Comprehensive inspection drone control method and system based on multi-source data analysis

Abstract

The present invention discloses a comprehensive inspection drone control method and system based on multi-source data analysis, which relates to the field of drone control technology, including: using GIS to obtain a geographic map of the inspection area, extracting elevation and ground cover data, and constructing a three-dimensional ground model; determining the coordinates of power grid inspection points and configuring them on the geographic map, and calculating inspection contribution parameters based on historical anomaly rates and preset weights; collecting real-time wind data and predicting wind characteristic sequences; generating multiple initial inspection paths, and calculating energy consumption parameters based on wind power predictions and path lengths; and comprehensively evaluating the sum of inspection contributions and energy consumption of each path within a preset period to select the optimal path. Through multi-source data fusion and dynamic path optimization, the present invention adapts to complex environments, improves power grid inspection efficiency, and reduces energy consumption, demonstrating high innovation and practical value.

Description

Technical Field

[0001] The present invention relates to the field of drone control technology, and in particular to a comprehensive inspection drone control method and system based on multi-source data analysis. Background Art

[0002] In the field of drone inspections, especially for critical infrastructure like power grids, optimizing inspection routes is crucial for improving efficiency and reducing costs. Traditional drone inspection methods often rely on pre-set static path planning, which often exhibits poor adaptability to complex and changing environmental conditions (such as wind speed fluctuations), leading to increased energy consumption and insufficient coverage of critical areas. Furthermore, existing technologies lack the ability to prioritize inspection points, often failing to fully consider the historical incidence of abnormalities and their importance at different inspection points, making it difficult to effectively focus on high-risk areas. Furthermore, while geographic information systems (GIS), historical inspection data, and real-time environmental monitoring data have been studied in drone applications, existing methods lack a systematic framework for integrating these multi-source data and using them for dynamic route optimization. This makes it difficult to balance energy efficiency and inspection effectiveness in actual drone inspections. To address these issues, a drone control method is urgently needed that can integrate multi-source data, dynamically adapt to environmental changes, and comprehensively consider energy consumption and inspection contribution. Summary of the Invention

[0003] The purpose of the present invention is to provide a comprehensive inspection drone control method and system capable of optimizing drone paths.

[0004] The present invention discloses a comprehensive inspection drone control method based on multi-source data analysis, comprising:

[0005] Step S100: using a GIS system to obtain a geographical map of the patrol area, determine the elevation data distribution and ground cover data of the geographical map of the patrol area, and construct a three-dimensional ground protrusion on the geographical map of the patrol area;

[0006] Step S200: Determine the coordinates of the power grid inspection points and place the corresponding power grid inspection points on a geographical map of the inspection area. Obtain historical inspection records and determine the abnormality incidence rates of different power grid inspection points based on the historical inspection records. Combined with the preset inspection weights of different power grid inspection points, determine the inspection contribution parameters of different power grid inspection points when an inspection is triggered.

[0007] Step S300: acquiring real-time wind monitoring data, and determining a wind characteristic prediction sequence based on the wind characteristics of the real-time wind monitoring data;

[0008] Step S400: constructing several initial inspection routes and determining the energy consumption parameters of each initial inspection route based on the wind force characteristic prediction sequence and the route length;

[0009] In step S500 , the total inspection contribution parameter and the total energy consumption parameter of each initial inspection path within a preset time period are comprehensively considered to select the optimal inspection path.

[0010] In some embodiments of the present invention, a method for determining a wind characteristic prediction sequence based on wind characteristics of real-time wind monitoring data includes:

[0011] Step S301: construct a wind prediction model, and use the wind prediction model to analyze the wind characteristics of real-time wind monitoring data to determine a wind characteristic prediction sequence. The method of constructing the wind prediction model includes:

[0012] Step S3011: deploying wind monitoring devices in the inspection area, and mapping the location nodes of the deployed wind monitoring devices onto a geographical map of the inspection area to obtain a number of wind performance nodes;

[0013] Step S3012: Acquire historical wind monitoring data from the wind monitoring device, align the historical wind monitoring data on a time axis, and map the corresponding wind performance nodes to form historical wind performance states for the wind performance nodes. The historical wind performance states include dynamically changing wind performance frames, each of which includes a wind direction line. The direction of the wind direction line represents the monitored wind direction, and the length represents the monitored wind force.

[0014] Step S3013: Obtain historical weather forecasts, align historical wind forecasts in the historical weather forecasts with a number of historical wind monitoring data on a time axis, and map the historical wind forecasts to each wind performance node on a geographic map of the inspection area to form a historical wind forecast performance state. The historical wind forecast performance state includes dynamically changing wind forecast performance frames, each wind forecast performance frame includes a wind forecast direction line, the direction of the wind forecast direction line is used to represent the forecast wind direction, and the length is used to represent the forecast wind force;

[0015] Step S3014 , compare the historical wind performance and the historical forecast performance, and based on the comparison results, classify the corresponding historical wind forecast and historical wind monitoring data, and build a retrieval logic framework for the classified data sets to obtain a wind prediction model.

[0016] In some embodiments of the present invention, the method for classifying corresponding historical wind forecasts and historical wind monitoring data includes:

[0017] Step S3015: Compare the historical wind performance state and the historical forecast performance state for the same time period, that is, compare several consecutive wind performance frames and wind forecast performance frames for each wind performance node, determine the difference characteristics of the directivity lines of each frame, and determine the inter-state consistency parameter between the historical wind performance state and the historical forecast performance state. If the inter-state consistency parameter is greater than or equal to a preset value, it is determined that the corresponding historical wind forecast and historical wind monitoring data are equivalent.

