Unmanned aerial vehicle intelligent inspection method and system in butt joint with intelligent management and control platform

Abstract

The invention provides an unmanned aerial vehicle intelligent inspection method and system in butt joint with an intelligent management and control platform, and the method comprises the steps: transmitting a butt joint signal through monitoring the voltage of a battery in real time; the charging base station with the minimum arrival cost is determined as a target charging base station through a Dijkstra algorithm, and the unmanned aerial vehicle is guided to a target charging position of the target charging base station through a multi-stage navigation module; and completing docking with the charging base station and charging, when the battery is fully charged, determining a return position with the minimum arrival cost according to the recorded navigation path sequence in combination with a Dijkstra algorithm to execute return, and continuing to execute the line patrol task. Through a multi-stage positioning strategy, the charging position of the charging platform can be accurately and automatically positioned for charging, and all-weather unmanned independent operation is realized. And the charging base station with the minimum arrival cost and the return position are determined through a Dijkstra algorithm, so that the energy consumption is reduced. And processing multi-modal input data through the constructed multi-modal neural network image visual navigation module, and outputting motor control parameters to realize accurate positioning and stable landing.

Description

Technical Field

[0001] The present application relates to the field of drone charging technology, and in particular to a drone intelligent inspection method and system connected to an intelligent management and control platform. Background Art

[0002] With the development of drone technology, the use of drones to replace manual inspections of power transmission channels has become increasingly common. Furthermore, with the rapid economic and social development of my country and the continued advancement of airspace management reforms, as well as the gradual opening up of low-altitude airspace, drones will see significant development and even wider application. For example, drones can be applied to fields such as power generation, communications, meteorology, agriculture and forestry, oceanography, exploration, and insurance.

[0003] Because drone inspection routes are long and they constantly perform filming and data transmission, they quickly deplete their battery, making it difficult to maintain smooth inspections. They can even be unable to return to base due to insufficient power. This short battery life restricts operational continuity. Current drones on the market typically require manual battery replacement, lacking automated charging and control. This makes them inadequate for practical use and prevents them from achieving unmanned, independent operation around the clock. Furthermore, the battery replacement process presents challenges with real-time detection.

[0004] Various docking devices are currently available on the market, connecting to the drone's charging port to enable charging. However, the inability of these devices to accurately locate and monitor battery voltage makes drone maintenance inconvenient. Furthermore, the diverse design structures of charging platforms still present obstacles to docking, which can easily lead to docking errors or damage.

[0005] Therefore, designing an intelligent inspection method for drones that is connected to an intelligent management and control platform to accurately navigate the charging platform, accurately locate the charging position, and accurately monitor the charging status is an important means and urgent task to ensure the stable operation of power grid inspection tasks. Summary of the Invention

[0006] In view of this, it is necessary to provide a drone intelligent inspection method and system that is connected to an intelligent management and control platform to solve the charging and docking problems of drones when performing inspection tasks. It is easy to promote and apply, has relatively simple operation, and is suitable for promotion and application in more scenarios.

[0007] The present application provides a method and system for intelligent inspection of unmanned aerial vehicles that is docked with an intelligent management and control platform. The method sends a docking signal by monitoring the voltage of the battery in real time; the charging base station with the minimum arrival cost is determined as the target charging base station by the Dijkstra algorithm, and the drone is guided to the target charging position of the target charging base station by the multi-level navigation module; the docking with the charging base station is completed and charging is performed. When the battery is fully charged, the return position with the minimum arrival cost is determined according to the recorded navigation path sequence combined with the Dijkstra algorithm, and the line patrol task is continued. Through the multi-level positioning strategy, the charging position of the charging platform can be accurately and automatically located for charging, realizing all-weather unmanned independent operation. The charging base station and return position with the minimum arrival cost are determined by the Dijkstra algorithm, saving energy. The multimodal input data is processed by the constructed multimodal neural network image vision navigation module, and the motor control parameters are output to achieve precise positioning and smooth landing.

[0008] In a first aspect, an embodiment of the present application provides a method for intelligent inspection of a drone connected to an intelligent management and control platform, the method comprising:

[0009] S1. Low voltage detection: The drone monitors the battery voltage in real time during the inspection process. When the battery voltage falls below the set voltage threshold, a docking signal is issued;

[0010] S2. UAV docking first navigation: The UAV navigation system determines the charging base station with the minimum arrival cost as the target charging base station based on the docking signal using the Dijkstra algorithm, and simultaneously activates the first navigation module to guide the UAV to the vicinity of the target charging base station;

[0011] S3. UAV docking second navigation: When the UAV reaches the first preset range from the target charging base station, the UAV navigation system starts the second navigation module to guide the UAV to the target charging base station above;

[0012] S4. UAV docking third navigation: When the UAV reaches the second preset range above the target charging base station, the UAV navigation system activates the third navigation module to guide the UAV to the target charging position of the charging base station;

[0013] S5. UAV docking: Use the docking module on the drone to complete docking with the charging base station;

[0014] S6. Drone Charging Monitoring: After docking is complete, the drone will pair with the charging base station via Bluetooth. Once the Bluetooth pairing is successful, the drone will send a charging signal to the charging base station, which will then charge the drone's battery pack via wireless power transmission. During the charging process, the drone will continuously monitor the battery level. When the battery is fully charged, the drone will send a stop charging signal to the charging base station.

[0015] S7. Return of the UAV: The UAV returns to the location with the minimum arrival cost determined by the recorded navigation path sequence combined with the Dijkstra algorithm and continues to perform the line patrol mission.

