Anti-explosion method, device, equipment and storage medium for tethered drone

By combining active and passive anti-crash modes, and utilizing tethered cable retraction equipment and safety airbags, the problem of safe recovery of tethered drones in stall situations has been solved, thus improving the safety and reliability of tethered drones.

CN118419296BActive Publication Date: 2024-11-22HUBEI CENT CHINA TECH DEV OF ELECTRIC POWER +1
View PDF 7 Cites 0 Cited by

Patent Information

Application Number
CN202410593723.0
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-05-14
Publication Date
2024-11-22
Estimated Expiration
2044-05-14

AI Technical Summary

Technical Problem

Existing tethered drones are prone to stalling during missions due to hardware failures, software malfunctions, operational errors, and external environmental factors. The lack of effective protective measures makes crashes inevitable.

Method used

The system employs a combination of active and passive anti-crash modes. The active anti-crash mode ensures security through a lock-in mode and unlock password verification, while the passive anti-crash mode calculates stall data by acquiring the drone's pose information and image data in real time, controls the tether line retraction device to perform the line retraction operation, and uses a catapult device to safely recover the drone.

Benefits of technology

It improves the safety and reliability of tethered drones in stall situations, ensures the safe recovery of drones in emergency situations, and reduces property damage and safety accidents.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118419296B_ABST
    Figure CN118419296B_ABST
Patent Text Reader

Abstract

The present invention provides a bomb-proofing method, device, equipment and storage medium for a tethered drone, the method comprising: controlling the drone device to enter an active bomb-proofing mode and a passive bomb-proofing mode; the active bomb-proofing mode comprises: when the drone device reaches the target position, controlling the drone device to enter a locking mode, in which the drone device stops receiving external signals; the passive bomb-proofing mode comprises: calculating the stall data of the drone device based on the posture information of the drone device and the real-time image taken, and then calculating the retraction and extension speed of the tethered line retraction and extension device, and controlling the tethered line retraction and extension device to retract the tethered line according to the retraction and extension speed; controlling the ejection device on the drone device to eject the airbag and inflate it, so as to achieve safe recovery of the drone device. The method improves the safety and reliability of the tethered drone through active and passive bomb-proofing modes and using the tethered line as a recovery tool for the drone in an emergency.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of tethered unmanned aerial vehicles, and in particular to an anti-explosion method, device, equipment and storage medium for a tethered unmanned aerial vehicle. Background Art

[0002] With the rapid development of drone technology, drones have been widely used in military reconnaissance, aerial photography, topographic mapping, traffic monitoring and other fields. In particular, tethered drones can obtain electricity directly from the ground through tethered lines to achieve long-term flight, which is particularly suitable for tasks such as long-term monitoring and communication relay. However, as the scope of application continues to expand, tethered drones face a variety of risks during the execution of their missions, including hardware failures, software failures, operational errors, and stalls caused by external environmental factors. These risks lead to drone crashes, resulting in property losses or safety accidents.

[0003] In the prior art, the safety measures for tethered drones mainly focus on preventing hardware and software failures and improving the accuracy and reliability of operations. For example, redundant design, real-time monitoring of system status, and the use of advanced flight control algorithms are used to reduce the occurrence of failures. However, crashes are still inevitable, and there is a lack of protective measures after the drone stalls. Summary of the invention

[0004] The main purpose of the present invention is to solve the technical problems that the existing tethered UAV anti-explosion machine inevitably explodes and lacks protective measures after the UAV stalls.

[0005] A first aspect of the present invention provides a bomb-proofing method for a tethered drone, the bomb-proofing method for a tethered drone is applied to a bomb-proofing system for a tethered drone, the bomb-proofing system for a tethered drone comprises a drone device, a tethering line, and a ground device that is communicatively connected to the drone device via the tethering line, the ground device comprises a tethering line retracting and releasing device, and the drone device is equipped with a camera device;

[0006] The anti-explosion method of the tethered UAV includes:

[0007] When the UAV device is performing aerial operations, the UAV device is controlled to simultaneously enter an active anti-bombing mode and a passive anti-bombing mode;

[0008] The active anti-explosion mode includes:

[0009] When the UAV device reaches the target position based on the control signal sent by the ground device, the UAV device is controlled to enter a locking mode, wherein the UAV device stops receiving external signals in the locking mode;

[0010] The input unlock password input by the ground device is sent to the drone device through the tethered line, and the drone device determines whether the input unlock password is the same as a preset set unlock password;

[0011] If so, the UAV device is controlled to exit the locking mode and receive external signals again to perform aerial operations;

[0012] The passive anti-bombing mode includes:

[0013] Acquire the posture information of the UAV device and the real-time image captured by the camera device in real time, calculate the stall data of the UAV device based on the posture information and the real-time image, and determine whether the UAV device is in a stall state based on the stall data;

[0014] If the UAV device is in a stall phase, the retraction speed of the mooring line retraction device is calculated based on the stall data and the environmental data collected by the UAV device, and the mooring line retraction device is controlled to retract the mooring line according to the retraction speed;

[0015] The ejection device on the drone equipment is controlled to eject the airbag and put the airbag into an inflated state, thereby achieving safe recovery of the drone equipment.

[0016] Optionally, in a first implementation manner of the first aspect of the present invention, the real-time acquisition of the posture information of the drone device and the real-time image captured by the camera device, calculating the stall data of the drone device based on the posture information and the real-time image, and judging whether the drone device is in a stall state based on the stall data includes:

[0017] Acquire the posture information of the drone device and the real-time image captured by the camera device in real time, and generate corresponding posture time series data and image time series data according to the time axis of the posture information and the real-time image;

[0018] Inputting the posture time series data and the image time series data into a preset stall detection model, wherein the stall detection model includes an input layer, an attention mechanism layer, a feature fusion layer, a multi-task detection layer and an output layer;

[0019] Preprocessing the posture time series data and the image time series data through the input layer, and performing data feature extraction and image feature extraction on the preprocessed posture time series data and image time series data, respectively, to obtain posture time series features and image time series features, respectively;

[0020] Calculate the first weight vector of the posture temporal feature and the second weight vector of the image temporal feature through the attention mechanism layer;

[0021] Performing feature fusion on the posture time series feature and the image time series feature according to the first weight vector and the second weight vector through the feature fusion layer to obtain a fused feature vector;

[0022] Performing multi-task detection according to the fused feature vector through a random forest network in the multi-task detection layer to obtain multiple types of stall data, and outputting the multiple types of stall data through the output layer;

[0023] Determine whether the drone device is in a stall state according to the multiple types of stall data.

