoring state objects corresponding to these objects is analyzed. Combined with the typical shape of cumulus clouds, a maximum coverage area is planned that covers both the location of the object with the changed state and is within the range of the location distribution of the monitoring state objects.

[0144] S2052. Calculate the distance between different object locations and the time interval of the change, and analyze the behavioral parameters of the cumulus cloud entity, including the direction of movement and the speed of movement.

[0145] S2053. Analyze the distance between each normal-state object and the location of the cumulus entity within the abnormal area. Obtain the installed capacity of photovoltaic equipment in each normal state object. Substitute the values ​​into the formula to calculate the anomaly coefficient for each anomaly region. :

[0146] ;

[0147] In the formula, This represents the number of normal objects within the abnormal region. and The first The current photovoltaic power consumption and grid power consumption of a normal state object.

[0148] Anomaly coefficients are used to screen high-threat areas characterized by high-density photovoltaic clusters, close-range cumulus cloud coverage, and strong power replenishment needs. The anomaly coefficient calculation quantifies the potential threat posed by cumulus clouds to the power grid, filtering out low-risk anomaly areas.

[0149] S300 uses a physical model combined with environmental parameters to simulate and analyze the movement trajectory of cumulus entities within a cumulus cloud region. It analyzes the facilities and equipment traversed by the movement trajectory on the GIS map and calculates the hazard index of cumulus entities within each cumulus cloud region. Specifically, this includes:

[0150] S301. Using the scale and behavior parameters of cumulus entities within the cumulus region, as well as environmental parameters, as initial data, an atmospheric hydrodynamic model is driven by temperature and humidity gradient field and wind speed and direction data to simulate the movement and evolution of cumulus entities.

[0151] S302. Combining topographic elevation and surface type spatial data from GIS maps, calculate the correction effect of topographic lifting and surface thermal disturbance on the movement of cumulus entities.

[0152] S303. By introducing the characteristics of the environmental pressure gradient distribution, we analyze its dynamic impact on the development pattern and movement speed of cumulus cloud entities at vertical height, thereby deducing the movement trend of cumulus cloud entities.

[0153] S304. Generate a time-varying spatial location sequence of cumulus entities on the GIS map based on numerical iteration, and output the warning duration. The movement trajectory within.

[0154] S305. Analyze the photovoltaic power stations and objects traversed by the movement trajectories of each cumulus entity on the GIS map, designating them as nodes. Calculate the hazard index of the corresponding cumulus entity based on the installed capacity of each node. The formula is as follows:

[0155] ;

[0156] After deformation, we obtain:

[0157] ;

[0158] In the formula, This represents the number of nodes traversed by the movement trajectory. For the first in the movement trajectory The installed capacity of each node, For the first in the movement trajectory The trajectory distance between each node and the cumulus cloud entity.

[0159] Calculate the sum of the hazard indices of all cumulus entities in the cumulus cloud region, and then sort and display all cumulus cloud regions in descending order of the sum of their hazard indices.

[0160] The S400 and data center issue warnings for cumulus cloud areas in descending order of risk index and display real-time images of the cumulus cloud areas.

[0161] By overlaying cumulus cloud trajectories, real-time monitoring images, and GIS layers, a spatiotemporally coupled emergency decision-making view is generated, improving the emergency response speed of power grid planning and dispatch.

[0162] Please see Figure 2 The present invention also provides a power grid data management system based on intelligent sensing, including an intelligent sensing module, an electricity consumption analysis module, a simulation calculation module and a visualization early warning module.

[0163] The intelligent sensing module is used to collect GIS maps and meteorological parameters within the power grid coverage area, as well as the configuration information and electricity consumption data of objects.

[0164] Collect multi-dimensional data of the power grid coverage area, including GIS geospatial information, meteorological parameters (such as temperature, humidity, and air pressure gradient), location coordinates and performance parameters of photovoltaic objects (installed capacity, conversion efficiency, etc.), and real-time electricity consumption data (power consumption of both photovoltaic and power grid).

