Substation inspection method and device based on unmanned aerial vehicle, and electronic equipment

Abstract

The invention discloses a transformer substation inspection method and device based on an unmanned aerial vehicle and electronic equipment. The method comprises the following steps: acquiring topographic data of a transformer substation based on an unmanned aerial vehicle and ground laser scanning equipment, and constructing a three-dimensional model of the transformer substation based on the topographic data; determining an initial path of the unmanned aerial vehicle in the transformer substation based on the transformer substation three-dimensional model; acquiring meteorological data of an area where the transformer substation is located and equipment states of a plurality of pieces of power equipment included in the transformer substation; and based on the meteorological data and the equipment state, optimizing the initial path to obtain a target path, and controlling the unmanned aerial vehicle to inspect the transformer substation along the target path. The technical problems that the manual participation degree is high, the inspection efficiency is low and the safety is difficult to guarantee in the substation inspection are solved.

Description

Technical Field

[0001] The present invention relates to the field of smart grids, and in particular, to a substation inspection method, device, and electronic device based on an unmanned aerial vehicle (UAV). Background Art

[0002] With the rapid development of the power system, as a key node for power transmission and distribution, the safe and stable operation of substations is of crucial importance. The traditional manual inspection method has problems such as long inspection cycles, low efficiency, and high safety hazards, and it is difficult to meet the intelligent and automated requirements of modern power grid operation and maintenance management. At present, although UAV inspection technology has been gradually applied to substations, most of the existing UAV inspection schemes only stay at the simple flight and shooting stage, still relying on manual rechecks, on-site inspections, etc. The operation of the UAV depends on the pilot. For a large number of substations across the country, it is impossible to meet the requirements of a large number of pilots. Moreover, as a high-tech product, the UAV has certain requirements for the learning ability of the pilot. Even if UAVs are distributed to substations, it is very easy to occur that the use of UAVs is abandoned due to insufficient pilots or insufficient technical capabilities of the pilots.

[0003] In view of the problems of high manual participation, low inspection efficiency, and difficult security guarantee in the above-mentioned substation inspection, no effective solution has been proposed yet. Summary of the Invention

[0004] Embodiments of the present invention provide a substation inspection method, device, and electronic device based on an unmanned aerial vehicle (UAV) to at least solve the technical problems of high manual participation, low inspection efficiency, and difficult security guarantee in substation inspection.

[0005] According to one aspect of the embodiments of the present invention, a substation inspection method based on an unmanned aerial vehicle (UAV) is provided, including: collecting topographic data of a substation based on the UAV and a ground laser scanning device, and constructing a three-dimensional model of the substation based on the topographic data; determining an initial path of the UAV in the substation based on the three-dimensional model of the substation; obtaining meteorological data of the area where the substation is located and the device status of a plurality of power devices included in the substation; optimizing the initial path based on the meteorological data and the device status to obtain a target path, and controlling the UAV to inspect the substation along the target path.

[0006] According to another aspect of the embodiments of the present invention, there is also provided a substation inspection device based on an unmanned aerial vehicle (UAV), including: a three-dimensional model construction module, configured to collect topographic data of a substation based on the UAV and a ground laser scanning device, and construct a three-dimensional model of the substation based on the topographic data; an initial path acquisition module, configured to determine an initial path of the UAV in the substation based on the three-dimensional model of the substation; a data acquisition module, configured to acquire meteorological data of the area where the substation is located and the device states of a plurality of power devices included in the substation; and a target path acquisition module, configured to optimize the initial path based on the meteorological data and the device states to obtain a target path, and control the UAV to inspect the substation along the target path.

[0007] According to another aspect of the embodiments of the present invention, there is also provided a non-volatile storage medium storing multiple instructions adapted to be loaded and executed by a processor to perform any one of the UAV-based substation inspection methods.

[0008] According to another aspect of the embodiments of the present invention, there is also provided an electronic device including one or more processors and a memory, where the memory is configured to store one or more programs, and when the one or more programs are executed by the one or more processors, the one or more processors are caused to implement any one of the UAV-based substation inspection methods.

[0009] In the embodiments of the present invention, by collecting topographic data of a substation based on a UAV and a ground laser scanning device, constructing a three-dimensional model of the substation based on the topographic data, determining an initial path of the UAV in the substation based on the three-dimensional model of the substation, acquiring meteorological data of the area where the substation is located and the device states of a plurality of power devices included in the substation, optimizing the initial path based on the meteorological data and the device states to obtain a target path, and controlling the UAV to inspect the substation along the target path, the purpose of collaborative collection of topographic data by the UAV and the ground laser scanning device, construction and dynamic update of the three-dimensional model of the substation, and intelligent adjustment of the UAV inspection path in combination with real-time meteorological data and power device states is achieved, thereby realizing the technical effects of improving inspection efficiency, reducing manual intervention, enhancing inspection safety and accuracy, and further solving the technical problems of high manual participation, low inspection efficiency and difficult guarantee of safety in substation inspection. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of this application. The illustrative embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation to the present invention. In the drawings:

[0011] Figure 1 is a flowchart of a substation inspection method based on an unmanned aerial vehicle according to an embodiment of the present invention;

[0012] Figure 2 is a flowchart of an alternative substation inspection method based on an unmanned aerial vehicle according to an embodiment of the present invention;

[0013] Figure 3 is a flowchart of another alternative substation inspection method based on an unmanned aerial vehicle according to an embodiment of the present invention;

[0014] Figure 4 is a schematic diagram of an alternative unmanned substation inspection device according to an embodiment of the present invention;

[0015] Figure 5 is a schematic diagram of a substation inspection device based on an unmanned aerial vehicle according to an embodiment of the present invention. Detailed implementation manners

[0016] In order to enable those skilled in the art to better understand the solution of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0017] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and do not have to be used to describe a specific order or sequence. It should be understood that such data can be interchanged under appropriate circumstances so that the embodiments of the present invention described here can be implemented in an order different from those illustrated or described here. In addition, the terms "comprising" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device that includes a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0018] First, for the convenience of understanding the embodiments of the present invention, some terms or nouns involved in the present invention will be explained below:

[0019] The A* algorithm is a heuristic search algorithm used to find the lowest-cost path from a starting point to an ending point in a graph. The A* algorithm combines the shortest-path search of the Dijkstra algorithm and a heuristic search method. It estimates the total cost of reaching the target node from the current node and the cost from the current node to the target node. The A* algorithm uses a priority queue to select the next expansion point, ensuring that nodes closer to the target are expanded first.