[0024] Optionally, in a second implementation of the first aspect of the present invention, the posture time series data and the image time series data are preprocessed by the input layer, and data feature extraction and image feature extraction are performed on the preprocessed posture time series data and image time series data, respectively, to obtain posture time series features and image time series features, respectively, including:

[0025] Performing data filtering and data denoising on the posture time series data and the image time series data;

[0026] Performing standard deviation threshold calculation on the posture time series data after data filtering and data denoising to obtain the standard deviation threshold of the posture time series data, and removing outliers from the posture time series data after data filtering and data denoising by using the standard deviation threshold;

[0027] Data feature extraction and image feature extraction are performed on the posture time series data after outliers are removed, and the image time series data after data filtering and data denoising, to obtain posture time series features and image time series features, respectively.

[0028] Optionally, in a third implementation of the first aspect of the present invention, the calculating the retracting speed of the mooring line retracting device based on the stall data and the environmental data collected by the drone device, and controlling the mooring line retracting device to retract the mooring line according to the retracting speed includes:

[0029] Calculating the unloading acceleration of the UAV device based on the stall data and the environmental data collected by the UAV device through a preset dynamic model;

[0030] Calculate a predicted descent speed of the UAV device at a next time step based on the descent acceleration and the current speed of the UAV device at the current time step;

[0031] Calculate the retraction and extension speed of the tethered line retraction and extension device according to the predicted descent speed of the UAV device at the next time step and the relative position of the UAV device and the ground device;

[0032] The mooring line is controlled to be reeled in according to the reeling-in speed by the mooring line reeling-in device.

[0033] Optionally, in a fourth implementation of the first aspect of the present invention, after calculating the retracting and releasing speed of the tethering line retracting and releasing device according to the predicted descent speed of the UAV device at the next time step and the relative position of the UAV device and the ground device, the method further includes:

[0034] Obtaining the predicted descent speed of the previous time step and the actual descent speed of the current time step, and calculating the predicted deviation value between the predicted descent speed of the previous time step and the actual descent speed of the current time step;

[0035] Obtaining the stall time of the UAV device in a stall state, and calculating the prediction correction value of the UAV device at the current time step according to the stall time and the prediction deviation value;

[0036] The retraction and extension speed of the mooring line retraction and extension device is adjusted according to the predicted correction value.

[0037] Optionally, in a fifth implementation of the first aspect of the present invention, the passive anti-explosion aircraft mode further includes:

[0038] Performing visual recognition on the real-time image collected by the drone device to determine the target object in the real-time image and the object classification of the target object;

[0039] Determining whether to perform position tracking on the target object according to the object classification;

[0040] If yes, the real-time image at the same time as the position tracking is performed is used as the initial frame image, and the real-time moving trajectory of the target object is generated in real time according to the initial frame image and the real-time images subsequent to the initial frame image;

[0041] generating a predicted trajectory of the target object based on the object classification and the real-time moving trajectory, and determining whether the predicted trajectory coincides with the position of the UAV device;

[0042] If they overlap, the tethering line is controlled to be retracted through the tethering line retracting and releasing device to achieve safe recovery of the UAV equipment.

[0043] Optionally, in a sixth implementation manner of the first aspect of the present invention, the performing visual recognition on the real-time image collected by the drone device to determine the target object in the real-time image and the object classification of the target object includes:

[0044] Performing image preprocessing on the real-time image to obtain a preprocessed image, and inputting the preprocessed image into the image detection model, wherein the image detection model includes an input layer, a feature extraction layer, a feature pyramid network, a feature fusion layer, a multi-task detection layer and an output layer;

[0045] Mapping the preprocessed image to a high-dimensional space through the input layer to obtain initial feature information in the high-dimensional space;

[0046] Extracting features from the initial feature information through the feature extraction layer to obtain convolution features at different levels;

[0047] Generating initial fusion features corresponding to convolutional features at different levels through the feature pyramid network, and fusing the initial fusion features at different levels through the feature fusion layer to obtain secondary fusion features;

[0048] The target object in the real-time image and the object classification of the target object are determined according to the secondary fusion features through the multi-task detection layer and the output layer.

[0049] A second aspect of the present invention provides an anti-explosion device for a tethered drone, the anti-explosion device for a tethered drone is applied to an anti-explosion system for a tethered drone, the anti-explosion system for a tethered drone comprises a drone device, a tethering line, and a ground device connected to the drone device through the tethering line, the ground device comprises a tethering line retracting and releasing device, and the drone device is equipped with a camera device;

[0050] The anti-explosion device of the tethered UAV comprises:

[0051] A control module, used for controlling the UAV device to enter an active anti-bombing mode and a passive anti-bombing mode simultaneously when the UAV device is performing aerial operations;

[0052] The active anti-explosion module is used to perform the following operations when the UAV device is in the active anti-explosion mode:

[0053] When the drone device reaches the target position based on the control signal sent by the ground device, the drone device is controlled to enter a locking mode, wherein the drone device stops receiving external signals in the locking mode; the input unlocking password input by the ground device is sent to the drone device through the tethered line, and the drone device determines whether the input unlocking password is the same as the preset set unlocking password; if so, the drone device is controlled to exit the locking mode and receive external signals again to perform aerial operations;

[0054] The passive anti-explosion module is used to perform the following operations when the UAV device is in the active anti-explosion mode:

[0055] The position information of the UAV device and the real-time image captured by the camera device are acquired in real time, the stall data of the UAV device is calculated based on the position information and the real-time image, and whether the UAV device is in a stall state is determined based on the stall data; if the UAV device is in a stall stage, the retraction and extension speed of the tethering line retraction and extension device is calculated based on the stall data and the environmental data collected by the UAV device, and the tethering line retraction and extension device is controlled to retract the tethering line according to the retraction and extension speed; the ejection device on the UAV device is controlled to eject the airbag and put the airbag into an inflated state, so as to achieve safe recovery of the UAV device.