[0018] Step S3016: classify the historical wind force forecasts and historical wind force monitoring data that have the same performance to obtain a data set corresponding to the same historical wind force, and classify the historical wind force forecasts and historical wind force monitoring data that have the same performance to obtain a data set corresponding to the different historical wind force;

[0019] Step S3017: Determine each historical wind forecast in the equivalent historical wind data set as a first search condition, and its corresponding historical wind monitoring data as a second search condition, to construct a search logic framework for the equivalent historical corresponding data set; and determine each historical wind monitoring data in the non-equivalent historical corresponding data set as a unique search condition, to construct a search logic framework for the non-equivalent historical corresponding data set;

[0020] Step S3018: construct the equivalent historical corresponding data set and the unequal historical corresponding data set into a database of the wind prediction model, and use the retrieval logic framework to retrieve the corresponding historical wind monitoring data.

[0021] In some embodiments of the present invention, a method for analyzing wind characteristics of real-time wind monitoring data using a wind prediction model includes:

[0022] Step S302 , dynamically mapping the real-time wind monitoring data to each wind performance node on the inspection area geographic map to form a real-time wind performance state of the wind performance node;

[0023] Step S303: Substitute the real-time wind monitoring data into the database of the wind prediction model. First, compare the real-time wind monitoring data corresponding to a certain time segment with the historical wind forecast. If there is a matching historical wind forecast, determine the historical wind monitoring data corresponding to the historical wind forecast, and compare the historical wind performance state corresponding to the historical wind monitoring data with the corresponding real-time wind performance state. Calculate the inter-state matching parameter. If the inter-state matching parameter is greater than or equal to a preset value, use the subsequent performance characteristics of the corresponding historical wind performance state as the predicted wind characteristics, and serialize the predicted wind characteristics to obtain a wind characteristic prediction sequence. If there is no matching historical wind forecast, compare the real-time wind performance state with the historical wind performance states of different historical wind monitoring data, and calculate the inter-state matching parameter. If the inter-state matching parameter is greater than or equal to the preset value, use the subsequent performance characteristics of the corresponding historical wind performance state as the predicted wind characteristics, and serialize the predicted wind characteristics to obtain a wind characteristic prediction sequence.

[0024] In some embodiments of the present invention, the expression for calculating the inter-state consistency parameter is:

[0025] ;

[0026] Among them, F is the inter-state consistency parameter, is the inter-state consistency parameter corresponding to the t-th time node, T is the total number of time nodes participating in the inter-state consistency comparison, L is the influence adjustment coefficient of the inter-state consistency parameter, Adjust the constant for the effect of the corresponding parameter between substates;

[0027] Among them, the expression for calculating the consistent parameters between sub-states is:

[0028] ;

[0029] in, is the line-to-line matching parameter of the wind direction line of the i-th wind performance node. If the angle between the wind direction lines is less than or equal to the preset value, and the length difference between the lines is less than or equal to the preset value, then Output the first parameter value. If the angle between the wind direction lines is less than or equal to the preset value, and the length difference between the lines is greater than the preset value, then Output the second parameter value. If the angle between the wind direction lines is greater than the preset value, and the length difference between the lines is greater than the preset value, then Output 0, where the first parameter value is greater than or equal to the second parameter value, the second parameter value is greater than 0, and n is the total number of wind performance nodes.

[0030] In some embodiments of the present invention, a method for determining energy consumption parameters of each initial inspection path based on a wind characteristic prediction sequence and a path length includes:

[0031] Step S401: constructing a virtual mapping point of the drone on the geographical map of the patrol area, and driving the virtual mapping point of the drone to move and map according to the initial patrol path;

[0032] Step S402: determining the wind characteristics corresponding to the virtual mapping point of the drone at different path nodes in the wind characteristics prediction sequence, and determining the wind resistance energy consumption parameter corresponding to the virtual mapping point of the drone based on the wind characteristics;

[0033] In step S403, the sum of the wind resistance energy consumption parameters corresponding to all path nodes is calculated, and the sum of the basic energy consumption parameters corresponding to all path nodes is calculated, and the sum of the sum of the wind resistance energy consumption parameters and the sum of the basic energy consumption parameters is calculated, and recorded as the energy consumption parameter of the initial inspection path.

[0034] In some embodiments of the present invention, a method for determining wind resistance energy consumption parameters corresponding to a drone mapping point based on wind characteristics includes:

[0035] Step S4021: Based on the wind characteristics, determine the angle between the wind direction and the initial patrol path and the wind strength, and based on the angle, determine the wind impact coefficient. The wind impact coefficient refers to the wind energy consumption conversion ratio of the wind to the energy consumption increase of the drone mapping point.

[0036] Step S4022: Calculate the product of the wind influence coefficient and the wind intensity to obtain the wind resistance energy consumption parameter.

[0037] In some embodiments of the present invention, the method of comprehensively considering the sum of inspection contribution parameters and the sum of energy consumption parameters of each initial inspection path within a preset time period includes:

[0038] Step S501: setting a number of parameter interval groups for the inspection contribution parameter sum and the energy consumption parameter sum, each parameter interval group including the inspection contribution parameter sum interval and the energy consumption parameter sum interval, and each parameter interval group is associated with an inspection contribution consideration coefficient and an energy consumption consideration coefficient;

[0039] Step S502: Determine the parameter intervals to which the sum of the inspection contribution parameters and the sum of the energy consumption belong, determine the corresponding consideration coefficients, and calculate the comprehensive consideration value of the initial inspection path based on the sum of the inspection contribution parameters and the sum of the energy consumption. Based on the comprehensive consideration value, determine the optimal inspection path.

[0040] Among them, the expression for calculating the comprehensive consideration value is:

[0041] ;

[0042] Among them, Z is the comprehensive consideration value, is the inspection contribution consideration coefficient, G is the sum of inspection contribution parameters, is the energy consumption consideration coefficient, and H is the total energy consumption.