[0020] The Dijkstra algorithm is an algorithm used to find the shortest paths from a starting point to all other nodes in a directed graph. The algorithm starts from the starting point and gradually expands to other nodes in the graph. By continuously updating the estimated values of the shortest paths, it finally finds the shortest paths for all nodes. The Dijkstra algorithm is applicable to graphs without negative-weight edges and can guarantee that the found paths are the shortest. It does not use any heuristic information but relies on exact path cost information for decision-making.

[0021] According to an embodiment of the present invention, there is provided an embodiment of a method for substation inspection based on an unmanned aerial vehicle (UAV). It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. And although the logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in a different order than here.

[0022] Figure 1 is a flowchart of the method for substation inspection based on an unmanned aerial vehicle according to an embodiment of the present invention, as Figure 1 shown, the method includes the following steps:

[0023] Step S102, based on the unmanned aerial vehicle and the ground laser scanning device, collect the topographic data of the substation, and based on the topographic data, construct a three-dimensional model of the substation;

[0024] Optionally, the unmanned aerial vehicle and the ground laser scanning device are used to cooperate to obtain high-precision topographic data of the substation. Specifically, the unmanned aerial vehicle is equipped with a high-definition camera and a light detection and ranging (LiDAR), and takes pictures and collects depth information of the substation from the air. At the same time, the ground laser scanner scans the surface around the substation in detail to obtain the elevation data and texture information of the ground. These data together constitute the three-dimensional topographic information of the substation and provide basic data materials for the construction of the three-dimensional model of the substation.

[0025] In an optional embodiment, constructing a three-dimensional model of the substation based on the topographic data includes: performing filtering processing on the topographic data to obtain the processed data, where the topographic data is point cloud data, and the filtering processing is used to remove the noise and redundancy of the point cloud data; performing surface reconstruction based on the processed data to convert the processed data into a three-dimensional model of the substation.

[0026] Optionally, the process of constructing a 3D substation model involves preprocessing the originally collected terrain data, especially filtering, to remove noise and redundant parts in the data. Subsequently, the processed data is used for 3D surface reconstruction to form a high-precision 3D substation model. Specifically, the terrain data of the substation collected by drones and terrestrial laser scanning devices exists in the form of point clouds. A point cloud is a collection composed of a large number of spatial coordinate points, which can accurately reflect the terrain features and structural details of the substation. The original point cloud data often contains noise and redundant information, which may stem from measurement errors, equipment errors, environmental factors, etc. during the scanning process. To improve the accuracy of the model and reduce the complexity of subsequent processing, it is necessary to filter the point cloud data. Filtering algorithms in the Point Cloud Library (PCL), such as statistical outlier filtering, voxel grid filtering, etc., can be used but are not limited to these to reduce the data density and remove abnormal points, thereby making the point cloud data more accurate and reliable. Through filtering, the noise and redundant information in the point cloud data are effectively removed, and the data quality is improved. To construct a 3D model of the substation, the processed point cloud data needs to be converted into a continuous surface model. Poisson surface reconstruction algorithm can be used but is not limited to this. This is an efficient point cloud surface reconstruction method that can generate a 3D model with high detail and accuracy. Through this algorithm, the point cloud data is converted into a 3D grid model of the substation, and the model contains the geometric features and structural information of the interior and surrounding environment of the substation. Through surface reconstruction, a 3D model of the substation is formed. This model is the basis for subsequent path planning and inspection task execution.

[0027] In the above way, a high-precision and dynamically updatable 3D substation model can be constructed based on the terrain data collected by drones and terrestrial laser scanning devices. This model can not only clearly reflect the physical structure of the substation but also be adjusted according to real-time data, providing accurate environmental information for the automatic inspection of drones and ensuring that the drones can complete the inspection tasks safely and efficiently.

[0028] Optionally, point cloud processing can be used to preprocess the terrain data of the substation to obtain the processed point cloud data; a grid generation algorithm is used to construct a high-precision 3D (i.e., 3D) model of the substation using the point cloud data, and the 3D model of the substation is dynamically updated through real-time data fusion and manual model optimization, and important areas, importance levels, and urgency levels are set. The filtering algorithm in the PCL library is used to filter the point cloud data to reduce the data density, and the Poisson surface reconstruction algorithm is used to convert the filtered point cloud data into a 3D model of the substation. Specifically:

[0029] Step S1021, 3D modeling of the substation, specifically including:

[0030] Point cloud processing: Use the PCL (Point Cloud Library) to process LiDAR data, generate point cloud data, and perform preprocessing operations such as filtering, denoising, and registration.

[0031] Filtering: P filtered = {p ∈ P ∣ distance(p, N(p)) < threshold}

[0032] where P filtered is the processed data, P is the original point cloud data, N(p) is the neighborhood of point p, and threshold is the filtering threshold.

[0033] Mesh generation: Use the Poisson Surface Reconstruction algorithm to convert the point cloud data into a three-dimensional mesh model.

[0034] Poisson equation: Δφ = ρ

[0035] where φ is the indicator function, ρ is the density function of the point cloud data, and the three-dimensional mesh model is obtained by solving the Poisson equation.

[0036] Step S1022, dynamic update of the three-dimensional model of the substation, specifically including:

[0037] Real-time data fusion: Use the sensor data fusion algorithm based on Kalman filtering to fuse the data of UAV aerial photography and ground scanning in real time to update the 3D model.

[0038]

[0039] where is the state estimate at the current time, A is the state transition matrix, B is the control input matrix, u k is the control input (such as UAV position, speed, etc.), K k is the Kalman gain, z k is the observation value (such as newly collected point cloud data), and H is the observation matrix.

[0040] Model optimization: Use the graphics processing algorithm accelerated by GPU to perform real-time rendering and optimization of the 3D model to improve the accuracy and real-time performance of the model.

[0041] Step S104, based on the three-dimensional model of the substation, determine the initial path of the UAV in the substation;

[0042] Optionally, in the method for substation inspection using an unmanned aerial vehicle (UAV), after constructing the 3D model of the substation, the initial inspection path of the UAV is further determined based on this 3D model. This process may or may not involve analyzing the space in the 3D model, identifying areas where the UAV can pass through, and on this basis, planning a safe and efficient inspection route as the initial path of the UAV within the substation.

[0043] In an optional embodiment, based on the 3D model of the substation, determining the initial path of the UAV within the substation includes: performing semantic segmentation on the 3D model of the substation to identify the obstacle areas and passable areas within the substation, where the passable area is the area where the UAV can pass through within the substation; based on the passable area, performing path planning for the UAV to obtain the initial path of the UAV within the substation.