[0016] Optionally, in an implementation of the first aspect of the present invention, the S1. low voltage detection: the drone monitors the battery voltage in real time during the inspection process, and sends a docking signal when the battery voltage is lower than a set voltage threshold, including:

[0017] Use high-frequency sampling to collect battery load voltage in real time;

[0018] Use hardware filtering circuit combined with software sliding window algorithm to eliminate voltage fluctuation interference;

[0019] A voltage-to-power mapping table is established through a dynamic threshold strategy, and an early warning method is established based on the mapping table, including: Level 1 warning: trigger status prompt, Level 2 warning: forced start of return procedure, and emergency protection: immediate execution of on-site landing;

[0020] The intelligent decision-making module evaluates the warning method, the estimated remaining battery life, the terrain complexity of the return path, the backup power status, and the mission criticality level to determine whether charging or returning is required;

[0021] When it is determined that charging is required, a docking signal is sent.

[0022] Optionally, in an implementation of the first aspect of the present invention, S2. UAV docking first navigation: The UAV navigation system determines, based on the docking signal, a charging base station with the minimum arrival cost using a Dijkstra algorithm as a target charging base station, and simultaneously activates a first navigation module to guide the UAV to the vicinity of the target charging base station, including:

[0023] S2.1, initialization: set the distance of the starting point to 0, the distance of all other nodes to infinity, and create an unvisited node set U containing all nodes;

[0024] S2.2, select the node with the minimum distance: select the node u that is currently closest to the starting point from the unvisited node set U, and move the node u from the unvisited set U to the visited set S;

[0025] S2.3, Update the distances of adjacent nodes: Traverse all adjacent nodes v of node u and calculate the distances to these adjacent nodes through node u; if the distance to adjacent node v through node u is less than the currently recorded distance, update the distance of v. After the update, add these adjacent nodes back to the set of unvisited nodes U;

[0026] S2.4, repeat steps S2.2 and S2.3: repeatedly select the minimum distance node and update the distances of its adjacent nodes until the unvisited node set U becomes empty or all nodes are processed;

[0027] S2.5, output result: the final distance array D, where d[i] represents the shortest path length from the starting point to node i, and the predecessor node is recorded to restore the specific shortest path;

[0028] The node set includes charging stations and key points on the path, and the edge weight is a cost function. The cost function formula is:

[0029] Cost total =α×E+β×T+γ×P+λ×T2;

[0030] E=(d×e_base×w_factor×a_factor)+e_hover×t_hover;

[0031]

[0032] Where d is the total flight distance, e_base is the basic energy consumption per unit distance, w_factor is the weather correction factor, w_factor is the altitude correction factor, e_hover is the energy consumption per unit time of hovering, t_hover is the expected hovering waiting time, E is the energy cost, T is the task delay penalty, P is the priority cost, and α, β, γ are the corresponding weight coefficients;

[0033] Use Dijkstra algorithm to optimize path planning to ensure the shortest and safest route;

[0034] S2.6, the UAV navigation system determines the priority of the navigation module based on the detection distance, wherein the priority includes a first priority, a second priority, and a third priority;

[0035] S2.7, when the priority is the first priority, matching the first navigation module;

[0036] S2.8. Start a first navigation module to guide the drone to the vicinity of the target charging base station, where the first navigation module is a GPS.

[0037] Optionally, in an implementation of the first aspect of the present invention, S3. UAV docking second navigation: When the UAV arrives within a first preset range from the target charging base station, the UAV navigation system activates a second navigation module to guide the UAV above the target charging base station, including:

[0038] Determining whether the first preset range meets the detection distance corresponding to the second priority;

[0039] When the conditions are met, the second navigation module is matched;

[0040] The second navigation module is activated by the drone navigation system to guide the drone to a preset position above the target charging base station; wherein the second navigation module is a GPS and ultrasonic combined navigation module.

[0041] Optionally, in one implementation of the first aspect of the present invention, S4. UAV docking third navigation: When the UAV arrives within a second preset range above the target charging base station, the UAV navigation system activates a third navigation module to guide the UAV to the target charging position of the charging base station, including:

[0042] Determining whether the second preset range meets the detection distance corresponding to the third priority;

[0043] When satisfied, match the third navigation module;

[0044] The drone navigation system activates a third navigation module to guide the drone to a target charging location of the charging base station; wherein the third navigation module is an image vision navigation module;

[0045] The image vision navigation module is a multimodal neural network architecture, which processes multimodal input data through the multimodal neural network architecture to output motor control parameters, and controls the motor according to the control parameters to achieve accurate positioning and smooth landing, wherein the second preset range is smaller than the first preset range.

[0046] Optionally, in an implementation of the first aspect of the present invention, processing the multimodal input data using a multimodal neural network architecture to output motor control parameters includes:

[0047] Constructing a multimodal neural network architecture, the multimodal neural network architecture comprising an input layer, a modality-specific encoder, and a multimodal fusion module;

[0048] The drone's acquisition module collects visual image data, the drone's attitude angle, height, and position information.

[0049] The visual image data is segmented through a multimodal neural network architecture to obtain the outer contour of the target charging platform, and the charging location area is further identified. The relative position is calculated by combining the height and position information, the attitude angle of the drone is adjusted, and the motor control parameters are output.

[0050] Optionally, in an implementation of the first aspect of the present invention, S6. Return of the UAV: The UAV determines a return location with a minimum arrival cost based on the recorded navigation path sequence in combination with the Dijkstra algorithm, and returns to continue performing the line patrol mission, including:

[0051] Use Dijkstra's algorithm to determine the return path with the minimum arrival cost, and record D', which represents the shortest path length from the X end point to node j;

[0052] Compare the minimum arrival costs consumed by array D' and array D, select the shortest path corresponding to the smaller minimum arrival cost array to perform the return trip, and continue to perform the line patrol task.