[0043] In some embodiments of the present invention, a comprehensive inspection drone control system based on multi-source data analysis is also disclosed, including:

[0044] The first module is used to obtain a geographical map of the patrol area using a GIS system, determine the elevation data distribution and ground cover data of the geographical map of the patrol area, and construct a three-dimensional ground protrusion on the geographical map of the patrol area;

[0045] The second module is used to determine the coordinates of the power grid inspection points and configure the corresponding power grid inspection points on the geographical map of the inspection area. It obtains historical inspection records and determines the abnormality incidence rate of different power grid inspection points based on the historical inspection records. Combined with the preset inspection weights of different power grid inspection points, it determines the inspection contribution parameters of different power grid inspection points when they are triggered by inspections;

[0046] The third module is used to obtain real-time wind monitoring data and determine a wind characteristic prediction sequence based on the wind characteristics of the real-time wind monitoring data;

[0047] The fourth module is used to construct several initial inspection paths and determine the energy consumption parameters of each initial inspection path based on the wind characteristics prediction sequence and path length;

[0048] The fifth module is used to comprehensively consider the sum of the inspection contribution parameters and the sum of the energy consumption parameters of each initial inspection path within a preset time period and select the optimal inspection path.

[0049] The present invention discloses a comprehensive inspection drone control method and system based on multi-source data analysis, which relates to the field of drone control technology, including: using GIS to obtain a geographic map of the inspection area, extracting elevation and ground cover data, and constructing a three-dimensional ground model; determining the coordinates of power grid inspection points and configuring them on the geographic map, and calculating inspection contribution parameters based on historical anomaly rates and preset weights; collecting real-time wind data and predicting wind characteristic sequences; generating multiple initial inspection paths, and calculating energy consumption parameters based on wind power predictions and path lengths; and comprehensively evaluating the sum of inspection contributions and energy consumption of each path within a preset period to select the optimal path. Through multi-source data fusion and dynamic path optimization, the present invention adapts to complex environments, improves power grid inspection efficiency, and reduces energy consumption, demonstrating high innovation and practical value.

[0050] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1This is a method step diagram of a comprehensive inspection drone control method based on multi-source data analysis disclosed in an embodiment of the present invention. DETAILED DESCRIPTION

[0052] The technical solution of the present invention is further described below with reference to the accompanying drawings and embodiments.

[0053] The following will be combined with the accompanying drawings and specific embodiments to clearly and completely describe the technical solutions of the present invention. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention and cannot be understood as limiting the scope of protection of the present invention. Those skilled in the art in this field can make some non-essential improvements and adjustments based on the content of the present invention described below. In the present invention, unless otherwise clearly specified and limited, the technical terms used in the present invention should have the common meanings understood by those skilled in the art of the present invention.

[0054] Example:

[0055] The present invention discloses a comprehensive inspection drone control method based on multi-source data analysis, see Figure 1 ,include:

[0056] Step S100: using a GIS system to obtain a geographical map of the patrol area, determine the elevation data distribution and ground cover data of the geographical map of the patrol area, and construct a three-dimensional ground protrusion on the geographical map of the patrol area.

[0057] The core of this step is to use a geographic information system (GIS) to obtain a geographic map of the inspection area and extract elevation and ground cover data to construct a three-dimensional ground model. Elevation data reflects the terrain's undulating characteristics, such as the height of hillsides or the depth of valleys, while ground cover data reveals the surface coverage, such as the distribution of vegetation, the location of water bodies, or the presence of buildings. By integrating this data into the GIS system, a three-dimensional model is generated that incorporates the terrain's raised features. This 3D model provides intuitive and accurate environmental foundational information for subsequent drone path planning. For example, in mountainous environments, drones need to adjust their flight altitude based on the terrain to avoid collisions. In urban areas, raised features such as tall buildings may require drones to circumvent or increase their flight altitude. In this way, step S100 provides drone inspections with the ability to dynamically adapt to the terrain. Compared to the two-dimensional maps commonly used in traditional technologies, the use of 3D modeling in this step significantly improves the accuracy and safety of path planning.

[0058] In step S200, the coordinate position of the power grid inspection point is determined, and the corresponding power grid inspection point is configured on the geographical map of the inspection area, the historical inspection records are obtained, and the abnormal occurrence rate of different power grid inspection points is determined based on the historical inspection records. In combination with the preset inspection weights of different power grid inspection points, the inspection contribution parameters of different power grid inspection points when the inspection is triggered are determined.

[0059] This step aims to determine the precise coordinates of the power grid inspection points and map them to the inspection area geographic map generated in step S100. At the same time, by analyzing historical inspection records, the inspection contribution parameters of each inspection point are quantified. Specifically, first, historical inspection data is obtained, and the abnormality incidence rate of each inspection point, that is, the frequency of occurrence of faults or problems, is calculated; then, combined with the preset inspection weight (usually based on the importance of the inspection point, for example, the substation has a higher weight than an ordinary pole), the inspection contribution parameter is calculated. This parameter comprehensively reflects the risk level and importance of the inspection point. For example, a tower that frequently fails may have a higher abnormality incidence rate, while a power supply core node has a higher weight. After combining the two, its inspection contribution parameter will increase significantly, so it will be given priority in path planning. Compared with the uniformly distributed inspection points or static priority methods commonly used in the prior art, this step dynamically quantifies inspection needs in a data-driven manner, making inspections more targeted and efficient.

[0060] Step S300: acquiring real-time wind monitoring data, and determining a wind characteristic prediction sequence based on wind characteristics of the real-time wind monitoring data.

[0061] Step S300 obtains real-time wind monitoring data, analyzes wind characteristics such as wind direction and wind speed, and predicts future wind change trends based on this data to generate a wind characteristic prediction sequence. This sequence provides dynamic environmental input for drone path planning. Wind has an important impact on drone flight: the size of the wind speed is directly related to flight resistance. Flying with the wind can reduce energy consumption, while flying against the wind requires increased power support; changes in wind direction may require the drone to adjust its heading to maintain stability. By predicting wind trends, such as predicting that the wind speed will increase or the wind direction will change within the next hour, the drone can adapt to these changes in advance when planning the path, thereby optimizing the flight strategy. Compared with static path planning in traditional technologies that often ignore dynamic wind factors, this step introduces a real-time monitoring and prediction mechanism, allowing drones to respond to complex weather conditions more intelligently and improve flight efficiency and safety.