[0044] Optionally, perform semantic segmentation on the 3D model of the substation to identify the obstacle areas and passable areas; then, based on the identified passable area, perform path planning to obtain the initial inspection path of the UAV. Specifically, semantic segmentation is used to classify each pixel or point cloud point in the 3D model of the substation into specific categories, such as "equipment", "building", or "open space", to identify the obstacle areas and passable areas within the substation. Further use a deep learning algorithm (such as DeepLabV3+) for semantic segmentation. This algorithm can learn the characteristics of different types of obstacles based on a large amount of training data and accurately identify these obstacles in new data. By inputting the 3D model of the substation into the trained model, each point in the model can be automatically classified into the corresponding category, such as equipment, building, grassland, etc. The semantic segmentation process can identify all the obstacles within the substation, including power equipment, buildings, and other fixed obstacles. At the same time, it will also identify the areas where the UAV can fly safely, that is, the "passable areas". This information provides the necessary environmental data for subsequent path planning. After identifying the passable area, the next step is to plan the inspection path for the UAV. This may or may not involve using the A* algorithm or Dijkstra algorithm to search for paths in the passable area of the 3D model. Through the calculation of this path planning algorithm, an initial path from the starting point of the UAV to the specified inspection point can be determined. This path will avoid all known obstacle areas to ensure the safety of the UAV during the inspection process.

[0045] Through the above method, based on the 3D model of the substation, obstacles and passable areas can be automatically identified, and then an initial safe inspection path can be planned for the UAV. This process is highly automated and intelligent, which can greatly reduce the need for manual intervention and improve the efficiency and safety of UAV inspection.

[0046] Optionally, a semantic segmentation algorithm based on deep learning (such as DeepLabV3+) can be used, but not limited to, to segment the 3D model and identify obstacles such as equipment and buildings in the substation. Further, based on the obstacle recognition results, a path space is constructed to divide the substation into passable areas and obstacle areas. Among them, during the application of the semantic segmentation algorithm, the construction of the model loss is as follows:

[0047]

[0048] where y i is the true label of pixel i, p i is the predicted probability of pixel i, N is the total number of pixels, and Loss represents the model loss.

[0049] In an alternative embodiment, based on the passable area, path planning is performed on the drone to obtain the initial path of the drone in the substation, including: determining the environmental complexity in the substation; in the case where the environmental complexity is greater than a preset complexity threshold, based on the passable area, using a first path planning algorithm to perform path planning on the drone to obtain the initial path of the drone in the substation; or in the case where the environmental complexity is less than or equal to the preset complexity threshold, based on the passable area, using a second path planning algorithm to perform path planning on the drone to obtain the initial path of the drone in the substation.

[0050] Optionally, the environmental complexity in the substation can be determined based on various factors such as the distribution of obstacles in the substation, the complexity of the terrain, and real-time meteorological data. For example, situations such as dense obstacles, large terrain undulations, and frequent wind speed changes can be determined as having a higher environmental complexity; conversely, few obstacles, flat terrain, and stable meteorological conditions are regarded as having a lower environmental complexity. Another example is that the environmental complexity in the substation can be determined based on the number of obstacles in the substation and / or the number of paths in the substation.

[0051] Optionally, the first path planning algorithm can be the A* algorithm, and the second path planning algorithm can be the Dijkstra algorithm. When the environmental complexity is greater than the preset complexity threshold, the first path planning algorithm (such as the A* algorithm) is used for path planning. The A* algorithm can efficiently find a lower-cost path from the starting point to the ending point in a complex environment through heuristic search. It not only considers the length of the path but also can introduce the prediction and avoidance of obstacles, thus being able to generate a safer and more efficient inspection path. When the environmental complexity is less than or equal to the preset complexity threshold, for a simple environment, the second path planning algorithm (such as the Dijkstra algorithm) is used for path planning. The Dijkstra algorithm can ensure finding the shortest paths from the starting point to all other nodes in a graph without negative-weight edges, and is suitable for situations with a relatively simple environment and fewer obstacles, where not much heuristic information is required and path planning can be directly based on the actual cost.

[0052] In the above way, it is possible to more intelligently adapt to different inspection environments and improve the adaptability and efficiency of unmanned aerial vehicle (UAV) inspections. Path planning in a complex environment pays more attention to safety and path optimization, while a simple environment is more inclined to find the most direct path to save time costs. This dynamic adjustment strategy can improve the efficiency and accuracy of UAV path planning while ensuring the efficient and safe operation of UAVs under different environmental conditions.

[0053] Step S106: Obtain the meteorological data of the area where the substation is located and the equipment status of multiple power equipment included in the substation.

[0054] Optionally, step S106 specifically includes two major parts: one is to collect the meteorological data of the area where the substation is located, and the other is to obtain the real-time equipment status of multiple power equipment in the substation. These information is crucial for optimizing the inspection path of the UAV and ensuring the safety and efficiency of the inspection work.

[0055] Optionally, the meteorological data includes but is not limited to wind speed, wind direction, temperature, humidity, air pressure, and possible weather events (such as rainfall, thunderstorms), etc. These data can be obtained through multiple channels such as meteorological stations, satellite data, and meteorological model predictions, and are transmitted to the control system of the UAV or the ground station in real time through a wireless network. The acquisition of meteorological data helps the UAV to fully understand the environmental conditions before and during flight, and has a decisive impact on flight safety, energy consumption management, and flight path planning. For example, strong winds may require the UAV to adjust its flight speed and altitude, while high temperature or humidity changes may affect the operating status of power equipment and need to be focused on during inspections.

[0056] Optionally, the equipment status data in the substation may include, but is not limited to, the temperature, vibration, sound, visual status (such as appearance damage) of power equipment, and other indicators that may reflect the health status and operating status of the equipment. These data are collected in real time by sensors installed on the power equipment and transmitted to the drone or the ground station. By monitoring and analyzing the equipment status data in real time, it is possible to help the inspection system identify potential faults or anomalies and evaluate the operating conditions of the equipment. According to the equipment status, the inspection path of the drone can be preferentially arranged to conduct a more detailed inspection near the equipment with poor status, or avoid the areas with abnormal status to avoid affecting the inspection task or increasing the risk of the drone.

[0057] Step S108: Optimize the initial path based on the meteorological data and equipment status to obtain the target path, and control the drone to inspect the substation along the target path.