[0165] A dynamic sensing network is constructed to provide a raw data pool of spatiotemporal correlations for subsequent analysis, supporting multi-source correlation modeling between electricity consumption behavior and meteorological disturbances.

[0166] The power consumption analysis module divides abnormal zones on a GIS map based on configuration information and the power consumption habits of the data analysis subjects. It then defines cumulus cloud entities for these abnormal zones, predicts their size and behavior, and calculates anomaly coefficients to define the cumulus cloud areas.

[0167] Based on the electricity consumption curve fluctuation index, abnormal state objects are identified, photovoltaic power generation is predicted through a linear regression model, and abnormal areas are divided by combining the spatiotemporal correlation of the objects.

[0168] Further analysis was conducted on the scale and behavior of cumulus entities within the anomalous area, and anomaly coefficients were calculated based on installed capacity and power consumption to screen high-threat cumulus cloud areas.

[0169] It enables intelligent mapping of abnormal power consumption to meteorological hazards, identifies local high-threat areas (cumulus cloud areas), reduces the cost of manual monitoring, and improves the accuracy of early warning.

[0170] The simulation module uses a physical model combined with environmental parameters to simulate and analyze the movement trajectory of cumulus entities within cumulus cloud areas. It analyzes the facilities and equipment traversed by the movement trajectory on the GIS map and calculates the hazard index of cumulus entities within each cumulus cloud area.

[0171] By coupling atmospheric hydrodynamics models with GIS topographic data, the cumulus cloud movement trajectory is simulated by driving the temperature and humidity gradient field, and the evolution path is dynamically corrected by superimposing the topographic lifting effect and pressure gradient.

[0172] The danger index is calculated by traversing the power grid facility nodes traversed by the trajectory, based on the installed capacity and trajectory distance.

[0173] Quantitatively predict the dynamic threat of cumulus clouds to power grid facilities, output spatiotemporally accurate trajectory risk sequences, and support priority-driven emergency decision-making.

[0174] The visualization warning module issues warnings for cumulus cloud areas in descending order of hazard index and displays real-time images of the cumulus cloud areas.

[0175] Sort the data in descending order by the sum of the danger index of the cumulus cloud area, retrieve real-time monitoring images (such as camera or satellite images) and overlay them with the GIS geographic layer to generate a spatiotemporally correlated visual early warning panel.

[0176] This enables a clear presentation and rapid location of risk situations, improving the response efficiency and accuracy of operations and maintenance personnel in dealing with high-threat cumulus cloud areas.

[0177] Example 1: Assume there are three nodes on the trajectory of the cumulus cloud entity, with installed capacities of 2.4MW, 1.8MW and 3.5MW respectively, and distances from the cumulus cloud entity to the trajectory of 2.5km, 3.6km and 5.2km respectively;

[0178] When distance The distance is 2km, and the interval is [time]. The warning duration is 0.16 hours. When the time is 1 hour, substitute the values ​​into the formula to calculate the hazard index of the cumulus cloud entity:

[0179] ;

[0180] The danger index of the cumulus entity is 5.25.