[0062] Step S400: construct a number of initial inspection paths, and determine the energy consumption parameters of each initial inspection path based on the wind characteristic prediction sequence and path length.

[0063] Based on the aforementioned three-dimensional model, inspection point configuration, and wind characteristics prediction, this step generates multiple initial inspection routes and evaluates the energy consumption parameters for each route. Specifically, route length is a fundamental factor affecting flight time and energy consumption, while wind characteristics (such as tailwind or headwind) further modulate actual energy consumption. For example, a shorter route may have the lowest energy consumption in calm conditions, but in strong headwinds, its energy consumption may be higher than a slightly longer route with a tailwind. By comprehensively analyzing the predicted sequence of route lengths and wind characteristics, the energy consumption parameters for each route are calculated, providing data support for subsequent optimization. This process transcends the limitations of traditional technologies that generate routes based solely on distance or fixed rules, and introduces dynamic environmental factors into the evaluation of energy consumption. For example, in areas with high wind speeds, a drone may choose to detour to take advantage of the tailwind, thereby achieving the optimal balance of energy consumption.

[0064] In step S500 , the total inspection contribution parameter and the total energy consumption parameter of each initial inspection path within a preset time period are comprehensively considered to select the optimal inspection path.

[0065] Step S500 is a key step in comprehensively evaluating the initial paths and selecting the optimal inspection path. The method involves performing a trade-off analysis between the sum of the inspection contribution parameters (reflecting inspection effectiveness) and the sum of the energy consumption parameters (reflecting energy consumption) for each path within a preset time period, ultimately determining the optimal path. The sum of the inspection contribution parameters is derived by summing the contribution parameters of each inspection point in step S200, reflecting the path's ability to cover high-risk or critical inspection points. The sum of the energy consumption parameters, calculated in step S400, reflects the path's energy consumption level. This comprehensive assessment is typically achieved through weighted summation or Pareto optimization methods. For example, a weighted ratio is set between inspection effectiveness and energy consumption, a comprehensive score is calculated for each path, and the highest-scoring path is selected. Compared to traditional techniques that focus solely on path coverage or a single energy consumption metric, this step achieves a balance between inspection effectiveness and energy consumption through multi-objective optimization. For example, when resources are limited, a drone may prioritize paths that cover critical inspection points and have lower energy consumption, thereby improving overall inspection efficiency.

[0066] In some embodiments of the present invention, a method for determining a wind characteristic prediction sequence based on wind characteristics of real-time wind monitoring data includes:

[0067] Step S301: construct a wind prediction model, and use the wind prediction model to analyze the wind characteristics of real-time wind monitoring data to determine a wind characteristic prediction sequence. The method of constructing the wind prediction model includes:

[0068] Step S3011 : deploying wind monitoring devices in the inspection area, and mapping the location nodes of the deployed wind monitoring devices onto a geographical map of the inspection area to obtain a number of wind performance nodes.

[0069] The principle of arranging wind monitoring devices and mapping nodes is to reasonably set up wind monitoring equipment in the patrol area, collect wind data at different locations in real time, and accurately map the locations of these devices to the geographical map of the patrol area to form a number of "wind performance nodes". These nodes provide a geographical reference for subsequent spatial wind analysis. The layout of wind monitoring devices needs to take into account factors such as the terrain and building distribution of the area to ensure the representativeness of the data. Compared with the single or a small number of fixed monitoring points commonly used in traditional technologies, this step can more comprehensively capture the wind distribution characteristics in the area through a multi-point distributed layout. For example, in complex terrain such as mountainous areas or urban environments, the wind may change significantly due to local obstacles, and multi-point monitoring can effectively reflect these differences. This method breaks through the spatial limitations of traditional wind monitoring and provides a more accurate environmental basis for the prediction of wind characteristics based on real-time data.

[0070] Step S3012: Acquire historical wind monitoring data from the wind monitoring device, align the historical wind monitoring data on the time axis, and map the corresponding wind performance nodes to form a historical wind performance state of the wind performance node. The historical wind performance state includes dynamically changing wind performance frames, and each wind performance frame includes a wind direction line. The direction of the wind direction line is used to represent the monitored wind direction, and the length is used to represent the monitored wind force.

[0071] The principle of forming a historical wind performance state is to use the historical wind data collected by the wind monitoring device, align it according to the time axis, and map it to the corresponding wind performance node, so as to construct a dynamic "historical wind performance state". This performance state is composed of multiple "wind performance frames", each frame contains a "wind direction line", whose direction indicates the monitored wind direction and its length reflects the wind force. This method converts the scattered wind data into an intuitive spatiotemporal sequence, which is convenient for analyzing the change pattern of wind force over time and space. In traditional technology, wind data is usually stored in the form of tables or numerical values, which makes it difficult to directly reflect dynamic characteristics. This step significantly improves the interpretability of the data through frame serialization and visualization design. For example, when analyzing the monsoon changes in a certain area, the periodic patterns of wind direction and wind force can be clearly identified through historical performance states, providing a reliable basis for subsequent predictions.

[0072] Step S3013, obtain historical weather forecasts, and align the historical wind forecasts in the historical weather forecasts and several historical wind monitoring data on the time axis, and map the historical wind forecasts to each wind performance node on the patrol area geographic map to form a historical wind forecast performance state. The historical wind forecast performance state includes dynamically changing wind forecast performance frames, and each wind forecast performance frame includes a wind forecast direction line. The direction of the wind forecast direction line is used to represent the forecast wind direction, and the length is used to represent the forecast wind force.