[0058] Optionally, step S108 is the dynamic path optimization stage in the method for inspecting a substation by a drone, aiming to adjust the preset initial path according to the real-time meteorological conditions and the operating status of the power equipment in the substation, so as to obtain the inspection path - the target path - that is most suitable for the current situation.

[0059] In an optional embodiment, optimizing the initial path based on the meteorological data and equipment status to obtain the target path includes: determining multiple objective functions, including: minimizing the total path length of the drone during the inspection of the substation; maximizing the safety of the drone inspection path, where the safety is inversely proportional to the number of obstacles and the safety level; minimizing the total energy consumption during the drone inspection process; maximizing the data collection efficiency of the drone for the preset key areas in the substation; the drone inspection path adapting to the real-time meteorological conditions, where the meteorological conditions include at least one of the following: wind speed, wind direction, temperature, humidity, air pressure; based on the meteorological data, equipment status, and multiple objective functions, using a multi-objective optimization algorithm to optimize the initial path to obtain the target path.

[0060] Optionally, to meet the various requirements of unmanned substation inspection, multiple objective functions are defined. These functions aim to optimize the inspection efficiency, safety, and energy consumption of the UAV, while ensuring the data acquisition efficiency for preset key areas. In the specific setting of objective functions, by setting the minimization of the total path length of UAV inspection, the inspection time can be reduced and the efficiency can be improved. By setting the maximization of the safety of the UAV inspection path, it can be quantified by the inverse ratio of the number of obstacles and the safety level to ensure the minimization of the risk when the UAV performs tasks. By minimizing the total energy consumption during UAV inspection, the path can be optimized to reduce power consumption and extend the continuous operation time of the UAV. By maximizing the data acquisition efficiency for preset key areas, the monitoring quality of key equipment and areas can be ensured, and potential problems can be detected in a timely manner. Moreover, the flight performance and safety of the UAV are significantly affected by meteorological conditions such as wind speed, wind direction, temperature, humidity, and air pressure. The inspection path of the UAV can be adjusted according to the real-time received meteorological data to ensure that the inspection task can still be performed under adverse weather conditions while avoiding possible flight risks. Based on the possible conflicts among the above multiple objective functions, a multi-objective optimization algorithm (such as NSGA-II, etc.) is adopted to comprehensively consider all objectives and globally optimize the initial path. This algorithm can find the optimal solution among multiple objectives, generate a set of objective paths, and select the best path that meets all objectives from them.

[0061] In the above way, it can be ensured that when the UAV performs the unmanned substation inspection task, it can not only efficiently cover the entire area, but also maintain safety and stability in a complex and changeable environment. At the same time, the monitoring of key areas is not compromised, effectively improving the intelligent level of substation operation and maintenance. In addition, by adjusting the path in real time to adapt to meteorological changes, the flexibility and adaptability of UAV inspection can be further enhanced, enabling it to maintain efficient inspection operations even in bad weather.

[0062] In an optional embodiment, the method further includes: obtaining multiple dimensions of detection data collected by the UAV for target power equipment in the substation during the inspection of the substation, where the multiple dimensions include at least two of the following: image data, video data, temperature data, vibration data, sound data, and the target power equipment is any one of multiple power equipment; performing wavelet transform processing on the detection data of multiple dimensions to obtain preprocessed data of multiple dimensions; performing standardization processing on the preprocessed data of multiple dimensions to obtain standardized data of multiple dimensions; performing fusion processing on the standardized data of multiple dimensions to obtain fusion data; and performing anomaly detection on the target power equipment based on the fusion data to obtain the anomaly detection result of the target power equipment.

[0063] Optionally, multi-dimensional detection data collection means that during the inspection by the UAV, it can collect multiple types of data for each power equipment (target power equipment) in the substation, including at least two or more types, such as image data, video data, temperature data, vibration data, sound data, etc. These data comprehensively cover the appearance, operating status, and surrounding environment of the equipment, providing a rich information source for subsequent anomaly detection. Considering that the acquired original detection data may contain noise and irrelevant information, preprocessing is required. Using wavelet transform to process various data can effectively remove noise while retaining important features in the data. Wavelet transform is particularly suitable for signal and image analysis, which can decompose data into different frequencies and scales, helping to identify detailed changes (such as cracks, wear) on the equipment surface and abnormal acoustic and vibration frequencies, providing clearer data for subsequent analysis. To ensure the comparability and consistency of data in different dimensions during fusion and analysis, it is necessary to standardize the preprocessed data after wavelet transform. Standardization processing (such as Z-score standardization) can convert the data into a form with the same mean and standard deviation, eliminating the influence of dimension and scale, enabling different types of sensor data to be compared and fused within the same framework. Fusing the standardized data in multiple dimensions results in a comprehensive fusion data containing multi-sensor information. Data fusion can use algorithms such as federated Kalman filtering, which can not only integrate data from different sensors but also consider the uncertainty of the data, improving the accuracy and reliability of the detection results. The fusion data can provide more comprehensive and accurate equipment status information for the anomaly detection algorithm. Finally, based on the fusion data, a dedicated anomaly detection algorithm (such as the isolation forest algorithm) is used to detect anomalies in the target power equipment. The anomaly detection algorithm can identify data points that are significantly different from the normal operating state, and these data points can indicate equipment failures, performance degradation, or other abnormal situations. Through anomaly detection, the operation and maintenance personnel can receive early warnings in a timely manner, thus quickly locating problems, taking measures, preventing power outages caused by equipment failures, and improving the stability and safety of the power system.

[0064] In the above method, through the collection, preprocessing, fusion, and analysis of multi-dimensional data, the UAV unmanned inspection method can finely monitor the status of power equipment, providing intelligent and automated support for power system operation and maintenance, and can significantly improve the accuracy and timeliness of fault detection.

[0065] Optionally, the drone collects multi-dimensional data, which at least includes one of image, video, temperature, vibration, and sound. The collected data is preprocessed using wavelet transform to obtain preprocessed data, and the preprocessed data is standardized using the Z-score standardization method to obtain standardized data. The federated Kalman filter algorithm is used to perform real-time fusion on the processed standardized data to obtain fusion data. The isolated forest algorithm is used to detect abnormal points in the fusion data. The long short-term memory (LSTM) algorithm can be used to perform intelligent analysis on the fusion data, extract key features, construct a fault prediction model, and perform abnormal detection on the target power equipment. Among them, the specific implementation process of data processing and data fusion is as follows:

[0066] Step S1081, data collection and preprocessing, specifically including:

[0067] Diversified data collection: The drone collects multi-dimensional data such as images, videos, temperatures, vibrations, and sounds.