[0181] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0182] Finally, it should be noted that the above descriptions are merely preferred embodiments 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 foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A smart sensing based power grid data management method, characterized in that: The method comprises: S100, collecting GIS maps and meteorological parameters in the power grid coverage area, and configuration information and power consumption data of objects; The power grid coverage area refers to a physical area jointly covered by power grid power supply services and data collection range; the meteorological parameter refers to the distribution of basic physical quantities reflecting the state of the atmosphere in the airspace; The object refers to a power consumption entity with dual access of photovoltaic and power grid, which supplements the power resources required by the object itself through self-configured photovoltaic panels combined with dual-channel power grid; The configuration information includes the location of the object and the performance data of the photovoltaic equipment; the power consumption data refers to the photovoltaic power consumption and power grid power consumption of the object at different times; S200, analyzing the power consumption habits of the object according to the configuration information and power consumption data, and dividing the abnormal area on the GIS map; setting cumulus entities for the abnormal area and predicting their size and behavior, and calculating the abnormal coefficient to set the cumulus area; specifically comprising: S201、analyze the power consumption data of each object, respectively draw the power consumption line chart; according to the power consumption habit analyzed according to the power consumption line chart, compare the current time power consumption data to calculate the abnormal coefficient, so as to set the abnormal state; specifically including: S2011、According to the object draws a power consumption line graph according to the power consumption data of the object, divides the cycle period and time points uniformly in the power consumption line graph; calculates the fluctuation index of each time point and establishes a combination, so as to set the power consumption interval in the cycle period; S2012, Analysis of meteorological parameters and objects The electricity consumption data is sampled at each time point within the electricity consumption interval in R cycles; the specific time corresponding to each time point is analyzed to obtain the meteorological parameters and photovoltaic power consumption at each time point; S2013, taking the photovoltaic power consumption at the same time as the dependent variable, and all meteorological parameters as independent variables and packing them as samples; all samples are input into the linear regression model for training to obtain the relationship expression of meteorological parameters and photovoltaic power consumption; S2014、if the current time is in the electricity consumption interval of the cycle period of the object , the current meteorological parameter is substituted into the expression to calculate the predicted photovoltaic electricity consumption power ; S2015、according to the current power consumption data and the predicted photovoltaic power consumption , calculate the transaction coefficient of the object Set the transaction state of the object with a transaction coefficient greater than the threshold value;​ S2016, respectively drawing power consumption line graphs for each object, dividing different cycle periods, and calculating abnormal motion coefficients, and setting abnormal motion states if the conditions are met; S202, mark the abnormal state object on the GIS map , and set the surrounding distance of other objects in the area as a monitoring state, and set the other objects as objects ; when the monitoring state of the object changes to an abnormal state, calculate the distance between the object and ; S203、analysis object Interval duration of state change And set the early warning duration ; Will Except interval duration After multiplying the early warning duration Get early warning distance ; S204、with the object the position as the center, the warning distance as the radius establish a circular area as the abnormal area with the position as the center and the warning distance as the radius, and set the cumulus entity in the abnormal area; S205、According to the state change and time interval of the object in the abnormal area , analyze the size parameters and behavior parameters of cumulus entities and calculate the abnormal coefficient, and take the abnormal area with an abnormal coefficient greater than the threshold value as a cumulus area. S300, using a physical model combined with environmental parameters to simulate and analyze the moving track of the cumulus entity in the cumulus area; analyzing the facilities and equipment crossed by the moving track on the GIS map, and calculating the danger index of the cumulus entity in each cumulus area; S400, the data center sequentially warns the cumulus area in descending order of the danger index, and displays the real-time picture of the cumulus area. 2.The smart sensing based power grid data management method of claim 1, wherein: S2011 comprises: S2011-1, obtaining the power consumption data of the object the sum of the power consumption of the photovoltaic and the power grid as the total power consumption; and analyzing the change of the total power consumption over time to draw a power consumption line graph. S2011-2, set the cycle length , in the electricity consumption line graph is divided into time length of the time period as a cycle period; each cycle period is evenly set time points, analysis of each time point corresponding to the total power ; S2011-3, according to the formula: The fluctuation coefficient of each time point is calculated respectively The time points with fluctuation coefficients greater than the threshold value are screened out; wherein, is the average value of total power consumption of all time points in the cycle period; S2011-4, calculate the standard deviation of the fluctuation coefficient at all cycle periods at the same time point, and take it as the fluctuation index ; freely combine two or more time points in a cycle period, a total of combinations are established; S2011-5, calculate the fluctuation index average value of all time points in each combination , filter out the minimum corresponding combination , and take the interval corresponding to the head and tail positions of the continuous time points in the combination as the power consumption interval of the cycle period. 3.The smart sensing based power grid data management method of claim 1, wherein: S2015 comprises: S2015-1, Analyzing time from power consumption data S2015-2, Calculating photovoltaic power S2015-3, Calculating grid power S2015-4, Mapping time to each power consumption interval S2015-5, Selecting the nearest time point ​ S2015-2, obtaining time point The total power consumption is obtained by summing the photovoltaic and grid power consumption at the same time, and the average value of the R total power consumptions is obtained ; S2015-3, analysis time point fluctuation index of , substitute formula calculation object transaction coefficient: ； In the formula, is a constant greater than 1, is the maximum fluctuation index among all time points in the power consumption section. 4.The smart sensing based power grid data management method of claim 1, wherein: S205 comprises: S2051, obtaining the change time of the object state in the abnormal area, analyzing the timing of the change time of different objects, combining the positions of the objects in the GIS map, analyzing the size parameters of the cumulus entity, including the position and coverage area; S2052, calculating the distance between different object positions and the interval time length of the change time, analyzing the behavior parameters of the cumulus entity, including the moving direction and moving speed; S2053、analyzing the distance between each normal state object and the cumulus entity position in the abnormal area , obtaining the installed capacity of the photovoltaic device of each normal state object ; substituting the formula to calculate the abnormal coefficient of each abnormal area : ； wherein is the number of normal state objects in the abnormal area, and are the current photovoltaic power and grid power of the th normal state object, respectively.