[0056] A third aspect of the present invention provides an anti-explosion device for a tethered drone, comprising: a memory and at least one processor, wherein instructions are stored in the memory, and the memory and the at least one processor are interconnected via a line; the at least one processor calls the instructions in the memory so that the anti-explosion device for the tethered drone executes the steps of the above-mentioned anti-explosion method for the tethered drone.

[0057] A fourth aspect of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores instructions, which, when executed on a computer, enable the computer to execute the steps of the above-mentioned method for preventing a tethered drone from bombing.

[0058] The above-mentioned anti-explosion method, device, equipment and storage medium of the tethered drone control the drone equipment to enter the active anti-explosion mode and the passive anti-explosion mode; the active anti-explosion mode includes: when the drone equipment reaches the target position, the drone equipment is controlled to enter the locking mode, and the drone equipment stops receiving external signals in the locking mode; the passive anti-explosion mode includes: based on the posture information of the drone equipment and the real-time image taken, the stall data of the drone equipment is calculated, and then the retraction and release speed of the tethered line retraction and release device is calculated, and the tethered line retraction and release device is controlled to retract the tethered line according to the retraction and release speed; the ejection device on the drone equipment is controlled to eject the airbag and inflate it, so as to achieve the safe recovery of the drone equipment. This method improves the safety and reliability of tethered drones through active and passive anti-explosion modes, and uses the tethered line as a recovery tool for drones in emergency situations.

[0059] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention are realized and obtained by the structures particularly pointed out in the description, claims and drawings.

[0060] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, preferred embodiments are given below and described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] Figure 1 It is a schematic diagram of a first embodiment of a method for tethering a drone to prevent bombing in an embodiment of the present invention;

[0062] Figure 2 A schematic diagram of an embodiment of an anti-explosion device for tethering a UAV in an embodiment of the present invention;

[0063] Figure 3 The figure is a schematic diagram of an embodiment of an anti-explosion device for tethering a UAV in an embodiment of the present invention. DETAILED DESCRIPTION

[0064] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0065] The terms "including" and "having" and any variations thereof mentioned in the embodiments of the present invention are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device end including a series of steps or units is not limited to the listed steps or units, but may optionally include other steps or units that are not listed, or may optionally include other steps or units that are inherent to these processes, methods, products or device ends.

[0066] To facilitate understanding of this embodiment, a method for preventing explosions of a tethered drone disclosed in an embodiment of the present invention is first described in detail. The method for preventing explosions of a tethered drone is applied to a system for preventing explosions of a tethered drone. The system for preventing explosions of a tethered drone includes a drone device, a tethering line, and ground equipment that is connected to the drone device through the tethering line. The ground equipment includes a tethering line retracting and releasing device. The drone device is equipped with a camera device. Figure 1 As shown, the method comprises the following steps:

[0067] 101. When the UAV equipment is performing aerial operations, the UAV equipment is controlled to enter the active anti-bombing mode and the passive anti-bombing mode at the same time;

[0068] In one embodiment of the present invention, tethered drones can generally be used to perform patrol and monitoring tasks, such as patrolling and monitoring pipelines, power transmission lines, forests, etc., in order to discover problems or abnormal conditions, or to build temporary communication relay stations to provide temporary communication signal coverage, such as playing an important role in disaster relief, major event security, etc. When the drone equipment is performing aerial operations, it simultaneously performs active anti-bombing mode and passive anti-bombing mode.

[0069] 102. In the active anti-explosion mode, when the UAV device reaches the target position based on the control signal sent by the ground device, the UAV device is controlled to enter the locking mode, wherein the UAV device stops receiving external signals in the locking mode;

[0070] In one embodiment of the present invention, in the active anti-collision mode, when the UAV device reaches the target position based on the control signal sent by the ground device, the UAV device is controlled to enter the locking mode, and the locking mode can realize active anti-explosion, because stopping receiving external signals can reduce interference with the UAV device and avoid the influence of unintentional interference signals or malicious attack signals from the outside on the control of the UAV device. The interference of external signals causes the UAV device to lose control, thereby causing accidents or dangerous situations. In addition, when reaching the target location, the tasks to be completed at the target location can be preset to avoid manual operation, so as to prevent flight abnormalities caused by misoperation or control signal errors. When the UAV performs a specific task, it needs to fly according to the predetermined path and operation. Stopping receiving external signals can ensure that the UAV flies according to the set planned path to ensure flight safety. In addition, in the locking mode, the UAV device tracks and locates the target through the internal system or sensor, which is not interfered by external signals, and improves the accuracy and stability of target locking. In addition, stopping receiving external signals can enhance the communication security of the UAV device and prevent unauthorized access or interference signals from affecting the communication link. By limiting the reception of external signals, the risk of eavesdropping, interference or tampering of the communication link can be reduced, and the confidentiality and integrity of the communication can be maintained.

[0071] 103. Send the input unlock password input by the ground device to the drone device through the tethered line, and determine whether the input unlock password is the same as the preset unlock password through the drone device;

[0072] In one embodiment of the present invention, the unlock password input by the ground device is transmitted to the drone device by a tethered line, and then the drone device compares and confirms whether the input unlock password matches the preset unlock password. This process ensures that the drone device can be unlocked and started only when the input unlock password completely matches the preset password. This security measure is intended to prevent unauthorized operators from manipulating the drone device, thereby ensuring the legality and safety of flight operations. By transmitting the unlock password by a tethered line, reliable transmission of password information can be achieved, reducing the risk of password leakage and external malicious interference. At the same time, the verification process of the unlock password by the drone device is highly autonomous and independent, effectively ensuring the safety and reliability of the entire unlocking process. This unlocking mechanism based on password verification provides a safe and reliable means for ground operators to ensure legal control of drone equipment, effectively prevents safety hazards caused by unauthorized operations, and further improves the flight safety level of drone equipment.