[0068] Data preprocessing: Wavelet Transform is used to denoise image and video data, and the Z-score standardization method is used to standardize sensor data.

[0069]

[0070] Among them, z is the standardized data, w is the original data, μ is the mean, and σ is the standard deviation.

[0071] Step S1082, data fusion and integration, specifically including:

[0072] Multi-source data fusion: The federated Kalman filter algorithm is used to fuse data from different sensors.

[0073]

[0074] Among them, is the state estimation at the current moment of the t-th sensor, A t , B t , H t are the state transition matrix, control input matrix, and observation matrix of the t-th sensor respectively, and K k,t is the Kalman gain of the t-th sensor.

[0075] Data integration and storage: The Hadoop Distributed File System (HDFS) of Hadoop is adopted to store the integrated data set, which facilitates subsequent data analysis and applications.

[0076] Optionally, the Isolation Forest algorithm can be used to perform anomaly detection on the integrated data set (i.e., the integrated data).

[0077] s(x,a) = 2 - E(h(a))

[0078] Among them, s(a,n) is the anomaly score of sample x, and E(h(a)) is the reciprocal of the expected value of the path length of sample a in the random forest.

[0079] The Long Short-Term Memory network (LSTM) can also be used to analyze the historical data of the device and predict possible future failures.

[0080] h t = σ(W h h t-1 + W z x t + b h )

[0081] C t = f t ⊙C t-1 + i t ⊙tanh(W c h t-1 + W z x t + b c )

[0082] o t = σ(W o h t-1 + W z x t + b o )

[0083] y t = h t = o t ⊙tanh(C t )

[0084] Among them, h t is the hidden state, C t is the cell state, f t , i t , o t are the activation values of the forget gate, input gate, and output gate respectively, and W h , W z , Wc ,W o and b h ,b c ,b o are the weight and bias terms respectively.

[0085] It is also possible to combine an image recognition algorithm (such as Faster R-CNN) and temperature distribution data to accurately locate and diagnose faults.

[0086]

[0087] Among them, L cls is the classification loss, L reg is the regression loss, p i is the predicted probability, u i is an indicator of whether it is a positive sample, t i is the parameter of the ground truth box, v i is the parameter of the predicted box, and λ is the balance coefficient.

[0088] In an alternative embodiment, the method further includes: during the process of controlling the drone to inspect the substation along the target path, collecting environmental data inside the substation through a plurality of sensors arranged on the drone; fusing the environmental data collected by the plurality of sensors respectively to construct a three-dimensional environmental perception map; performing target detection on the three-dimensional environmental perception map to detect whether there are target obstacles in the substation; in the case where target obstacles are detected in the substation, generating an obstacle avoidance path based on the position information of the target obstacles and the target path; controlling the drone to inspect the substation along the obstacle avoidance path.

[0089] Optionally, during the inspection of the UAV along the target path, multiple sensors carried on it (such as visual sensors, lidar, ultrasonic sensors, infrared sensors, etc.) will continuously collect environmental data inside the substation. These sensors can capture real-time images of the substation, the precise positions of obstacles, changes in environmental temperature and humidity, etc. from different angles and frequencies, providing basic data for constructing a three-dimensional environmental perception map. To more accurately understand the environment where the UAV is located, the environmental data collected by multiple sensors are fused. The Kalman filtering algorithm can be used to process and fuse these data to generate a high-precision three-dimensional environmental perception map. This perception map not only contains the structural information of the substation but also can update the positions of obstacles and dynamic environmental changes in real time, providing an intuitive and comprehensive environmental view for subsequent obstacle detection and obstacle avoidance path planning. Using a deep learning-based object detection algorithm (such as YOLOv5) to perform real-time analysis on the constructed three-dimensional environmental perception map to detect whether there are target obstacles inside the substation. The object detection algorithm can identify and locate objects in the image, so on the three-dimensional perception map, it can also identify and locate any obstacles that may affect the flight safety of the UAV, including suddenly appearing non-fixed obstacles (such as staff, wild animals). If an obstacle is identified during the object detection process, an obstacle avoidance path can be generated based on the position information of the obstacle and the current target path of the UAV. The A* algorithm or Dijkstra algorithm can be combined to re-plan the travel route of the UAV to ensure that the UAV can avoid obstacles and continue to perform the inspection task along a safe path. The generated obstacle avoidance path should not only consider physical obstacles but also consider the flight performance limitations of the UAV and the accessibility of the safe landing point. According to the generated obstacle avoidance path, adjust the flight control strategy of the UAV to guide the UAV to avoid obstacles and continue to inspect the substation. This process requires the autonomous decision-making ability of the UAV and real-time communication with the ground station to ensure the dynamic adjustment of the path and the precise control of the UAV position.

[0090] In the above method, through real-time environmental data fusion and obstacle detection, the UAV is provided with the ability of dynamic obstacle avoidance during the unmanned inspection of the substation, ensuring the safe flight and efficient inspection of the UAV in a complex environment. The UAV can autonomously respond to emergencies during the inspection process without human intervention, reducing the need for manual operation and improving the autonomy and flexibility of the inspection task.

[0091] Through the above steps S102 to S108, the purpose of the drone and the ground laser scanning device to cooperate in collecting terrain data, constructing and dynamically updating the 3D model of the substation, and intelligently adjusting the inspection path of the drone in combination with real-time meteorological data and the status of power equipment can be achieved, so as to achieve the technical effects of improving the inspection efficiency, reducing manual intervention, enhancing the safety and accuracy of the inspection, and further solving the technical problems of high manual participation, low inspection efficiency and difficult safety guarantee in substation inspection.

[0092] Based on the above embodiments and alternative embodiments, the present invention proposes an alternative implementation manner. Figure 2 It is a flowchart of an alternative method for inspecting a substation based on a drone according to an embodiment of the present invention. Figure 3 It is a flowchart of another alternative method for inspecting a substation based on a drone according to an embodiment of the present invention, as Figure 2 and Figure 3 shown. The method includes:

[0093] S1. Use the drone to combine aerial photography and ground laser scanning to collect the terrain of the substation, obtain the terrain data of the substation, and make a dynamically updatable 3D model of the substation according to the terrain data of the substation.

[0094] Specifically, use point cloud processing to preprocess the terrain data of the substation to obtain the processed point cloud data; use a mesh generation algorithm to construct a high-precision three-dimensional (i.e., 3D) model of the substation using the point cloud data, and dynamically update the 3D model of the substation through real-time data fusion and manual model optimization, and set important areas, importance levels and urgency levels. Use the filtering algorithm in the PCL library to filter the point cloud data to reduce the data density, and use the Poisson surface reconstruction algorithm to convert the filtered point cloud data into a 3D model of the substation.