[0073] The principle of forming a historical wind forecast representation is to extract the wind forecast data from the historical meteorological forecast, align it with the historical wind monitoring data on the time axis, and map it to each wind representation node in the patrol area to form a "historical wind forecast representation". Similar to the historical wind representation, this representation is composed of "wind forecast representation frames", each frame contains a "wind forecast pointing line", whose direction and length respectively represent the predicted wind direction and wind force. Through this design, the forecast data is integrated into the same framework as the monitoring data, which is convenient for subsequent comparison and analysis. In traditional technologies, wind forecast data is mostly used for post-verification or independent analysis, while this step converts it into a dynamic representation, giving full play to the role of forecast data in pattern recognition. For example, if a historical forecast accurately predicts a strong wind event, the forecast pattern can be recorded and used to improve the accuracy of future forecasts.

[0074] In some embodiments of the present invention, the method for classifying corresponding historical wind forecasts and historical wind monitoring data includes:

[0075] Step S3015 compares the historical wind performance state and the historical forecast performance state for the same time period, that is, compares several consecutive wind performance frames and wind forecast performance frames of each wind performance node, determines the difference characteristics of the directional lines of each frame, and determines the inter-state consistency parameter between the historical wind performance state and the historical forecast performance state. If the inter-state consistency parameter is greater than or equal to a preset value, it is determined that the corresponding historical wind forecast and historical wind monitoring data are equivalent.

[0076] The principle of comparing historical performance states with forecast performance states is to quantify the degree of similarity between historical wind monitoring data and forecast data within the same time period by calculating the "inter-state consistency parameter." Specifically, for each wind performance node, consecutive wind performance frames and wind forecast performance frames are compared to analyze the differences in the direction and length of the pointing lines. If the difference is small (i.e., the inter-state consistency parameter is greater than or equal to the preset value), the historical forecast and monitoring data for that section are considered to be "equivalent." This step uses dynamic comparison to identify historically accurate forecast scenarios, providing data support for the prediction model. Traditionally, forecast accuracy assessments rely heavily on statistical indicators such as mean square error. This step, however, uses a pattern matching method for frame sequences to more meticulously capture the spatiotemporal characteristics of wind changes. For example, in a storm, if the forecast pointing line is highly consistent with the actual pointing line, this pattern can be used as a reference for predicting similar events in the future.

[0077] Step S3016: classify the historical wind force forecasts and historical wind force monitoring data with equivalent performance to obtain a corresponding data set of equivalent historical wind force, and classify the historical wind force forecasts and historical wind force monitoring data with unequal performance to obtain a corresponding data set of unequal historical wind force.

[0078] The principle of data classification is to divide historical wind forecasts and monitoring data into two categories based on the comparison results: "equivalent historical wind corresponding data sets" (forecast and monitoring performance are equivalent) and "not equivalent historical wind corresponding data sets" (forecast and monitoring performance are not equivalent). This classification lays the foundation for subsequent retrieval and prediction. The equivalent data sets contain historically accurate forecast scenarios and can be used as a priority reference for prediction, while the not equivalent data sets provide supplementary information for dealing with situations where forecast deviations are large. In traditional technologies, historical data are often processed uniformly and lack targeted classification. This step improves the accuracy of data utilization by distinguishing between data with equivalent and unequal performance. For example, when predicting strong winds, giving priority to using patterns in the equivalent data sets can effectively improve the reliability of the prediction results.

[0079] Step S3017: identify each historical wind forecast in the equivalent historical wind corresponding data set as the first search condition, and its corresponding historical wind monitoring data as the second search condition, and build a search logic framework for the equivalent historical corresponding data set; identify each historical wind monitoring data in the unequal historical corresponding data set as the only search condition, and build a search logic framework for the unequal historical corresponding data set.

[0080] The principle of constructing a retrieval logic framework is to design different retrieval conditions and logic for the classified data sets to achieve rapid matching of historical patterns. For data sets corresponding to equivalent historical wind power, the historical wind power forecast is used as the "first retrieval condition" and the corresponding monitoring data is used as the "second retrieval condition"; for data sets corresponding to unequal historical wind power, only the historical monitoring data is used as the "only retrieval condition". This design makes full use of the predictive value of the forecast data while retaining the authenticity of the monitoring data, ensuring the accuracy and applicability of the retrieval results. In traditional technologies, data retrieval is mostly based on simple time or location conditions, while this step achieves more efficient pattern matching through multi-condition combination retrieval. For example, in real-time prediction, if the current wind power characteristics are similar to a historical forecast, the corresponding monitoring data can be directly called as the basis for prediction.

[0081] Step S3018: construct the equivalent historical corresponding data set and the unequal historical corresponding data set into a database of the wind prediction model, and use the retrieval logic framework to retrieve the corresponding historical wind monitoring data.

[0082] The principle of establishing a model database is to integrate the classified equivalent and unequal data sets into a database of wind prediction models, and through a retrieval logic framework, extract historical patterns that match the real-time wind monitoring data from the database to predict future wind characteristics. This database provides support for dynamic wind prediction through pattern matching of historical data. Compared with traditional wind prediction methods based on statistical models or time series analysis, this step combines pattern retrieval with historical data, and can respond quickly to complex or sudden scenarios. For example, during drone inspections, if a sudden change in wind speed is detected in real time, the model can quickly retrieve similar historical scenarios and generate a wind characteristic prediction sequence, thereby optimizing the flight path.

[0083] In some embodiments of the present invention, a method for analyzing wind characteristics of real-time wind monitoring data using a wind prediction model includes:

[0084] Step S302 : Dynamically map the real-time wind monitoring data to each wind performance node on the inspection area geographic map to form a real-time wind performance state of the wind performance node.