[0073] 104. If yes, the UAV device is controlled to exit the lock mode and receive external signals again to perform aerial operations;

[0074] In one embodiment of the present invention, after receiving the unlock password transmitted by the ground device, the drone device will perform internal decoding processing and compare the input unlock password with the preset unlock password for verification. If the comparison result shows that the two are the same, the drone device confirms that the unlock password is valid. Secondly, after confirming that the unlock password is valid, the drone device will start the corresponding unlock program and execute the operation of exiting the lock mode. This includes stopping the current lock state and restoring the function of receiving external signals so as to re-receive control instructions and operation signals from the ground device. Subsequently, the drone device will start the relevant flight program according to the instructions sent by the ground device or the preset mission plan, including take-off, route planning, flight altitude adjustment and other operations, so as to achieve the smooth execution of the aerial work task. In the process of re-receiving the external signal, the drone device will continue to perform the aerial work task according to the instructions of the ground device or the preset program, ensure the safety and stability of the flight process, and make real-time adjustments and feedback as needed to complete the specified task.

[0075] 105. In the passive anti-bombing mode, the posture information of the UAV device and the real-time image taken by the camera device are obtained in real time, the stall data of the UAV device is calculated based on the posture information and the real-time image, and whether the UAV device is in a stall state is determined based on the stall data;

[0076] In one embodiment of the present invention, the real-time acquisition of the posture information of the UAV device and the real-time image taken by the camera device, the calculation of the stall data of the UAV device based on the posture information and the real-time image, and the judgment of whether the UAV device is in a stall state based on the stall data include: real-time acquisition of the posture information of the UAV device and the real-time image taken by the camera device, and generating corresponding posture timing data and image timing data according to the time axis of the posture information and the real-time image; inputting the posture timing data and the image timing data into a preset stall detection model, wherein the stall detection model includes an input layer, an attention mechanism layer, a feature fusion layer, a multi-task detection layer and an output layer; the input layer performs an operation on the posture timing data and the image timing data; and the image timing data is input into a preset stall detection model. The timing data is preprocessed, and data feature extraction and image feature extraction are performed on the preprocessed posture timing data and image timing data, respectively, to obtain posture timing features and image timing features, respectively; the first weight vector of the posture timing features and the second weight vector of the image timing features are calculated through the attention mechanism layer; the posture timing features and the image timing features are feature fused according to the first weight vector and the second weight vector through the feature fusion layer to obtain a fused feature vector; multi-task detection is performed according to the fused feature vector through the random forest network in the multi-task detection layer to obtain multiple types of stall data, and the multiple types of stall data are output through the output layer; and whether the UAV device is in a stall state is determined according to the multiple types of stall data.

[0077] Specifically, in the process of generating the posture time series data and the image time series data, first, based on the posture information of the drone device acquired in real time, record the position, posture and other parameters at each time point. Arrange the recorded posture information in chronological order to form a posture time series data sequence, which represents the posture state of the drone device at different time points. Acquire the real-time image taken by the camera device in real time, and record the timestamp of each frame. Arrange the images in order according to the timestamps to form an image time series data sequence, which represents the real-time image sequence taken by the drone device at different time points.

[0078] Specifically, in the input layer of the stall detection model, the input layer cleans and normalizes the pose time series data to ensure the consistency and comparability of the data. Then, various feature extraction methods, such as statistical feature extraction and frequency domain feature extraction, can be used to extract representative feature information from the pose time series data. After data feature extraction, the pose time series feature vector representing the change in the pose state of the drone is obtained. The image time series data is preprocessed, including image denoising, size unification and other operations to ensure image quality and consistency. Using deep learning models (such as CNN) or traditional feature extraction methods (such as SIFT, HOG, etc.), high-level semantic features are extracted from each frame of the image to obtain the image time series feature vector.

[0079] Specifically, for the pose temporal features (P) and the image temporal features (I), their correlation matrices are calculated respectively. This can be achieved by calculating the similarity between them, such as using methods such as cosine similarity. The pose temporal feature matrix (P) and the image temporal feature matrix (I) are multiplied by a learnable parameter matrix to obtain new representations, namely P' and I'. For P' and I', their attention weights are calculated respectively. This can be achieved by inputting them into an activation function, such as a softmax function, to obtain normalized weight values. P' and I' are weighted and summed according to their attention weights to obtain a weighted representation of the pose temporal features and a weighted representation of the image temporal features. Finally, these weighted representations can be used in subsequent feature fusion layers to generate a fused feature vector for multi-task detection. In the multi-task detection process, the fused feature vector is input into a random forest network. The random forest network consists of multiple decision trees, each of which makes a prediction on the input fused feature vector. Each decision tree outputs prediction results on multiple types of stall data. The final output will be the integrated result of multiple decision trees, which can obtain comprehensive prediction results for various types of stall data. Among them, the stall data can include the classification of stall causes (such as excessive wind speed, insufficient battery power, sensor failure, etc.), the severity level of the stall (such as mild stall, severe stall, etc.), and by analyzing the posture information and image information, the movement of the aircraft in the air can be inferred, such as changes in pitch angle and roll angle. This information can be used to infer whether the aircraft is in a stall state. Stall duration, stall type, such as curling stall, dive stall, etc.

[0080] Furthermore, the pose time series data and the image time series data are preprocessed through the input layer, and data feature extraction and image feature extraction are performed on the preprocessed pose time series data and image time series data, respectively, to obtain pose time series features and image time series features, respectively, including: data filtering and data denoising are performed on the pose time series data and the image time series data; standard deviation threshold calculation is performed on the pose time series data after data filtering and data denoising to obtain the standard deviation threshold of the pose time series data, and outliers are removed from the pose time series data after data filtering and data denoising by using the standard deviation threshold; data feature extraction and image feature extraction are performed on the pose time series data after outliers are removed and on the image time series data after data filtering and data denoising to obtain pose time series features and image time series features, respectively.