[0053] In a second aspect, an embodiment of the present application provides a drone intelligent inspection system connected to an intelligent management and control platform, which is applied to the drone intelligent inspection method connected to an intelligent management and control platform as described in any one of claims 1 to 7, and is characterized by comprising:

[0054] A low voltage detection module is used to monitor the battery voltage of the drone in real time during the inspection process. When the battery voltage is lower than the set voltage threshold, a docking signal is issued;

[0055] The drone docks with the first navigation module, which is used for the drone navigation system to determine the charging base station with the minimum arrival cost as the target charging base station based on the docking signal using the Dijkstra algorithm, and at the same time activate the first navigation module to guide the drone to the vicinity of the target charging base station;

[0056] The drone is docked with a second navigation module, and when the drone reaches a first preset range from the target charging base station, the drone navigation system activates the second navigation module to guide the drone to above the target charging base station;

[0057] The drone is docked with a third navigation module, and when the drone reaches a second preset range above the target charging base station, the drone navigation system activates the third navigation module to guide the drone to the target charging position of the charging base station;

[0058] The UAV docking module is used to connect the UAV to the charging base station using the docking module provided on the UAV.

[0059] The drone charging monitoring module is used to match the drone with the charging base station via Bluetooth. After the Bluetooth matching is successful, the drone will send a charging signal to the charging base station, and the charging base station will charge the drone's battery pack via wireless power transmission. During the charging process, the drone will continuously monitor the battery level. When the battery is fully charged, the drone will send a stop charging signal to the charging base station.

[0060] The drone return module is used for the drone to return to the return location with the minimum arrival cost according to the recorded navigation path sequence combined with the Dijkstra algorithm and continue to perform the line patrol mission.

[0061] In a third aspect, an embodiment of the present application provides an electronic device, characterized by including:

[0062] processor;

[0063] a memory for storing processor-executable instructions;

[0064] Among them, the processor is configured to implement the drone intelligent inspection method docking with the intelligent management and control platform as described in the first aspect when executing the instructions.

[0065] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, characterized in that the computer-readable storage medium stores a program, and the program instructs a device to execute the drone intelligent inspection method docked with the intelligent management and control platform as described in the first aspect.

[0066] The present application provides a method and system for intelligent inspection of drones docked with an intelligent management and control platform. The drone monitors the battery voltage in real time during the inspection process. When the battery voltage is lower than a set voltage threshold, a docking signal is sent. The drone navigation system determines the charging base station with the minimum arrival cost as the target charging base station by the Dijkstra algorithm based on the docking signal, and simultaneously activates the first navigation module to guide the drone to the vicinity of the target charging base station. When the drone arrives within a first preset range from the target charging base station, the drone navigation system activates the second navigation module to guide the drone above the target charging base station. When the drone arrives within a second preset range above the target charging base station, The drone's navigation system activates the third navigation module to guide the drone to the target charging location of the charging base station; the docking module set on the drone is used to complete the docking with the charging base station; after the docking is completed, the drone will perform Bluetooth matching with the charging base station. After the Bluetooth matching is successful, the drone will send a charging signal to the charging base station, and the charging base station will charge the drone's battery pack through wireless power transmission; during the charging process, the drone will continuously monitor the battery level. When the battery is full, the drone will send a stop charging signal to the charging base station; the drone will determine the return location with the minimum arrival cost based on the recorded navigation path sequence combined with the Dijkstra algorithm, and then return to continue the line patrol mission.

[0067] Beneficial effects:

[0068] (1) Through the multi-level positioning strategy, the charging platform can be accurately and automatically located for charging, achieving unmanned independent operation around the clock.

[0069] (2) The Dijkstra algorithm is used to determine the charging base station with the minimum arrival cost and the return location, which provides strong support for the smooth charging process while saving energy consumption.

[0070] (3) The multimodal neural network image vision navigation module is constructed to process multimodal input data and output motor control parameters to achieve precise positioning and smooth landing. The powerful data fusion capability and optimization algorithm of the multimodal neural network are utilized, and a variety of sensor data and advanced control strategies are combined to ensure the efficiency and accuracy of the navigation task.

[0071] (4) It is easy to promote and apply, and the operation is relatively simple, making it suitable for promotion and application in more application scenarios. BRIEF DESCRIPTION OF THE DRAWINGS

[0072] Figure 1 A flowchart of a method for intelligent inspection of drones connected to an intelligent management and control platform provided in one embodiment of the present application.

[0073] Figure 2 A flowchart of navigation using the Dijkstra algorithm and the first navigation module provided in one embodiment of the present application.

[0074] Figure 3 A schematic diagram of a drone intelligent inspection system module connected to an intelligent management and control platform provided in one embodiment of the present application.

[0075] Figure 4 A schematic diagram of an electronic device provided in one embodiment of the present application. DETAILED DESCRIPTION

[0076] The technical solutions in the embodiments of the present application will be described clearly and completely below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments.

[0077] It should be noted that, in the embodiments of the present application, "at least one" refers to one or more, and "more" refers to two or more. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art in the art to which this application relates. The terms used in the specification of this application are only for the purpose of describing specific embodiments and are not intended to limit this application.

[0078] It should be noted that, in the embodiments of the present application, words such as "first" and "second" are only used for the purpose of distinguishing descriptions, and cannot be understood as indicating or implying relative importance, nor can they be understood as indicating or implying an order. Features defined as "first" and "second" may explicitly or implicitly include one or more of the features. In the description of the embodiments of the present application, words such as "exemplary" or "for example" are used to indicate examples, illustrations or explanations. Any embodiment or design scheme described as "exemplary" or "for example" in the embodiments of the present application should not be interpreted as being more preferred or more advantageous than other embodiments or design schemes. Specifically, the use of words such as "exemplary" or "for example" is intended to present related concepts in a concrete way.