5. The smart sensing based power grid data management method of claim 1, wherein: S300 comprises: S301, taking the size parameters and behavior parameters of the cumulus entity in the cumulus area, and the environmental parameters as initial data, using the temperature and humidity gradient field and wind speed and direction data to drive the atmospheric fluid mechanics model to simulate the moving evolution process of the cumulus entity; S302, combining the terrain elevation and surface type spatial data in the GIS map, calculating the correction effect of the terrain lifting effect and the surface thermal disturbance on the movement of the cumulus entity; S303, by introducing the environmental pressure gradient distribution characteristics, analyzing its dynamic influence on the development form and moving speed change of the cumulus entity in the vertical height, and deducing the moving trend of the cumulus entity; S304、According to the numerical iteration, a sequence of spatial positions of the cumulus entity over time on the GIS map is generated, and a warning duration is output The moving trajectory within the inner S305, analyze the photovoltaic power station and the object crossed by the moving track of each cumulus entity on the GIS map, as nodes respectively; combine the installed capacity of each node to calculate the risk index of the corresponding cumulus entity respectively . 6.The smart sensing based power grid data management method of claim 5, wherein: In S305, the danger index calculation formula is as follows: ； In the formula, This represents the number of nodes traversed by the movement trajectory. For the first in the movement trajectory The installed capacity of each node, For the first in the movement trajectory The trajectory distance between each node and the cumulus entity.

7. The smart sensing based power grid data management system as claimed in claim 1, wherein the said smart sensing based power grid data management system is applied to the smart sensing based power grid data management method as claimed in claim 1. The system comprises an intelligent sensing module, a power consumption analysis module, a simulation calculation module, and a visual warning module; The intelligent sensing module is used to collect GIS maps and meteorological parameters in the power grid coverage area, and configuration information and power consumption data of objects; The power consumption analysis module analyzes the power consumption habit of the object according to the configuration information and the power consumption data, divides abnormal areas on the GIS map, sets cumulus entities for the abnormal areas, predicts the scale and behavior of the cumulus entities, calculates an abnormal coefficient to set a cumulus area; The simulation calculation module simulates and analyzes the moving track of the cumulus entities in the cumulus area by using a physical model combined with environmental parameters, analyzes the facilities and equipment crossed by the moving track on the GIS map, and calculates the danger index of the cumulus entities in each cumulus area; The visual early warning module early warns the cumulus areas in order from large to small according to the danger index, and displays the real-time picture of the cumulus area.

Images