[0081] Specifically, the pose time series data and the image time series data are preprocessed, including data filtering and data denoising. Data filtering can use common filtering algorithms, such as mean filtering, median filtering, Gaussian filtering, etc., to smooth the data and reduce noise. Data denoising can use various denoising algorithms, such as wavelet denoising, mean denoising, median denoising, etc., to eliminate noise in the data. Standard deviation threshold calculation is performed on the pose time series data after data filtering and data denoising. The standard deviation threshold can be calculated based on the statistical characteristics of the data, for example, by calculating the standard deviation of the data to determine the threshold. Through the standard deviation threshold, outliers are removed, which can be achieved by comparing with the threshold. Feature extraction is performed on the image time series data after data filtering and data denoising. Image feature extraction can use various computer vision technologies and image processing algorithms, such as Local Binary Patterns (LBP), Histogram of Oriented Gradients (HOG), Convolutional Neural Networks (CNNs), etc. These technologies can extract various features from images, such as texture features, shape features, color features, etc.

[0082] 106. If the UAV device is in a stall phase, the retraction speed of the mooring line retraction device is calculated based on the stall data and the environmental data collected by the UAV device, and the mooring line retraction device is controlled to retract the mooring line according to the retraction speed;

[0083] In one embodiment of the present invention, the calculating the retracting and releasing speed of the mooring line retracting and releasing device based on the stall data and the environmental data collected by the drone device, and controlling the mooring line retracting and releasing device to reel in the mooring line according to the retracting and releasing speed includes: calculating the unloading acceleration of the drone device based on the stall data and the environmental data collected by the drone device through a preset dynamic model; calculating the predicted descent speed of the drone device in the next time step based on the descent acceleration and the current speed of the drone device in the current time step; calculating the retracting and releasing speed of the mooring line retracting and releasing device according to the predicted descent speed of the drone device in the next time step and the relative position of the drone device and the ground device; and controlling the mooring line retracting and releasing device to reel in the mooring line according to the retracting and releasing speed.

[0084] Specifically, a preset dynamic model is established, which can calculate the dynamic behavior of the drone, especially the descent acceleration, based on the stall data and environmental data (such as wind speed, temperature, humidity, etc.), and obtain the following formula:

[0085] [ a_{\text{descent}} = f_{\text{dynamics}}(\text{stall data}, \text{environmental data}) ]

[0086] Where (a_{\text{down}}) represents the descent acceleration, and (f_{\text{dynamics}}) is a function based on the preset dynamics model. Then predict the descent speed of the next time step: [ v_{\text{down}, t+1} = v_{\text{down}, t} + a_{\text{down}} \times \Delta t ] Where (v_{\text{down}, t}) is the descent speed of the current time step, (v_{\text{down}, t+1}) is the predicted descent speed of the next time step, and (\Delta t) is the time step length. Based on the predicted descent speed of the drone and the relative position of the drone and the ground equipment, the retraction and release speed of the tethering line retraction device is calculated to adjust the length of the tethering line to ensure the safe recovery or stable flight of the drone. The calculation formula is as follows: [ v_{\text{Retract}} = g(v_{\text{Descent}, t+1}, \text{Relative Position}) ] where (v_{\text{Retract}}) is the retracting speed of the tethered line, and (g) is a function determined based on the predicted descent speed and the relative position of the UAV and the ground equipment. Finally, based on the calculated retracting speed of the tethered line, the tethered line retracting device is controlled to perform the retracting operation to adapt to the real-time flight status and environmental conditions of the UAV to ensure its safety.

[0087] Furthermore, after calculating the retraction and extension speed of the tethered line retraction and extension device according to the predicted descent speed of the UAV device in the next time step and the relative position of the UAV device and the ground device, it also includes: obtaining the predicted descent speed of the previous time step and the actual descent speed of the current time step, and calculating the predicted deviation value between the predicted descent speed of the previous time step and the actual descent speed of the current time step; obtaining the stall time of the UAV device in a stall state, and calculating the predicted correction value of the UAV device in the current time step according to the stall time and the predicted deviation value; adjusting the retraction and extension speed of the tethered line retraction and extension device according to the predicted correction value.

[0088] Specifically, when adjusting the retraction and release speed of the tethered line retraction and release device, first collect data including the current altitude, speed, acceleration of the drone, and external environmental factors such as wind speed and wind direction, and use dynamic and aerodynamic models, combined with input data, to predict the descent trajectory and speed of the drone in the future. Then, based on the predicted descent trajectory, calculate the length of the tethered line that needs to be adjusted in the future to maintain a stable connection between the drone and the ground station. Monitor the descent status of the drone in real time, including altitude, speed, acceleration, etc., and compare them with the predicted values. Calculate the error between the predicted value and the actual value. Dynamically adjust the retraction and release speed of the tethered line according to the size and direction of the error. This can be achieved through a PID (proportional-integral-differential) controller, which adjusts the output (i.e., the retraction and release speed of the tethered line) according to the proportion, integral, and differential of the error to reduce the error.

[0089] 107. Control the ejection device on the drone equipment to eject the airbag and put the airbag into an inflated state, thereby achieving safe recovery of the drone equipment.

[0090] In one embodiment of the present invention,

[0091] Furthermore, the passive anti-explosion mode also includes:

[0092] Perform visual recognition on the real-time image collected by the drone device to determine the target object in the real-time image and the object classification of the target object; determine whether to track the position of the target object based on the object classification; if so, use the real-time image at the same time as the position tracking determination as the initial frame image, and generate the real-time movement trajectory of the target object in real time based on the initial frame image and the real-time images subsequent to the initial frame image; generate the predicted trajectory of the target object based on the object classification and the real-time movement trajectory, and determine whether the predicted trajectory coincides with the position of the drone device; if coincident, control the tethering line retracting and releasing device to retract the tethering line to achieve safe recovery of the drone device.