[0079] Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0080] Example 1

[0081] The present application provides a method and system for intelligent inspection of unmanned aerial vehicles that is docked with an intelligent management and control platform. The method sends a docking signal by monitoring the voltage of the battery in real time; the charging base station with the minimum arrival cost is determined as the target charging base station by the Dijkstra algorithm, and the drone is guided to the target charging position of the target charging base station by the multi-level navigation module; the docking with the charging base station is completed and charging is performed. When the battery is fully charged, the return position with the minimum arrival cost is determined according to the recorded navigation path sequence combined with the Dijkstra algorithm, and the line patrol task is continued. Through the multi-level positioning strategy, the charging position of the charging platform can be accurately and automatically located for charging, realizing all-weather unmanned independent operation. The charging base station and return position with the minimum arrival cost are determined by the Dijkstra algorithm, saving energy. The multimodal input data is processed by the constructed multimodal neural network image vision navigation module, and the motor control parameters are output to achieve precise positioning and smooth landing.

[0082] Figure 1 A flowchart of a method for intelligent inspection of drones connected to an intelligent management and control platform provided in one embodiment of the present application.

[0083] like Figure 1 As shown, a drone intelligent inspection method connected to an intelligent management and control platform includes:

[0084] S1. Low voltage detection: The drone monitors the battery voltage in real time during the inspection process. When the battery voltage is lower than the set voltage threshold, a docking signal is sent.

[0085] It is understood that in this embodiment, the S1. low voltage detection: the drone monitors the battery voltage in real time during the inspection process, and sends a docking signal when the battery voltage is lower than the set voltage threshold, including:

[0086] Use high-frequency sampling to collect battery load voltage in real time;

[0087] Use hardware filtering circuit combined with software sliding window algorithm to eliminate voltage fluctuation interference;

[0088] A voltage-to-power mapping table is established through a dynamic threshold strategy, and an early warning method is established based on the mapping table, including: Level 1 warning: trigger status prompt, Level 2 warning: forced start of return procedure, and emergency protection: immediate execution of on-site landing;

[0089] The intelligent decision-making module evaluates the warning method, the estimated remaining battery life, the terrain complexity of the return path, the backup power status, and the mission criticality level to determine whether charging or returning is required;

[0090] When it is determined that charging is required, a docking signal is sent.

[0091] Specifically, the drone monitors battery voltage in real time during inspections. When the battery voltage falls below a set threshold, a docking signal is triggered. This feature ensures that the drone can promptly return to the charging station for recharging when the battery is low, thus avoiding mission interruptions due to battery depletion. For example, the drone uses a preset battery alarm threshold to trigger a return command. When the battery level falls below the set value, the drone will automatically return to the charging platform to complete charging. Furthermore, the drone monitors the remaining battery charge and compares it to a set threshold. When the battery level falls below the threshold, it will issue an alarm signal or execute a return operation.

[0092] S2. UAV docking first navigation: The UAV navigation system determines the charging base station with the minimum arrival cost as the target charging base station based on the docking signal using the Dijkstra algorithm, and simultaneously activates the first navigation module to guide the UAV to the vicinity of the target charging base station.

[0093] Figure 2 This is a flowchart of navigation by the Dijkstra algorithm and the first navigation module provided in one embodiment of the present application. Figure 2 As shown, it can be understood that, in this embodiment, the S2. UAV docking first navigation: the UAV navigation system determines the charging base station with the minimum arrival cost as the target charging base station based on the docking signal by the Dijkstra algorithm, and simultaneously activates the first navigation module to guide the UAV to the vicinity of the target charging base station, including:

[0094] S2.1, initialization: set the distance of the starting point to 0, the distance of all other nodes to infinity, and create an unvisited node set U containing all nodes;

[0095] S2.2, select the node with the minimum distance: select the node u that is currently closest to the starting point from the unvisited node set U, and move the node u from the unvisited set U to the visited set S;

[0096] S2.3, Update the distances of adjacent nodes: Traverse all adjacent nodes v of node u and calculate the distances to these adjacent nodes through node u; if the distance to adjacent node v through node u is less than the currently recorded distance, update the distance of v. After the update, add these adjacent nodes back to the set of unvisited nodes U;

[0097] S2.4, repeat steps S2.2 and S2.3: repeatedly select the minimum distance node and update the distances of its adjacent nodes until the unvisited node set U becomes empty or all nodes are processed;

[0098] S2.5, output result: the final distance array D, where d[i] represents the shortest path length from the starting point to node i, and the predecessor node is recorded to restore the specific shortest path;

[0099] The node set includes charging stations and key points on the path, and the edge weight is a cost function. The cost function formula is:

[0100] Cost total =α×E+β×T+γ×P+λ×T2;

[0101] E=(d×e_base×w_factor×a_factor)+e_hover×t_hover;

[0102]

[0103] Where d is the total flight distance, e_base is the basic energy consumption per unit distance, w_factor is the weather correction factor, w_factor is the altitude correction factor, e_hover is the energy consumption per unit time of hovering, t_hover is the expected hovering waiting time, E is the energy cost, T is the task delay penalty, P is the priority cost, and α, β, γ are the corresponding weight coefficients;

[0104] Use Dijkstra algorithm to optimize path planning to ensure the shortest and safest route;

[0105] S2.6, the UAV navigation system determines the priority of the navigation module based on the detection distance, wherein the priority includes a first priority, a second priority, and a third priority;

[0106] S2.7, when the priority is the first priority, matching the first navigation module;

[0107] S2.8. Start a first navigation module to guide the drone to the vicinity of the target charging base station, where the first navigation module is a GPS.