[0095] S2. Use the 3D model of the substation to calculate the initial path available for the drone to pass through in the substation space, and at the same time import real-time meteorological data and equipment status to dynamically adjust the initial path to obtain the target path.

[0096] Specifically, use the semantic segmentation algorithm of deep learning to construct a path space in the 3D model of the substation, and divide the substation into passable areas and obstacle areas; use the A* algorithm in complex environments and the Dijkstra algorithm in simple environments to plan paths in the passable areas to obtain the initial path, and use a multi-objective optimization algorithm to globally optimize the initial path to obtain the final path.

[0097] S3. Use the multi-sensors of the drone to sense the surrounding environment in real time, adopt a dynamic obstacle avoidance strategy and an emergency landing mechanism, and control the drone to perform inspections along the target path.

[0098] Among them, the multi-sensor includes at least one of an integrated vision sensor, a lidar, and an inertial navigation system. The multi-sensor collects and senses the surrounding environment information. According to the environmental information collected by the multi-sensor, the Kalman filtering algorithm is used to process and fuse the environmental information to generate a high-precision three-dimensional environmental perception map. The object detection algorithm based on deep learning is used to detect obstacles in the high-precision three-dimensional environmental perception map in real time, and a safe obstacle avoidance path is planned in cooperation with the substation 3D model.

[0099] S4. According to the aerial data obtained by the drone for substation inspection, the fusion and analysis algorithm is used to perform intelligent processing and analysis on the aerial data.

[0100] Specifically, the drone collects multi-dimensional data, and the multi-dimensional data includes at least one of images, videos, temperature, vibration, and sound. The collected data is preprocessed using wavelet transform to obtain preprocessed data, and the preprocessed data is standardized using the Z-score normalization method to obtain standardized data.

[0101] The federated Kalman filtering algorithm can be used to perform real-time fusion on the processed standardized data to obtain fusion data. The isolation forest algorithm is used to detect outliers in the fusion data, and the LSTM algorithm is used to perform intelligent analysis on the fusion data to extract key features and construct a fault prediction model.

[0102] The inspection data obtained by the drone inspection can be intelligently processed and analyzed through data fusion and analysis, improving the accuracy and timeliness of fault identification. The application of specific algorithms such as federated Kalman filtering, isolation forest, and LSTM provides decision-making support for maintenance personnel.

[0103] This embodiment also provides a substation unmanned inspection device. Figure 4 It is a schematic diagram of an optional substation unmanned inspection device according to an embodiment of the present invention. This substation unmanned inspection device can be applied to the drone-based substation inspection method as shown in Figure 2 and Figure 3 and includes a drone and an airport for taking off and landing the drone. The drone is equipped with an environment collection module, a wireless network module, a status monitoring and fault warning module, and a carrier receiving module. The airport includes a takeoff and landing control module, a charging module, a data reading module, a multi-node data interaction module, and a carrier transmitting module.

[0104] Both the drone and the airport are equipped with a processor and a battery. The carrier receiving module of the drone is communicatively connected to the carrier transmitting modules of multiple airports at the same time.

[0105] The airport serves as the ground station, maintaining communication with the UAV during its operation and real-time positioning of the UAV's location. The positioning method uses Real-time kinematic (RTK). The carrier wave transmitting module of the reference station (the airport) transmits carrier wave observation values and station coordinate information to the carrier wave receiving module of the user station (the UAV) in real time. The carrier wave receiving module of the user station (the UAV) combines the received GPS satellite carrier phase with the reference station carrier phase to form a phase difference observation value, which is processed in real time, thereby quickly giving a positioning result accurate to the centimeter level. The positioning accuracy formula is as follows:

[0106] Positioning accuracy = 2 × phase difference observation value / carrier wavelength;

[0107] Through the precise positioning of the UAV, real-time monitoring of the UAV's position is achieved. Then, by importing the 3D model of the substation, the movement path of the UAV in the substation can be precisely controlled, preventing the UAV from misjudging its own position due to inaccurate positioning and avoiding the UAV from entering the wrong path.

[0108] By setting up a takeoff and landing control module, during the takeoff and landing process of the UAV, the airport locates and controls the UAV, enabling the UAV to accurately stop on the airport. And combined with the UAV's status monitoring and fault warning module, it can judge whether the UAV is suitable for takeoff and automatically control the takeoff and landing of the UAV.

[0109] The UAV transmits various information collected by the environmental collection module to the airport through the wireless network module. The data reading module of the airport reads the relevant data. And to ensure the stability of data transmission, the wireless network connection of the UAV can automatically switch to the airport with the best signal, and the data is transmitted in segments to each airport, so that the UAV can continuously transmit data during the inspection process. Then each airport uses the multi-node data interaction module to verify the data with each other, complete the data and finally transmit it to the control center, with a more stable signal and less likely to have data errors.

[0110] It should be noted that in this embodiment, by performing high-precision 3D modeling and dynamic update on the substation, the accuracy and efficiency of the inspection are improved, and the real-time performance and accuracy of the model are ensured; intelligent path planning and optimization are carried out for the behavior of the UAV: according to the real-time meteorological data and equipment status, the path is dynamically adjusted, the inspection efficiency is optimized, and the utilization rate of the UAV is increased; the UAV can avoid obstacles autonomously and perform emergency handling during operation: through measures such as multi-sensor fusion, dynamic obstacle avoidance strategy, and emergency landing mechanism, the safety and reliability of the UAV are improved, and the failure rate is reduced. The application of specific algorithms such as Kalman filter, YOLOv5, and A* algorithm further improves the ability of obstacle avoidance and emergency handling, so that in the inspection work, the UAV only needs technical support during the first initial setting, and then the UAV can perform inspections by itself, greatly reducing manual participation and lowering the usage threshold.

[0111] This embodiment also uses a status detection and fault warning module to warn the UAV status during inspection according to real-time meteorological data, UAV usage data, etc., so as to lower the UAV in advance before predicting a possible failure, preventing the UAV from crashing or hitting other equipment. By setting a carrier receiving module on the UAV and a carrier transmitting module on the airport, signals are simultaneously sent and received to the UAV by multiple airports, and the position and altitude of the UAV are located by integrating the data of each airport, and the UAV is more accurately located in cooperation with the UAV's own positioning devices such as GPS.