[0085] Step S303: Substitute the real-time wind monitoring data into the database of the wind prediction model. First, compare the real-time wind monitoring data corresponding to a certain time segment with the historical wind forecast. If there is a matching historical wind forecast, determine the historical wind monitoring data corresponding to the historical wind forecast, and compare the historical wind performance state corresponding to the historical wind monitoring data with the corresponding real-time wind performance state. Calculate the inter-state matching parameter. If the inter-state matching parameter is greater than or equal to a preset value, use the subsequent performance characteristics of the corresponding historical wind performance state as the predicted wind characteristics, and serialize the predicted wind characteristics to obtain a wind characteristic prediction sequence. If there is no matching historical wind forecast, compare the real-time wind performance state with the historical wind performance states of different historical wind monitoring data, and calculate the inter-state matching parameter. If the inter-state matching parameter is greater than or equal to the preset value, use the subsequent performance characteristics of the corresponding historical wind performance state as the predicted wind characteristics, and serialize the predicted wind characteristics to obtain a wind characteristic prediction sequence.

[0086] In some embodiments of the present invention, the expression for calculating the inter-state consistency parameter is:

[0087] .

[0088] Among them, F is the inter-state consistency parameter, is the inter-state consistency parameter corresponding to the t-th time node, T is the total number of time nodes participating in the inter-state consistency comparison, L is the influence adjustment coefficient of the inter-state consistency parameter, Adjust the constant for the effect of the corresponding parameter between substates;

[0089] Among them, the expression for calculating the consistent parameters between sub-states is:

[0090] .

[0091] in, is the line-to-line matching parameter of the wind direction line of the i-th wind performance node. If the angle between the wind direction lines is less than or equal to the preset value, and the length difference between the lines is less than or equal to the preset value, then Output the first parameter value. If the angle between the wind direction lines is less than or equal to the preset value, and the length difference between the lines is greater than the preset value, then Output the second parameter value. If the angle between the wind direction lines is greater than the preset value, and the length difference between the lines is greater than the preset value, then Output 0, where the first parameter value is greater than or equal to the second parameter value, the second parameter value is greater than 0, and n is the total number of wind performance nodes.

[0092] In some embodiments of the present invention, a method for determining energy consumption parameters of each initial inspection path based on a wind characteristic prediction sequence and path length includes:

[0093] In step S401 , a virtual mapping point of the drone is constructed on a geographical map of the patrol area, and the virtual mapping point of the drone is driven to move and map according to an initial patrol path.

[0094] Step S402: determining the wind characteristics corresponding to the virtual mapping point of the drone at different path nodes in the wind characteristics prediction sequence, and determining the wind resistance energy consumption parameter corresponding to the virtual mapping point of the drone based on the wind characteristics.

[0095] In step S403, the sum of the wind resistance energy consumption parameters corresponding to all path nodes is calculated, and the sum of the basic energy consumption parameters corresponding to all path nodes is calculated, and the sum of the sum of the wind resistance energy consumption parameters and the sum of the basic energy consumption parameters is calculated, and recorded as the energy consumption parameter of the initial inspection path.

[0096] In some embodiments of the present invention, a method for determining wind resistance energy consumption parameters corresponding to a drone mapping point based on wind characteristics includes:

[0097] Step S4021: Based on the wind characteristics, determine the angle between the wind direction and the initial patrol path and the wind intensity, and based on the angle, determine the wind impact coefficient. The wind impact coefficient refers to the wind energy consumption conversion ratio of the wind to the energy consumption increase of the drone mapping point.

[0098] Step S4022: Calculate the product of the wind influence coefficient and the wind intensity to obtain the wind resistance energy consumption parameter.

[0099] In some embodiments of the present invention, the method of comprehensively considering the sum of inspection contribution parameters and the sum of energy consumption parameters of each initial inspection path within a preset time period includes:

[0100] Step S501, setting several parameter interval groups for the inspection contribution parameter sum and the energy consumption parameter sum, each parameter interval group includes the inspection contribution parameter sum interval and the energy consumption parameter sum interval, and each parameter interval group is associated with the inspection contribution consideration coefficient and the energy consumption consideration coefficient.

[0101] Step S502: determine the parameter intervals to which the sum of the inspection contribution parameters and the sum of the energy consumption belong, and determine the corresponding consideration coefficients. Combined with the sum of the inspection contribution parameters and the sum of the energy consumption, the comprehensive consideration value of the initial inspection path is calculated, and based on the comprehensive consideration value, the selected optimal inspection path is determined.

[0102] Among them, the expression for calculating the comprehensive consideration value is:

[0103] .

[0104] Among them, Z is the comprehensive consideration value, is the inspection contribution consideration coefficient, G is the sum of inspection contribution parameters, is the energy consumption consideration coefficient, and H is the total energy consumption.

[0105] In some embodiments of the present invention, a comprehensive inspection drone control system based on multi-source data analysis is also disclosed, including:

[0106] The first module is used to obtain the patrol area geographic map using the GIS system, determine the elevation data distribution and ground cover data of the patrol area geographic map, and construct a three-dimensional ground protrusion on the patrol area geographic map.

[0107] The second module is used to determine the coordinate position of the power grid inspection point, and configure the corresponding power grid inspection point on the geographical map of the inspection area, obtain historical inspection records, and determine the abnormal occurrence rate of different power grid inspection points based on the historical inspection records. Combined with the preset inspection weights of different power grid inspection points, the inspection contribution parameters of different power grid inspection points when they are triggered by inspection are determined.

[0108] The third module is used to obtain real-time wind monitoring data and determine a wind characteristic prediction sequence based on the wind characteristics of the real-time wind monitoring data.

[0109] The fourth module is used to construct several initial inspection paths and determine the energy consumption parameters of each initial inspection path based on the wind characteristic prediction sequence and path length.

[0110] The fifth module is used to comprehensively consider the sum of the inspection contribution parameters and the sum of the energy consumption parameters of each initial inspection path within a preset time period and select the optimal inspection path.