[0108] Specifically, after receiving the docking signal, the drone's navigation system uses the Dijkstra algorithm to determine the location of the target charging base station and activates the first navigation module to guide the drone to the vicinity of the target charging base station. This process involves path planning and navigation control. The drone uses GPS to locate its own position and uses it to locate and navigate charging platforms within the detection range of the UWB base station. The drone's flight control system automatically searches for and lands at the wireless charging base station, completing docking.

[0109] S3. UAV docking second navigation: When the UAV arrives within the first preset range from the target charging base station, the UAV navigation system activates the second navigation module to guide the UAV to the top of the target charging base station.

[0110] It is understood that, in this embodiment, the second navigation of the drone docking is as follows: when the drone arrives within the first preset range from the target charging base station, the second navigation module is activated by the drone navigation system to guide the drone to the top of the target charging base station, including:

[0111] Determining whether the first preset range meets the detection distance corresponding to the second priority;

[0112] When the conditions are met, the second navigation module is matched;

[0113] The second navigation module is activated by the drone navigation system to guide the drone to a preset position above the target charging base station; wherein the second navigation module is a GPS and ultrasonic combined navigation module.

[0114] Specifically, when the drone reaches the target charging station (within a first preset range), the navigation system activates the second navigation module to guide the drone above the charging station. This step further refines the drone's precise positioning capabilities, facilitating subsequent adjustments to ensure proper docking when the drone approaches the charging station.

[0115] S4. UAV docking third navigation: When the UAV reaches the second preset range above the target charging base station, the UAV navigation system activates the third navigation module to guide the UAV to the target charging position of the charging base station.

[0116] It is understood that, in this embodiment, the third navigation for drone docking in S4: when the drone arrives within the second preset range above the target charging base station, the drone navigation system activates the third navigation module to guide the drone to the target charging position of the charging base station, including:

[0117] Determining whether the second preset range meets the detection distance corresponding to the third priority;

[0118] When satisfied, match the third navigation module;

[0119] The drone navigation system activates a third navigation module to guide the drone to a target charging location of the charging base station; wherein the third navigation module is an image vision navigation module;

[0120] The image vision navigation module is a multimodal neural network architecture, which processes multimodal input data through the multimodal neural network architecture to output motor control parameters, and controls the motor according to the control parameters to achieve accurate positioning and smooth landing, wherein the second preset range is smaller than the first preset range.

[0121] Specifically, the multimodal input data is processed by a multimodal neural network architecture to output motor control parameters, including:

[0122] Constructing a multimodal neural network architecture, the multimodal neural network architecture comprising an input layer, a modality-specific encoder, and a multimodal fusion module;

[0123] The drone's acquisition module collects visual image data, the drone's attitude angle, height, and position information.

[0124] The visual image data is segmented through a multimodal neural network architecture to obtain the outer contour of the target charging platform, and the charging location area is further identified. The relative position is calculated by combining the height and position information, the attitude angle of the drone is adjusted, and the motor control parameters are output.

[0125] Specifically, when the drone reaches the target charging base station (the second preset range), the navigation system activates the third navigation module to guide the drone to the target charging location of the charging base station. This step further improves the drone's ability to precisely control space. Through multi-information collaborative positioning technology, the drone can gradually move from coarse positioning to millimeter-level precision positioning.

[0126] Specifically, the image vision navigation module processes multimodal input data through a multimodal neural network and outputs motor control parameters to achieve precise positioning and smooth landing. This process relies on the powerful data fusion capabilities and optimization algorithms of the multimodal neural network, combining multiple sensor data with advanced control strategies to ensure efficient and accurate navigation tasks.

[0127] Specifically, the multimodal neural network architecture consists of an input layer, a modality-specific encoder, and a multimodal fusion module. This architecture can integrate data from different sensors, such as visual image data, the drone's attitude angle, altitude, and position information, thereby improving the accuracy and robustness of drone navigation.

[0128] Drones use their built-in sensors (such as cameras, IMUs, and GPS) to acquire multimodal data in real time, including visual image data, attitude angles, altitude, and position information. After preprocessing (such as filtering, dimensionality reduction, and normalization), these data are converted into training and test sets for subsequent model training and verification.

[0129] In a multimodal neural network, data from each modality is passed through a specific encoder for feature extraction. For example, spatial features can be extracted from visual image data using a convolutional neural network (CNN), while attitude angles, height, and position information can be extracted as time series features using a fully connected layer or recurrent neural network (RNN). These features are then integrated in a multimodal fusion module to generate a unified representation using a self-attention mechanism or other fusion methods.

[0130] A multimodal neural network is used to segment visual image data, identify the outer contours of the target charging platform, and further determine the charging location area. Combining altitude and position information, the relative position between the drone and the target is calculated, and the drone's attitude angle is adjusted to achieve precise positioning.

[0131] Based on the fused multimodal features, the multimodal neural network outputs motor control parameters. These parameters are used to adjust the drone's flight state (such as speed, direction, and attitude) to ensure that the drone can land smoothly at the target location.

[0132] To improve model performance, a random forest algorithm can be used to optimize the combination of multiple teacher models and verify the effectiveness of the model using a test set. In addition, a lightweight student network can be used to further reduce computational complexity and thus improve operational efficiency.

[0133] S5. UAV docking: Use the docking module set on the UAV to complete the docking with the charging base station.

[0134] It is understood that in this embodiment, the docking module provided on the drone is used to achieve docking with the charging base station. The drone achieves precise docking with the charging battery holder through the auxiliary battery exchange module and positioning device. The drone adjusts its posture during the docking process to ensure the contact point is accurate.