[0093] Specifically, first, the drone equipment will collect real-time images of the surrounding environment through the camera or sensor it carries. Computer vision technology is used to detect targets on the collected real-time images and identify the target objects in the images. This involves using target detection algorithms (such as YOLO, Faster R-CNN, etc.) to calibrate the position of the target object in the image. Once the position of the target object is identified, the target object can be classified using image recognition technology. This can be achieved through deep learning models such as convolutional neural networks (CNNs) to determine which category the target object belongs to (such as people, vehicles, buildings, etc.). Based on the results of object classification, it is determined whether the target object needs to be tracked. If so, the real-time image at the current moment is recorded as the initial frame image. Based on the initial frame image and subsequent real-time images, motion tracking algorithms (such as optical flow method, Kalman filter, etc.) can be used to generate the movement trajectory of the target object in real time. Finally, combined with object classification and real-time movement trajectory, prediction algorithms (such as Kalman filter, particle filter, etc.) can be used to generate the predicted trajectory of the target object.

[0094] Furthermore, the visual recognition of the real-time image collected by the UAV device to determine the target object in the real-time image and the object classification of the target object includes: performing image preprocessing on the real-time image to obtain a preprocessed image, and inputting the preprocessed image into the image detection model, wherein the image detection model includes an input layer, a feature extraction layer, a feature pyramid network, a feature fusion layer, a multi-task detection layer and an output layer; mapping the preprocessed image to a high-dimensional space through the input layer to obtain initial feature information in the high-dimensional space; performing feature extraction on the initial feature information through the feature extraction layer to obtain convolution features of different levels; generating initial fusion features corresponding to convolution features of different levels through the feature pyramid network, and performing feature fusion on the initial fusion features of different levels through the feature fusion layer to obtain secondary fusion features; determining the target object in the real-time image and the object classification of the target object according to the secondary fusion features through the multi-task detection layer and the output layer.

[0095] Specifically, the model structure above uses a multi-task detection network, such as Faster R-CNN, SSD (Single Shot MultiBox Detector) or YOLO (You Only Look Once). First, the preprocessed image data is received and mapped to a high-dimensional space in preparation for subsequent processing. The initial feature information is convolved to extract convolutional features at different levels. These features will capture various information in the image, such as edges, textures, etc. The feature pyramid network generates initial fusion features corresponding to convolutional features at different levels. Generally, the feature pyramid network aims to generate multi-scale feature maps to detect target objects at different scales. The feature fusion layer fuses the initial fusion features at different levels to obtain a richer and more representative feature representation. This step can usually fuse feature information from different levels to enhance the model's ability to represent the target. The multi-task detection layer is usually used to simultaneously predict the location and category of the target, as well as other tasks related to target detection. This involves regression of bounding box coordinates and classification of target categories. Finally, the output layer determines the target object and the classification of the object in the real-time image based on the secondary fusion features. The output layer generates the target’s location information and category probability for the final target detection results.

[0096] In this embodiment, the drone device is controlled to enter the active anti-explosion mode and the passive anti-explosion mode; the active anti-explosion mode includes: when the drone device reaches the target position, the drone device is controlled to enter the locking mode, and the drone device stops receiving external signals in the locking mode; the passive anti-explosion mode includes: based on the posture information of the drone device and the real-time image taken, the stall data of the drone device is calculated, and then the retraction speed of the tethered line retraction device is calculated, and the tethered line retraction device is controlled to retract the tethered line according to the retraction speed; the ejection device on the drone device is controlled to eject the airbag and inflate it, so as to achieve the safe recovery of the drone device. This method improves the safety and reliability of the tethered drone through the active and passive anti-explosion modes and uses the tethered line as a recovery tool for the drone in an emergency.

[0097] The above describes the anti-explosion method of a tethered drone in an embodiment of the present invention. The anti-explosion device of a tethered drone is applied to an anti-explosion system of a tethered drone. The anti-explosion system of a tethered drone includes a drone device, a tethering line, and a ground device that is communicatively connected to the drone device through the tethering line. The ground device includes a tethering line retracting and releasing device. The drone device is equipped with a camera device.

[0098] The following is a description of the anti-explosion device for tethered drones in an embodiment of the present invention. Figure 2In one embodiment of the present invention, an anti-explosion device for tethering a UAV includes:

[0099] The control module 201 is used to control the UAV device to enter the active anti-bombing mode and the passive anti-bombing mode simultaneously when the UAV device is performing aerial operations;

[0100] The active anti-explosion module 202 is used to perform the following operations when the UAV device is in the active anti-explosion mode:

[0101] When the drone device reaches the target position based on the control signal sent by the ground device, the drone device is controlled to enter a locking mode, wherein the drone device stops receiving external signals in the locking mode; the input unlocking password input by the ground device is sent to the drone device through the tethered line, and the drone device determines whether the input unlocking password is the same as the preset set unlocking password; if so, the drone device is controlled to exit the locking mode and receive external signals again to perform aerial operations;

[0102] The passive anti-explosion module 203 is used to perform the following operations when the UAV device is in the active anti-explosion mode:

[0103] The position information of the UAV device and the real-time image captured by the camera device are acquired in real time, the stall data of the UAV device is calculated based on the position information and the real-time image, and whether the UAV device is in a stall state is determined based on the stall data; if the UAV device is in a stall stage, the retraction and extension speed of the tethering line retraction and extension device is calculated based on the stall data and the environmental data collected by the UAV device, and the tethering line retraction and extension device is controlled to retract the tethering line according to the retraction and extension speed; the ejection device on the UAV device is controlled to eject the airbag and put the airbag into an inflated state, so as to achieve safe recovery of the UAV device.

[0104] In an embodiment of the present invention, the anti-explosion device of the tethered drone runs the anti-explosion method of the tethered drone, and the anti-explosion device of the tethered drone enters the active anti-explosion mode and the passive anti-explosion mode by controlling the drone device; the active anti-explosion mode includes: when the drone device reaches the target position, the drone device is controlled to enter the locking mode, and the drone device stops receiving external signals in the locking mode; the passive anti-explosion mode includes: based on the posture information of the drone device and the real-time image taken, the stall data of the drone device is calculated, and then the retraction speed of the tethered line retraction device is calculated, and the tethered line retraction device is controlled to retract the tethered line according to the retraction speed; the ejection device on the drone device is controlled to eject the airbag and inflate it, so as to achieve the safe recovery of the drone device. This method improves the safety and reliability of the tethered drone through the active and passive anti-explosion modes, and uses the tethered line as a recovery tool for the drone in an emergency.