[0111] The present invention discloses a comprehensive inspection drone control method and system based on multi-source data analysis, which relates to the field of drone control technology, including: using GIS to obtain a geographic map of the inspection area, extracting elevation and ground cover data, and constructing a three-dimensional ground model; determining the coordinates of power grid inspection points and configuring them on the geographic map, and calculating inspection contribution parameters based on historical anomaly rates and preset weights; collecting real-time wind data and predicting wind characteristic sequences; generating multiple initial inspection paths, and calculating energy consumption parameters based on wind power predictions and path lengths; and comprehensively evaluating the sum of inspection contributions and energy consumption of each path within a preset period to select the optimal path. Through multi-source data fusion and dynamic path optimization, the present invention adapts to complex environments, improves power grid inspection efficiency, and reduces energy consumption, demonstrating high innovation and practical value.

[0112] Through the above description of the embodiments, those skilled in the art will clearly understand that the present invention can be implemented via hardware or via software combined with a necessary general-purpose hardware platform. Based on this understanding, the technical solution of the present invention can be embodied in the form of a software product. This software product can be stored on a non-volatile storage medium (such as a CD-ROM, USB flash drive, or external hard drive) and includes instructions for enabling a computer device (such as a personal computer, server, or network device) to execute the methods described in various implementation scenarios of the present invention.

[0113] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit the same. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that they can still modify or replace the technical solutions of the present invention with equivalents, and these modifications or equivalent replacements cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A comprehensive inspection drone control method based on multi-source data analysis, characterized in that: include: Step S100: using a GIS system to obtain a geographical map of the patrol area, determine the elevation data distribution and ground cover data of the geographical map of the patrol area, and construct a three-dimensional ground protrusion on the geographical map of the patrol area; Step S200: Determine the coordinates of the power grid inspection points and place the corresponding power grid inspection points on a geographical map of the inspection area. Obtain historical inspection records and determine the abnormality incidence rates of different power grid inspection points based on the historical inspection records. Combined with the preset inspection weights of different power grid inspection points, determine the inspection contribution parameters of different power grid inspection points when an inspection is triggered. Step S300: acquiring real-time wind monitoring data, and determining a wind characteristic prediction sequence based on the wind characteristics of the real-time wind monitoring data; Step S400: constructing several initial inspection routes and determining the energy consumption parameters of each initial inspection route based on the wind force characteristic prediction sequence and the route length; Step S500 , comprehensively considering the sum of inspection contribution parameters and energy consumption parameters of each initial inspection path within a preset time period, and selecting the optimal inspection path; The method for determining the wind characteristic prediction sequence based on the wind characteristics of real-time wind monitoring data includes: Step S301: construct a wind prediction model, and use the wind prediction model to analyze the wind characteristics of real-time wind monitoring data to determine a wind characteristic prediction sequence. The method of constructing the wind prediction model includes: Step S3011: deploying wind monitoring devices in the inspection area, and mapping the location nodes of the deployed wind monitoring devices onto a geographical map of the inspection area to obtain a number of wind performance nodes; Step S3012: Acquire historical wind monitoring data from the wind monitoring device, align the historical wind monitoring data on a time axis, and map the corresponding wind performance nodes to form historical wind performance states for the wind performance nodes. The historical wind performance states include dynamically changing wind performance frames, each of which includes a wind direction line. The direction of the wind direction line represents the monitored wind direction, and the length represents the monitored wind force. Step S3013: Obtain historical weather forecasts, align historical wind forecasts in the historical weather forecasts with a number of historical wind monitoring data on a time axis, and map the historical wind forecasts to each wind performance node on a geographic map of the inspection area to form a historical wind forecast performance state. The historical wind forecast performance state includes dynamically changing wind forecast performance frames, each wind forecast performance frame includes a wind forecast direction line, the direction of the wind forecast direction line is used to represent the forecast wind direction, and the length is used to represent the forecast wind force; Step S3014: comparing historical wind performance and historical forecast performance, and based on the comparison results, classifying the corresponding historical wind forecast and historical wind monitoring data, and constructing a retrieval logic framework for the classified data sets to obtain a wind prediction model; Methods for classifying the corresponding historical wind forecasts and historical wind monitoring data include: Step S3015: Compare the historical wind performance state and the historical forecast performance state for the same time period, that is, compare several consecutive wind performance frames and wind forecast performance frames for each wind performance node, determine the difference characteristics of the directivity lines of each frame, and determine the inter-state consistency parameter between the historical wind performance state and the historical forecast performance state. If the inter-state consistency parameter is greater than or equal to a preset value, it is determined that the corresponding historical wind forecast and historical wind monitoring data are equivalent. Step S3016: classify the historical wind force forecasts and historical wind force monitoring data that have the same performance to obtain a data set corresponding to the same historical wind force, and classify the historical wind force forecasts and historical wind force monitoring data that have the same performance to obtain a data set corresponding to the different historical wind force; Step S3017: Determine each historical wind forecast in the equivalent historical wind data set as a first search condition, and its corresponding historical wind monitoring data as a second search condition, to construct a search logic framework for the equivalent historical corresponding data set; and determine each historical wind monitoring data in the non-equivalent historical corresponding data set as a unique search condition, to construct a search logic framework for the non-equivalent historical corresponding data set; Step S3018: construct the equivalent historical corresponding data set and the unequal historical corresponding data set into a database of the wind prediction model, and use the retrieval logic framework to retrieve the corresponding historical wind monitoring data.