[0135] S6. Drone Charging Monitoring: After docking is complete, the drone will perform Bluetooth pairing with the charging base station. Once Bluetooth pairing is successful, the drone will send a charging signal to the charging base station, and the charging base station will charge the drone's battery pack via wireless power transmission. During the charging process, the drone will continuously monitor the battery level. When the battery is fully charged, the drone will send a stop charging signal to the charging base station.

[0136] It will be appreciated that in this embodiment, after docking is complete, the drone will pair with the charging base station via Bluetooth and transmit a charging signal. The charging base station charges the drone's battery pack via wireless power transmission. During the charging process, the drone continuously monitors the battery level and transmits a stop signal when the battery is fully charged. The drone charges using wireless electromagnetic induction wireless power transmission technology and monitors the battery status in real time. During the charging process, current and voltage are monitored in real time to ensure safety and efficiency.

[0137] S7. Return of the UAV: The UAV returns to the location with the minimum arrival cost determined by the recorded navigation path sequence combined with the Dijkstra algorithm and continues to perform the line patrol mission.

[0138] It is understood that, in this embodiment, the step S7. Return of the UAV: The UAV determines the return position with the minimum arrival cost according to the recorded navigation path sequence combined with the Dijkstra algorithm, and then returns to continue the line patrol mission, including:

[0139] Use Dijkstra's algorithm to determine the return path with the minimum arrival cost, and record D', which represents the shortest path length from the X end point to node j;

[0140] Compare the minimum arrival costs consumed by array D' and array D, select the shortest path corresponding to the smaller minimum arrival cost array to perform the return trip, and continue to perform the line patrol task.

[0141] During the inspection process, the drone achieves a fully automated inspection and charging process through real-time monitoring of battery voltage, precise navigation docking, wireless power transmission charging, and return route planning. The combination of these functions and technologies not only improves inspection efficiency but also effectively solves the problem of insufficient battery life.

[0142] Example 2

[0143] like Figure 3As shown, the present application provides a drone intelligent inspection system docked with an intelligent management and control platform, which is applied to the drone intelligent inspection method docked with an intelligent management and control platform as described in Example 1, including: a low voltage detection module 11, a drone docking first navigation module 12, a drone docking second navigation module 13, a drone docking third navigation module 14, a drone docking module 15, a drone charging monitoring module 16, and a drone return module 17.

[0144] It can be understood that, in this embodiment, the low voltage detection module 11 is used for the drone to monitor the battery voltage in real time during the inspection process, and to send a docking signal when the battery voltage is lower than a set voltage threshold;

[0145] It can be understood that in this embodiment, the drone is docked with the first navigation module 12, and the drone navigation system is used to determine the charging base station with the minimum arrival cost as the target charging base station based on the docking signal by the Dijkstra algorithm, and at the same time start the first navigation module to guide the drone to the vicinity of the target charging base station;

[0146] It can be understood that, in this embodiment, the drone is docked with the second navigation module 13, so that when the drone arrives within the first preset range from the target charging base station, the drone navigation system activates the second navigation module to guide the drone to above the target charging base station;

[0147] It can be understood that, in this embodiment, the drone is docked with the third navigation module 14, so that when the drone arrives within the second preset range above the target charging base station, the drone navigation system activates the third navigation module to guide the drone to the target charging position of the charging base station;

[0148] It can be understood that, in this embodiment, the drone docking module 15 is used to complete the docking with the charging base station by utilizing the docking module provided on the drone;

[0149] It is understood that in this embodiment, the drone charging monitoring module 16 is used to perform Bluetooth matching between the drone and the charging base station after the docking is completed. After the Bluetooth matching is successful, the drone will send a charging signal to the charging base station, and the charging base station will charge the drone battery pack by wireless power transmission; during the charging process, the drone will continuously monitor the battery power. When the battery is fully charged, the drone will send a stop charging signal to the charging base station;

[0150] It can be understood that, in this embodiment, the drone return module 17 is used for the drone to return to the return position with the minimum arrival cost determined by the Dijkstra algorithm according to the recorded navigation path sequence, and continue to perform the line patrol mission.

[0151] The present application provides a method and system for intelligent inspection of unmanned aerial vehicles that is docked with an intelligent management and control platform. The method sends a docking signal by monitoring the voltage of the battery in real time; the charging base station with the minimum arrival cost is determined as the target charging base station by the Dijkstra algorithm, and the drone is guided to the target charging position of the target charging base station by the multi-level navigation module; the docking with the charging base station is completed and charging is performed. When the battery is fully charged, the return position with the minimum arrival cost is determined according to the recorded navigation path sequence combined with the Dijkstra algorithm, and the line patrol task is continued. Through the multi-level positioning strategy, the charging position of the charging platform can be accurately and automatically located for charging, realizing all-weather unmanned independent operation. The charging base station and return position with the minimum arrival cost are determined by the Dijkstra algorithm, saving energy. The multimodal input data is processed by the constructed multimodal neural network image vision navigation module, and the motor control parameters are output to achieve precise positioning and smooth landing.

[0152] Figure 4 This is an electronic device provided by an embodiment of the present application. Figure 4 As shown, the electronic device includes at least the following parts: a processor 101 and a memory 100 , a communication interface 103 , and a bus 102 .

[0153] In the embodiment of the present application, the memory 100 is used to store instructions executable by the processor 101. The processor 101 is configured to execute the instructions to implement the following Figure 3 The device module shown is for drone intelligent inspection, which is connected to the intelligent management and control platform.

[0154] In an embodiment of the present application, a computer-readable storage medium includes instructions, and the instructions instruct a device to execute the method of the first aspect. For example, the instructions instruct the device to execute Figure 1 The method is shown in the process steps.