[0105] above Figure 2 The anti-explosion device of the tethered drone in the embodiment of the present invention is described in detail from the perspective of modular functional entities. The anti-explosion device of the tethered drone in the embodiment of the present invention is described in detail from the perspective of hardware processing.

[0106] Figure 3 1 is a schematic diagram of the structure of a bomb-proof device for a tethered drone provided by an embodiment of the present invention. The bomb-proof device 300 for a tethered drone may have relatively large differences due to different configurations or performances, and may include one or more processors (central processing units, CPU) 310 (for example, one or more processors) and a memory 320, and one or more storage media 330 (for example, one or more mass storage device terminals) storing application programs 333 or data 332. Among them, the memory 320 and the storage medium 330 may be temporary storage or permanent storage. The program stored in the storage medium 330 may include one or more modules (not shown in the figure), and each module may include a series of instruction operations in the bomb-proof device 300 for a tethered drone. Furthermore, the processor 310 may be configured to communicate with the storage medium 330, and execute a series of instruction operations in the storage medium 330 on the bomb-proof device 300 for a tethered drone to implement the steps of the bomb-proof method for a tethered drone.

[0107] The anti-explosion device 300 for the tethered drone may also include one or more power supplies 340, one or more wired or wireless network interfaces 350, one or more input and output interfaces 360, and / or one or more operating systems 331, such as Windows Serve, Mac OS X, Unix, Linux, FreeBSD, etc. It will be appreciated by those skilled in the art that Figure 3 The structure of the anti-explosion device for the tethered drone shown does not constitute a limitation on the anti-explosion device for the tethered drone provided by the present invention, and may include more or fewer components than shown in the figure, or a combination of certain components, or a different arrangement of components.

[0108] The present invention also provides a computer-readable storage medium, which may be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium. Instructions are stored in the computer-readable storage medium. When the instructions are executed on a computer, the computer executes the steps of the anti-explosion method for a tethered drone.

[0109] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the above-described system, device, or unit can refer to the corresponding process in the aforementioned method embodiment and will not be repeated here.

[0110] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art or the whole or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk and other media that can store program code.

[0111] As described above, the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features thereof may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for preventing a tethered drone from bombing. It is characterized in that The anti-explosion method of a tethered UAV is applied to an anti-explosion system of a tethered UAV, the anti-explosion system of a tethered UAV comprising a UAV device, a tethering line, and a ground device connected to the UAV device through the tethering line, the ground device comprising a tethering line retracting and releasing device, and the UAV device is equipped with a camera device; The anti-explosion method of the tethered UAV includes: When the UAV device is performing aerial operations, the UAV device is controlled to simultaneously enter an active anti-bombing mode and a passive anti-bombing mode; The active anti-explosion mode includes: When the UAV device reaches the target position based on the control signal sent by the ground device, the UAV device is controlled to enter a locking mode, wherein the UAV device stops receiving external signals in the locking mode; The input unlock password input by the ground device is sent to the drone device through the tethered line, and the drone device determines whether the input unlock password is the same as a preset set unlock password; If so, the UAV device is controlled to exit the locking mode and receive external signals again to perform aerial operations; The passive anti-bombing mode includes: Acquire the posture information of the UAV device and the real-time image captured by the camera device in real time, calculate the stall data of the UAV device based on the posture information and the real-time image, and determine whether the UAV device is in a stall state based on the stall data; If the UAV device is in a stall phase, the retraction speed of the mooring line retraction device is calculated based on the stall data and the environmental data collected by the UAV device, and the mooring line retraction device is controlled to retract the mooring line according to the retraction speed; The ejection device on the drone equipment is controlled to eject the airbag and put the airbag into an inflated state, thereby achieving safe recovery of the drone equipment.

2. The method for preventing a tethered drone from exploding according to claim 1, It is characterized in that The step of acquiring the posture information of the drone device and the real-time image captured by the camera device in real time, calculating the stall data of the drone device based on the posture information and the real-time image, and determining whether the drone device is in a stall state based on the stall data includes: Acquire the posture information of the drone device and the real-time image captured by the camera device in real time, and generate corresponding posture time series data and image time series data according to the time axis of the posture information and the real-time image; Inputting the posture time series data and the image time series data into a preset stall detection model, wherein the stall detection model includes an input layer, an attention mechanism layer, a feature fusion layer, a multi-task detection layer and an output layer; Preprocessing the posture time series data and the image time series data through the input layer, and performing data feature extraction and image feature extraction on the preprocessed posture time series data and image time series data, respectively, to obtain posture time series features and image time series features, respectively; Calculate the first weight vector of the posture temporal feature and the second weight vector of the image temporal feature through the attention mechanism layer; Performing feature fusion on the posture time series feature and the image time series feature according to the first weight vector and the second weight vector through the feature fusion layer to obtain a fused feature vector; Performing multi-task detection according to the fused feature vector through a random forest network in the multi-task detection layer to obtain multiple types of stall data, and outputting the multiple types of stall data through the output layer; Determine whether the drone device is in a stall state according to the multiple types of stall data.

3. The anti-explosion method of a tethered drone according to claim 2, It is characterized in that The step of preprocessing the posture time series data and the image time series data through the input layer, and performing data feature extraction and image feature extraction on the preprocessed posture time series data and image time series data, respectively, to obtain posture time series features and image time series features, respectively, includes: Performing data filtering and data denoising on the posture time series data and the image time series data; Performing standard deviation threshold calculation on the posture time series data after data filtering and data denoising to obtain the standard deviation threshold of the posture time series data, and removing outliers from the posture time series data after data filtering and data denoising by using the standard deviation threshold; Data feature extraction and image feature extraction are performed on the posture time series data after outliers are removed, and the image time series data after data filtering and data denoising, to obtain posture time series features and image time series features, respectively.