[0112] In this embodiment, a substation inspection device based on a UAV is also provided. This device is used to implement the above-mentioned embodiments and preferred implementation manners, and those that have been described will not be repeated. As used hereinafter, the terms "module" and "device" can be a combination of software and / or hardware that can achieve a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, implementation in hardware, or a combination of software and hardware is also possible and contemplated.

[0113] According to an embodiment of the present invention, an apparatus embodiment for implementing the above-mentioned substation inspection method based on a UAV is also provided. Figure 5 is a schematic structural diagram of a substation inspection device based on a UAV according to an embodiment of the present invention, as Figure 5 shown, the above-mentioned substation inspection device based on a UAV includes: a three-dimensional model construction module 500, an initial path acquisition module 502, a data acquisition module 504, and a target path acquisition module 506, wherein:

[0114] The three-dimensional model construction module 500 is used to collect the terrain data of the substation based on the UAV and the ground laser scanning device, and construct a three-dimensional model of the substation based on the terrain data;

[0115] An initial path acquisition module 502, connected to the three-dimensional model construction module 500, is used to determine the initial path of the UAV in the substation based on the three-dimensional model of the substation;

[0116] A data acquisition module 504, connected to the initial path acquisition module 502, is used to acquire the meteorological data of the area where the substation is located and the equipment status of multiple power equipment included in the substation;

[0117] A target path acquisition module 506, connected to the data acquisition module 504, is used to optimize the initial path based on the meteorological data and the equipment status to obtain a target path, and control the UAV to inspect the substation along the target path.

[0118] In the embodiment of the present invention, by setting a three-dimensional model construction module 500, which is used to collect the topographic data of the substation based on the UAV and the ground laser scanning device, and construct a three-dimensional model of the substation based on the topographic data; an initial path acquisition module 502, connected to the three-dimensional model construction module 500, is used to determine the initial path of the UAV in the substation based on the three-dimensional model of the substation; a data acquisition module 504, connected to the initial path acquisition module 502, is used to acquire the meteorological data of the area where the substation is located and the equipment status of multiple power equipment included in the substation; a target path acquisition module 506, connected to the data acquisition module 504, is used to optimize the initial path based on the meteorological data and the equipment status to obtain a target path, and control the UAV to inspect the substation along the target path, the purpose of the UAV and the ground laser scanning device to cooperate to collect topographic data, construct and dynamically update the three-dimensional model of the substation, and intelligently adjust the UAV inspection path in combination with real-time meteorological data and power equipment status is achieved, thereby realizing the technical effects of improving the inspection efficiency, reducing manual intervention, enhancing the inspection safety and accuracy, and further solving the technical problems of high manual participation, low inspection efficiency and difficult to guarantee safety in substation inspection.

[0119] It should be noted that the above-mentioned various modules can be implemented by software or hardware. For example, for the latter, it can be implemented in the following ways: the above-mentioned various modules can be located in the same processor; or, the above-mentioned various modules are located in different processors in any combination.

[0120] It should be noted here that the above-mentioned three-dimensional model construction module 500, initial path acquisition module 502, data acquisition module 504, and target path acquisition module 506 correspond to steps S102 to S108 in the embodiment. The examples and application scenarios implemented by the above-mentioned modules and the corresponding steps are the same, but are not limited to the content disclosed in the above-mentioned embodiments. It should be noted that the above-mentioned modules can run in a computer terminal as part of the device.

[0121] It should be noted that the optional or preferred implementation manners of this embodiment can be referred to the relevant descriptions in the embodiments, and will not be elaborated here.

[0122] The above-mentioned substation inspection device based on a drone may further include a processor and a memory. The above-mentioned three-dimensional model construction module 500, initial path acquisition module 502, data acquisition module 504, target path acquisition module 506, etc. are all stored in the memory as program modules, and the processor executes the above program modules stored in the memory to implement corresponding functions.

[0123] The processor contains a kernel, and the kernel retrieves the corresponding program modules from the memory. One or more of the above kernels can be set. The memory may include non-permanent memory in a computer-readable medium, forms such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash memory (flash RAM), and the memory includes at least one storage chip.

[0124] According to an embodiment of the present application, an embodiment of a non-volatile storage medium is also provided. Optionally, in this embodiment, the above non-volatile storage medium includes a stored program, wherein when the above program runs, it controls the device where the non-volatile storage medium is located to execute any one of the above-mentioned substation inspection methods based on a drone.

[0125] Optionally, in this embodiment, the above non-volatile storage medium can be located in any one of the computer terminals in a computer terminal group in a computer network, or in any one of the mobile terminals in a mobile terminal group. The above non-volatile storage medium includes a stored program.

[0126] Optionally, when the program runs, it controls the device where the non-volatile storage medium is located to execute the following functions: collecting topographic data of the substation based on a drone and a ground laser scanning device, constructing a three-dimensional model of the substation based on the topographic data; determining an initial path of the drone in the substation based on the three-dimensional model of the substation; obtaining meteorological data of the area where the substation is located and the device status of a plurality of power equipment included in the substation; optimizing the initial path based on the meteorological data and the device status to obtain a target path, and controlling the drone to inspect the substation along the target path.

[0127] According to an embodiment of the present application, an embodiment of a processor is also provided. Optionally, in this embodiment, the above processor is used to run a program, wherein when the above program runs, it executes any one of the above-mentioned substation inspection methods based on a drone.

[0128] According to an embodiment of the present application, an embodiment of a computer program product is further provided. When executed on a data processing device, it is adapted to execute a program that initializes the steps of the drone-based substation inspection method described above.

[0129] Optionally, when the above computer program product is executed on a data processing device, it is adapted to execute a program with the following method steps: collect topographic data of the substation based on a drone and a ground laser scanning device, and construct a 3D model of the substation based on the topographic data; determine an initial path of the drone within the substation based on the 3D model of the substation; obtain meteorological data of the area where the substation is located and the device status of multiple power equipment included in the substation; optimize the initial path based on the meteorological data and the device status to obtain a target path, and control the drone to inspect the substation along the target path.

[0130] An embodiment of the present invention provides an electronic device. The electronic device includes a processor, a memory, and a program stored on the memory and executable on the processor. When the processor executes the program, the following steps are implemented: collect topographic data of the substation based on a drone and a ground laser scanning device, and construct a 3D model of the substation based on the topographic data; determine an initial path of the drone within the substation based on the 3D model of the substation; obtain meteorological data of the area where the substation is located and the device status of multiple power equipment included in the substation; optimize the initial path based on the meteorological data and the device status to obtain a target path, and control the drone to inspect the substation along the target path.