[0155] The program running in the electronic device involved in one embodiment of the present application can be a program that controls a central processing unit (CPU) and the like to realize the functions of the above-mentioned embodiment involved in one embodiment of the present invention (a program that enables a computer to function). Then, the information processed by these devices is temporarily stored in a random access memory (RAM) during its processing, and then stored in various ROMs such as read-only memory (Flash ROM) and hard disk drive (HDD), and is read, modified, and written by the CPU as needed.

[0156] It should be noted that a portion of the electronic device of the above embodiment may also be implemented by a computer. In this case, a program for implementing the control function may be recorded on a computer-readable recording medium, and the program recorded on the recording medium may be read into a computer and executed.

[0157] It should be noted that the "computer" mentioned here refers to a computer built into an electronic device, employing hardware including an operating system (OS) and peripheral devices. Furthermore, "computer-readable recording medium" refers to removable media such as floppy disks, magneto-optical disks, ROMs, and CD-ROMs, as well as storage devices such as hard disks built into computers.

[0158] Furthermore, "computer-readable recording media" may include: media that dynamically store programs for a short period of time, such as communication lines when transmitting programs via networks such as the Internet or communication lines such as telephone lines; and media that store programs for a fixed period of time, such as volatile memory within computers acting as servers or clients in this context. Furthermore, the aforementioned program may be a program for implementing a portion of the aforementioned functions, or a program that can achieve the aforementioned functions by combining with a program already stored in a computer.

[0159] Furthermore, the electronic device in the above-described embodiments can also be implemented as a collection (device group) consisting of multiple devices. Each device constituting the device group may have a portion or all of the functions or functional blocks of the electronic device in the above-described embodiments. A device group only needs to have all the functions or functional blocks of the electronic device.

[0160] Those skilled in the art should recognize that the above embodiments are merely intended to illustrate the present application and are not intended to limit the present application. As long as they are within the spirit of the present application, appropriate changes and modifications to the above embodiments are within the scope of protection claimed in the present application.

Claims

1. A drone intelligent inspection method connected to an intelligent management and control platform, characterized in that: The method comprises: S1. Low voltage detection: The drone monitors the battery voltage in real time during the inspection process. When the battery voltage falls below the set voltage threshold, a docking signal is issued; S2. UAV docking first navigation: The UAV navigation system determines the charging base station with the minimum arrival cost as the target charging base station based on the docking signal using the Dijkstra algorithm, and simultaneously activates the first navigation module to guide the UAV to the vicinity of the target charging base station; S3. UAV docking second navigation: When the UAV reaches the first preset range from the target charging base station, the UAV navigation system starts the second navigation module to guide the UAV to the target charging base station above; S4. UAV docking third navigation: When the UAV reaches the second preset range above the target charging base station, the UAV navigation system activates the third navigation module to guide the UAV to the target charging position of the charging base station; S5. UAV docking: Use the docking module on the drone to complete docking with the charging base station; S6. Drone Charging Monitoring: After docking is complete, the drone will pair with the charging base station via Bluetooth. Once the Bluetooth pairing is successful, the drone will send a charging signal to the charging base station, which will then charge the drone's battery pack via wireless power transmission. During the charging process, the drone will continuously monitor the battery level. When the battery is fully charged, the drone will send a stop charging signal to the charging base station. S7. Return of the UAV: The UAV returns to the location with the minimum arrival cost determined by the recorded navigation path sequence combined with the Dijkstra algorithm and continues to perform the line patrol mission.

2. The method for intelligent inspection of unmanned aerial vehicles connected to an intelligent management and control platform according to claim 1 is characterized in that: S1. Low voltage detection: The drone monitors the battery voltage in real time during the inspection process. When the battery voltage is lower than the set voltage threshold, it sends a docking signal, including: Use high-frequency sampling to collect battery load voltage in real time; Use hardware filtering circuit combined with software sliding window algorithm to eliminate voltage fluctuation interference; A voltage-to-power mapping table is established through a dynamic threshold strategy, and an early warning method is established based on the mapping table, including: Level 1 warning: trigger status prompt, Level 2 warning: forced start of return procedure, and emergency protection: immediate execution of on-site landing; The intelligent decision-making module evaluates the warning method, the estimated remaining battery life, the terrain complexity of the return path, the backup power status, and the mission criticality level to determine whether charging or returning is required; When it is determined that charging is required, a docking signal is sent.

3. The method for intelligent inspection of unmanned aerial vehicles connected to an intelligent management and control platform according to claim 2 is characterized in that: S2. UAV docking first navigation: The UAV navigation system determines the charging base station with the minimum arrival cost as the target charging base station based on the docking signal using the Dijkstra algorithm, and simultaneously activates the first navigation module to guide the UAV to the vicinity of the target charging base station, including: S2.1, initialization: set the distance of the starting point to 0, the distance of all other nodes to infinity, and create an unvisited node set U containing all nodes; S2.2, select the node with the minimum distance: select the node u that is currently closest to the starting point from the unvisited node set U, and move the node u from the unvisited set U to the visited set S; S2.3, Update the distances of adjacent nodes: Traverse all adjacent nodes v of node u and calculate the distances to these adjacent nodes through node u; if the distance to adjacent node v through node u is less than the currently recorded distance, update the distance of v. After the update, add these adjacent nodes back to the set of unvisited nodes U; S2.4, repeat steps S2.2 and S2.3: repeatedly select the minimum distance node and update the distances of its adjacent nodes until the unvisited node set U becomes empty or all nodes are processed; S2.5, output result: the final distance array D, where d[i] represents the shortest path length from the starting point to node i, and the predecessor node is recorded to restore the specific shortest path; The node set includes charging stations and key points on the path, and the edge weight is a cost function. The cost function formula is: Cost total =α×E+β×T+γ×P+λ×T2; E=(d×e_base×w_factor×a_factor)+e_hover×t_hover; Where d is the total flight distance, e_base is the basic energy consumption per unit distance, w_factor is the weather correction factor, w_factor is the altitude correction factor, e_hover is the energy consumption per unit time of hovering, t_hover is the expected hovering waiting time, E is the energy cost, T is the task delay penalty, P is the priority cost, and α, β, γ are the corresponding weight coefficients; Use Dijkstra algorithm to optimize path planning to ensure the shortest and safest route; S2.6, the UAV navigation system determines the priority of the navigation module based on the detection distance, wherein the priority includes a first priority, a second priority, and a third priority; S2.7, when the priority is the first priority, matching the first navigation module; S2.