4. The method for preventing a tethered drone from exploding according to claim 1, It is characterized in that The calculating the retracting speed of the mooring line retracting device based on the stall data and the environmental data collected by the drone device, and controlling the mooring line retracting device to retract the mooring line according to the retracting speed includes: Calculating the descent acceleration of the UAV device based on the stall data and the environmental data collected by the UAV device through a preset dynamic model; Calculate a predicted descent speed of the UAV device at a next time step based on the descent acceleration and the current speed of the UAV device at the current time step; Calculate the retraction and extension speed of the tethered line retraction and extension device according to the predicted descent speed of the UAV device at the next time step and the relative position of the UAV device and the ground device; The mooring line retracting and releasing device is controlled to retract the mooring line according to the retracting and releasing speed.

5. The method for preventing a tethered drone from exploding according to claim 4, It is characterized in that After calculating the retracting and releasing speed of the tethered line retracting and releasing device according to the predicted descent speed of the UAV device at the next time step and the relative position of the UAV device and the ground device, the method further includes: Obtaining the predicted descent speed of the previous time step and the actual descent speed of the current time step, and calculating the predicted deviation value between the predicted descent speed of the previous time step and the actual descent speed of the current time step; Obtaining the stall time of the UAV device in a stall state, and calculating the prediction correction value of the UAV device at the current time step according to the stall time and the prediction deviation value; The retraction and extension speed of the mooring line retraction and extension device is adjusted according to the predicted correction value.

6. The method for preventing a tethered drone from exploding according to claim 1, It is characterized in that The passive anti-bombing mode also includes: Performing visual recognition on the real-time image collected by the drone device to determine the target object in the real-time image and the object classification of the target object; Determining whether to perform position tracking on the target object according to the object classification; If yes, the real-time image at the same time as the position tracking is performed is used as the initial frame image, and the real-time moving trajectory of the target object is generated in real time according to the initial frame image and the real-time images subsequent to the initial frame image; generating a predicted trajectory of the target object based on the object classification and the real-time moving trajectory, and determining whether the predicted trajectory coincides with the position of the UAV device; If they overlap, the mooring line retracting and releasing device is controlled to reel in the mooring line to achieve safe recovery of the UAV equipment.

7. The method for preventing a tethered drone from exploding according to claim 6, It is characterized in that The visual recognition of the real-time image collected by the drone device to determine the target object in the real-time image and the object classification of the target object includes: Performing image preprocessing on the real-time image to obtain a preprocessed image, and inputting the preprocessed image into an image detection model, wherein the image detection model includes an input layer, a feature extraction layer, a feature pyramid network, a feature fusion layer, a multi-task detection layer and an output layer; Mapping the preprocessed image to a high-dimensional space through the input layer to obtain initial feature information in the high-dimensional space; Extracting features from the initial feature information through the feature extraction layer to obtain convolution features at different levels; Generating initial fusion features corresponding to convolutional features at different levels through the feature pyramid network, and fusing the initial fusion features at different levels through the feature fusion layer to obtain secondary fusion features; The target object in the real-time image and the object classification of the target object are determined according to the secondary fusion features through the multi-task detection layer and the output layer. 8.An anti-explosion device for a tethered drone, It is characterized in that The anti-explosion device for a tethered UAV is applied to an anti-explosion system for a tethered UAV, the anti-explosion system for a tethered UAV comprising a UAV device, a tethering line, and ground equipment that is communicatively connected to the UAV device via the tethering line, the ground equipment comprising a tethering line retracting and releasing device, and the UAV device is equipped with a camera device; The anti-explosion device of the tethered UAV comprises: A control module, used for controlling the UAV device to enter an active anti-bombing mode and a passive anti-bombing mode simultaneously when the UAV device is performing aerial operations; The active anti-explosion module is used to perform the following operations when the UAV device is in the active anti-explosion mode: When the drone device reaches the target position based on the control signal sent by the ground device, the drone device is controlled to enter a locking mode, wherein the drone device stops receiving external signals in the locking mode; the input unlocking password input by the ground device is sent to the drone device through the tethered line, and the drone device determines whether the input unlocking password is the same as the preset set unlocking password; if so, the drone device is controlled to exit the locking mode and receive external signals again to perform aerial operations; The passive anti-explosion module is used to perform the following operations when the UAV device is in the active anti-explosion mode: The position information of the UAV device and the real-time image captured by the camera device are acquired in real time, the stall data of the UAV device is calculated based on the position information and the real-time image, and whether the UAV device is in a stall state is determined based on the stall data; if the UAV device is in a stall stage, the retraction and extension speed of the tethering line retraction and extension device is calculated based on the stall data and the environmental data collected by the UAV device, and the tethering line retraction and extension device is controlled to retract the tethering line according to the retraction and extension speed; the ejection device on the UAV device is controlled to eject the airbag and put the airbag into an inflated state, so as to achieve safe recovery of the UAV device. 9.An anti-explosion device for tethered drones, It is characterized in that The anti-explosion device for the tethered drone includes: a memory and at least one processor, wherein the memory stores instructions; The at least one processor calls the instructions in the memory to enable the anti-explosion device of the tethered drone to perform the steps of the anti-explosion method of the tethered drone as described in any one of claims 1-7.

10. A computer-readable storage medium having instructions stored thereon, It is characterized in that When the instructions are executed by the processor, the steps of the anti-explosion method for a tethered drone as described in any one of claims 1 to 7 are implemented.

Citation Information

Patent Citations

  • Unmanned plane supervision cloud platform based on Beidou geofence differential positioning module

    CN107132852A

  • Unmanned aerial vehicle and grounding device

    CN113697105A

  • Dual-power control system applied to mooring aircraft

    CN116374247A

  • Tethered unmanned aerial vehicle fire fighting system

    US20170043872A1

  • Methods and Systems for User Interaction and Feedback Via Control of Tether

    US20180072421A1