2. The comprehensive inspection drone control method based on multi-source data analysis according to claim 1 is characterized in that: Methods for analyzing wind characteristics of real-time wind monitoring data using a wind prediction model include: Step S302 , dynamically mapping the real-time wind monitoring data to each wind performance node on the inspection area geographic map to form a real-time wind performance state of the wind performance node; Step S303: Substitute the real-time wind monitoring data into the database of the wind prediction model. First, compare the real-time wind monitoring data corresponding to a certain time segment with the historical wind forecast. If there is a matching historical wind forecast, determine the historical wind monitoring data corresponding to the historical wind forecast, and compare the historical wind performance state corresponding to the historical wind monitoring data with the corresponding real-time wind performance state. Calculate the inter-state matching parameter. If the inter-state matching parameter is greater than or equal to a preset value, use the subsequent performance characteristics of the corresponding historical wind performance state as the predicted wind characteristics, and serialize the predicted wind characteristics to obtain a wind characteristic prediction sequence. If there is no matching historical wind forecast, compare the real-time wind performance state with the historical wind performance states of different historical wind monitoring data, and calculate the inter-state matching parameter. If the inter-state matching parameter is greater than or equal to the preset value, use the subsequent performance characteristics of the corresponding historical wind performance state as the predicted wind characteristics, and serialize the predicted wind characteristics to obtain a wind characteristic prediction sequence.

3. The comprehensive inspection drone control method based on multi-source data analysis according to claim 1 is characterized in that: The expression for calculating the inter-state consistent parameter is: ； Among them, F is the inter-state consistency parameter, is the inter-state consistency parameter corresponding to the t-th time node, T is the total number of time nodes participating in the inter-state consistency comparison, L is the influence adjustment coefficient of the inter-state consistency parameter, and b1 is the influence adjustment constant of the inter-state consistency parameter; Among them, the expression for calculating the consistent parameters between sub-states is: ； in, is the line-to-line matching parameter of the wind direction line of the i-th wind performance node. If the angle between the wind direction lines is less than or equal to the preset value, and the length difference between the lines is less than or equal to the preset value, then Output the first parameter value. If the angle between the wind direction lines is less than or equal to the preset value, and the length difference between the lines is greater than the preset value, then Output the second parameter value. If the angle between the wind direction lines is greater than the preset value, and the length difference between the lines is greater than the preset value, then Output 0, where the first parameter value is greater than or equal to the second parameter value, the second parameter value is greater than 0, and n is the total number of wind performance nodes.

4. The comprehensive inspection drone control method based on multi-source data analysis according to claim 1 is characterized in that: The method for determining the energy consumption parameters of each initial inspection path based on the wind characteristic prediction sequence and path length includes: Step S401: constructing a virtual mapping point of the drone on the geographical map of the patrol area, and driving the virtual mapping point of the drone to move and map according to the initial patrol path; Step S402: determining the wind characteristics corresponding to the virtual mapping point of the drone at different path nodes in the wind characteristics prediction sequence, and determining the wind resistance energy consumption parameter corresponding to the virtual mapping point of the drone based on the wind characteristics; In step S403, the sum of the wind resistance energy consumption parameters corresponding to all path nodes is calculated, and the sum of the basic energy consumption parameters corresponding to all path nodes is calculated, and the sum of the sum of the wind resistance energy consumption parameters and the sum of the basic energy consumption parameters is calculated, and recorded as the energy consumption parameter of the initial inspection path.

5. The comprehensive inspection drone control method based on multi-source data analysis according to claim 4 is characterized in that: Based on wind characteristics, the method for determining the wind resistance energy consumption parameters corresponding to the UAV mapping point includes: Step S4021: Based on the wind characteristics, determine the angle between the wind direction and the initial patrol path and the wind strength, and based on the angle, determine the wind impact coefficient. The wind impact coefficient refers to the wind energy consumption conversion ratio of the wind to the energy consumption increase of the drone mapping point. Step S4022: Calculate the product of the wind influence coefficient and the wind intensity to obtain the wind resistance energy consumption parameter.

6. The comprehensive inspection drone control method based on multi-source data analysis according to claim 1 is characterized in that: The method for comprehensively considering the sum of inspection contribution parameters and the sum of energy consumption parameters of each initial inspection path within a preset time period includes: Step S501: setting a number of parameter interval groups for the inspection contribution parameter sum and the energy consumption parameter sum, each parameter interval group including the inspection contribution parameter sum interval and the energy consumption parameter sum interval, and each parameter interval group is associated with an inspection contribution consideration coefficient and an energy consumption consideration coefficient; Step S502: Determine the parameter intervals to which the sum of the inspection contribution parameters and the sum of the energy consumption belong, determine the corresponding consideration coefficients, and calculate the comprehensive consideration value of the initial inspection path based on the sum of the inspection contribution parameters and the sum of the energy consumption. Based on the comprehensive consideration value, determine the optimal inspection path. Among them, the expression for calculating the comprehensive consideration value is: ； Among them, Z is the comprehensive consideration value, K1 is the inspection contribution consideration coefficient, G is the sum of inspection contribution parameters, K2 is the energy consumption consideration coefficient, and H is the total energy consumption.

7. Comprehensive inspection drone control system based on multi-source data analysis, characterized by: A comprehensive inspection drone control method for executing any one of claims 1-6, comprising: The first module is used to obtain a geographical map of the patrol area using a GIS system, determine the elevation data distribution and ground cover data of the geographical map of the patrol area, and construct a three-dimensional ground protrusion on the geographical map of the patrol area; The second module is used to determine the coordinates of the power grid inspection points and configure the corresponding power grid inspection points on the geographical map of the inspection area. It obtains historical inspection records and determines the abnormality incidence rate of different power grid inspection points based on the historical inspection records. Combined with the preset inspection weights of different power grid inspection points, it determines the inspection contribution parameters of different power grid inspection points when they are triggered by inspections; The third module is used to obtain real-time wind monitoring data and determine a wind characteristic prediction sequence based on the wind characteristics of the real-time wind monitoring data; The fourth module is used to construct several initial inspection paths and determine the energy consumption parameters of each initial inspection path based on the wind characteristics prediction sequence and path length; The fifth module is used to comprehensively consider the sum of the inspection contribution parameters and the sum of the energy consumption parameters of each initial inspection path within a preset time period and select the optimal inspection path.

Images