8. Start a first navigation module to guide the drone to the vicinity of the target charging base station, where the first navigation module is a GPS.

4. The method for intelligent inspection of unmanned aerial vehicles connected to an intelligent management and control platform according to claim 3 is characterized in that: The S3. UAV docking second navigation: When the UAV arrives within the first preset range from the target charging base station, the UAV navigation system activates the second navigation module to guide the UAV to the top of the target charging base station, including: Determining whether the first preset range meets the detection distance corresponding to the second priority; When the conditions are met, the second navigation module is matched; The second navigation module is activated by the drone navigation system to guide the drone to a preset position above the target charging base station; wherein the second navigation module is a GPS and ultrasonic combined navigation module.

5. The method for intelligent inspection of unmanned aerial vehicles connected to an intelligent management and control platform according to claim 2 is characterized in that: Said S4. UAV docking third navigation: When the UAV reaches the second preset range above the target charging base station, the UAV navigation system activates the third navigation module to guide the UAV to the target charging position of the charging base station, including: Determining whether the second preset range meets the detection distance corresponding to the third priority; When satisfied, match the third navigation module; The drone navigation system activates a third navigation module to guide the drone to a target charging location of the charging base station; wherein the third navigation module is an image vision navigation module; The image vision navigation module is a multimodal neural network architecture, which processes multimodal input data through the multimodal neural network architecture to output motor control parameters, and controls the motor according to the control parameters to achieve accurate positioning and smooth landing, wherein the second preset range is smaller than the first preset range.

6. The method for intelligent inspection of a drone connected to an intelligent management and control platform according to claim 5 is characterized in that: The method of processing multimodal input data and outputting motor control parameters through a multimodal neural network architecture includes: Constructing a multimodal neural network architecture, the multimodal neural network architecture comprising an input layer, a modality-specific encoder, and a multimodal fusion module; The drone's acquisition module collects visual image data, the drone's attitude angle, height, and position information. The visual image data is segmented through a multimodal neural network architecture to obtain the outer contour of the target charging platform, and the charging location area is further identified. The relative position is calculated by combining the height and position information, the attitude angle of the drone is adjusted, and the motor control parameters are output.

7. The method for intelligent inspection of unmanned aerial vehicles connected to an intelligent management and control platform according to claim 2, characterized in that: S7. Return of the UAV: The UAV determines the return location with the minimum arrival cost based on the recorded navigation path sequence and the Dijkstra algorithm, and then returns to continue the line patrol mission, including: Use Dijkstra's algorithm to determine the return path with the minimum arrival cost, and record D', which represents the shortest path length from the X end point to node j; Compare the minimum arrival costs consumed by array D' and array D, select the shortest path corresponding to the smaller minimum arrival cost array to perform the return trip, and continue to perform the line patrol task.

8. An intelligent inspection system for unmanned aerial vehicles connected to an intelligent management and control platform, applied to the intelligent inspection method for unmanned aerial vehicles connected to an intelligent management and control platform as claimed in any one of claims 1 to 7, characterized in that: include: A low voltage detection module is used to monitor the battery voltage of the drone in real time during the inspection process. When the battery voltage is lower than the set voltage threshold, a docking signal is issued; The drone docks with the first navigation module, which is used for the drone navigation system to determine the charging base station with the minimum arrival cost as the target charging base station based on the docking signal using the Dijkstra algorithm, and at the same time activate the first navigation module to guide the drone to the vicinity of the target charging base station; The drone is docked with a second navigation module, and when the drone reaches a first preset range from the target charging base station, the drone navigation system activates the second navigation module to guide the drone to above the target charging base station; The drone is docked with a third navigation module, and when the drone reaches a second preset range above the target charging base station, the drone navigation system activates the third navigation module to guide the drone to the target charging position of the charging base station; The UAV docking module is used to connect the UAV to the charging base station using the docking module provided on the UAV. The drone charging monitoring module is used to match the drone with the charging base station via Bluetooth. After the Bluetooth matching is successful, the drone will send a charging signal to the charging base station, and the charging base station will charge the drone's battery pack via wireless power transmission. During the charging process, the drone will continuously monitor the battery level. When the battery is fully charged, the drone will send a stop charging signal to the charging base station. The drone return module is used for the drone to return to the return location with the minimum arrival cost according to the recorded navigation path sequence combined with the Dijkstra algorithm and continue to perform the line patrol mission.

9. An electronic device, characterized in that: include: processor; a memory for storing processor-executable instructions; The processor is configured to implement the intelligent inspection method for a drone connected to an intelligent management and control platform as described in any one of claims 1 to 7 when executing the instructions.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a program, and the program instructs the device to execute the drone intelligent inspection method docked with the intelligent management and control platform as described in any one of claims 1 to 7.

Images