[0131] The order of the above embodiments of the present invention is only for description and does not represent the superiority or inferiority of the embodiments.

[0132] In the above embodiments of the present invention, the descriptions of each embodiment have their own emphases. For parts not detailed in a certain embodiment, reference may be made to the relevant descriptions of other embodiments.

[0133] In several embodiments provided by the present application, it should be understood that the disclosed technical content can be implemented in other ways. Among them, the device embodiments described above are only illustrative. For example, the above module division can be a logical function division. In actual implementation, there can be other division methods. For example, multiple modules or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed coupling or direct coupling or communication connection between each other can be through some interfaces. The indirect coupling or communication connection of modules or modules can be in an electrical or other form.

[0134] The modules described above as separate components may or may not be physically separated. The components shown as modules may or may not be physical modules, that is, they may be located in one place or distributed across multiple modules. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0135] In addition, the functional modules in each embodiment of the present invention can be integrated into one processing module, or each module can exist physically alone, or two or more modules can be integrated into one module. The above-mentioned integrated modules can be implemented in the form of hardware or in the form of software functional modules.

[0136] If the above-mentioned integrated module is implemented in the form of a software functional module and sold or used as an independent product, it can be stored in a computer-readable non-volatile storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a non-volatile storage medium and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods in each embodiment of the present invention. The aforementioned non-volatile storage medium includes: USB flash drives, read-only memories (ROMs), random access memories (RAMs), mobile hard disks, magnetic disks, or optical discs and other various media that can store program codes.

[0137] The above are only the preferred embodiments of the present invention. It should be noted that for those of ordinary skill in the art, without departing from the principle of the present invention, several improvements and refinements can be made, and these improvements and refinements should also be regarded as the protection scope of the present invention.

Claims

1. A substation inspection method based on drone, characterized in that: include: Using drones and ground laser scanning equipment, collect terrain data of the substation, and build a three-dimensional model of the substation based on the terrain data; Determining an initial path of the UAV within the substation based on the three-dimensional model of the substation; Acquiring meteorological data of the area where the substation is located, and equipment status of multiple power equipment included in the substation; Based on the meteorological data and the equipment status, the initial path is optimized to obtain a target path, and the UAV is controlled to inspect the substation along the target path.

2. The method according to claim 1, characterized in that: The step of constructing a three-dimensional model of a substation based on the terrain data includes: Performing filtering processing on the terrain data to obtain processed data, wherein the terrain data is point cloud data, and wherein the filtering processing is used to remove noise and redundancy of the point cloud data; Surface reconstruction is performed based on the processed data, and the processed data is converted into the three-dimensional model of the substation.

3. The method according to claim 1, characterized in that The determining, based on the three-dimensional model of the substation, an initial path of the drone within the substation includes: Performing semantic segmentation on the three-dimensional model of the substation to identify obstacle areas and passable areas within the substation, wherein the passable area is an area within the substation through which the drone can pass; Based on the traversable area, path planning is performed on the UAV to obtain the initial path of the UAV in the substation.

4. The method according to claim 3, characterized in that The performing path planning for the UAV based on the traversable area to obtain the initial path of the UAV in the substation includes: Determining the environmental complexity within the substation; When the environmental complexity is greater than a preset complexity threshold, based on the traversable area, a first path planning algorithm is used to perform path planning for the UAV to obtain the initial path of the UAV in the substation; or When the environmental complexity is less than or equal to the preset complexity threshold, a second path planning algorithm is used to plan the path of the UAV based on the traversable area to obtain the initial path of the UAV in the substation.

5. The method according to claim 1, characterized in that The optimizing the initial path based on the meteorological data and the device status to obtain a target path includes: Determining multiple objective functions includes: minimizing the total path length of the drone inspection in the substation; maximizing the safety of the drone inspection path, wherein the safety is inversely proportional to the number of obstacles and the safety level; minimizing the total energy consumption during the drone inspection process; maximizing the data collection efficiency of the drone for the preset key areas in the substation; the drone inspection path adapts to real-time meteorological conditions, wherein the meteorological conditions include at least one of the following: wind speed, wind direction, temperature, humidity, and air pressure; Based on the meteorological data, the equipment status and the multiple objective functions, a multi-objective optimization algorithm is used to optimize the initial path to obtain the target path.

6. The method according to any one of claims 1 to 5, characterized in that The method further comprises: Acquire detection data of multiple dimensions collected by the drone for a target power device in the substation during the inspection of the substation, wherein the multiple dimensions include at least two of the following: image data, video data, temperature data, vibration data, and sound data, and the target power device is any power device among the multiple power devices; Performing wavelet transformation on the detection data of the multiple dimensions to obtain pre-processed data of the multiple dimensions; Performing standardization processing on the preprocessed data of the multiple dimensions to obtain standardized data of the multiple dimensions; Performing fusion processing on the standardized data of the multiple dimensions to obtain fused data; Anomaly detection is performed on the target power equipment based on the fused data to obtain an abnormality detection result of the target power equipment.

7. The method according to any one of claims 1 to 5, characterized in that The method further comprises: In the process of controlling the drone to inspect the substation along the target path, collecting environmental data in the substation through a plurality of sensors provided on the drone; Fusing the environmental data collected by the multiple sensors to construct a three-dimensional environmental perception map; Performing target detection on the three-dimensional environment perception map to detect whether there is a target obstacle in the substation; When the target obstacle is detected in the substation, an obstacle avoidance path is generated based on the location information of the target obstacle and the target path; The drone is controlled to inspect the substation along the obstacle avoidance path.

8. A substation inspection device based on drone, characterized in that: include: A three-dimensional model building module is used to collect terrain data of the substation based on drones and ground laser scanning equipment, and build a three-dimensional model of the substation based on the terrain data; An initial path acquisition module, used to determine an initial path of the UAV within the substation based on the three-dimensional model of the substation; A data acquisition module, used to acquire meteorological data of the area where the substation is located, and the equipment status of multiple power equipment included in the substation; The target path acquisition module is used to optimize the initial path based on the meteorological data and the equipment status to obtain the target path, and control the UAV to inspect the substation along the target path.

9. A non-volatile storage medium, characterized in that: The non-volatile storage medium stores a plurality of instructions, and the instructions are suitable for being loaded by a processor and executing the substation inspection method based on a drone as described in any one of claims 1 to 7.

10. An electronic device, characterized in that: It includes one or more processors and a memory, wherein the memory is used to store one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement the substation inspection method based on a drone as described in any one of claims 1 to 7.

Images