Beacon data processing method and device, electronic equipment and storage medium

Abstract

The embodiment of the invention provides a beacon data processing method and device, electronic equipment and a storage medium. The method comprises the steps that vibration data, collected by a vibration sensor, of the unmanned aerial vehicle in the operation process are acquired; the vibration sensor is arranged on the unmanned aerial vehicle; processing the vibration data to obtain a multi-dimensional time domain feature corresponding to the vibration data; calling a pre-trained beacon abnormal vibration detection model to process the multi-dimensional time domain features, and detecting whether an unmanned aerial vehicle beacon arranged on the unmanned aerial vehicle is in an abnormal vibration mode; and when it is determined that the unmanned aerial vehicle beacon is in a physical removal state according to the detected condition that the unmanned aerial vehicle beacon is in the abnormal vibration mode, data in the unmanned aerial vehicle beacon is removed. Data leakage in the beacon of the unmanned aerial vehicle can be avoided, and the safety and reliability of the unmanned aerial vehicle are improved.

Description

Technical Field

[0001] The present application relates to the field of drone technology, and in particular to a beacon data processing method, device, electronic device and storage medium. Background Art

[0002] With the widespread application of drone technology, the safety and reliability of drones have become important issues. As a key device for communication between drones and ground control stations, the safety of drone beacons is directly related to the operational safety and data security of drones.

[0003] Currently, drone beacons are mainly attached by a small device that is attached to the drone to achieve a one-to-one physical binding. The small device has a built-in anti-disassembly mechanism, such as a spring sheet. When an attacker attempts to dismantle the drone beacon, the spring sheet pops out, triggering the anti-disassembly mechanism and clearing the data in the drone beacon. However, this anti-disassembly method that relies on physical means such as spring sheets has the following risks: 1. Insufficient security: If the attacker knows the specific location of the spring piece in advance, he or she may deliberately avoid it during the process of dismantling the drone beacon, so that the anti-dismantling mechanism cannot be triggered, resulting in data leakage in the drone beacon.

[0004] 2. Lack of dynamism: In a complex and changing environment, this physical anti-dismantling mechanism cannot adjust the anti-dismantling strategy according to the real-time operating status of the drone and changes in the environment. It lacks a flexible and effective response mechanism and poses a greater security risk. Summary of the invention

[0005] The technical problem to be solved by the embodiments of the present application is to provide a beacon data processing method, device, electronic device and storage medium to avoid data leakage in drone beacons and improve the safety and reliability of drones.

[0006] In a first aspect, an embodiment of the present application provides a beacon data processing method, the method comprising: Acquiring vibration data of the UAV during operation collected by a vibration sensor; the vibration sensor is arranged on the UAV; Processing the vibration data to obtain multi-dimensional time domain features corresponding to the vibration data; Calling a pre-trained beacon abnormal vibration detection model to process the multi-dimensional time domain features to detect whether the drone beacon set on the drone is in an abnormal vibration mode; When it is determined that the drone beacon is in a physically dismantled state based on the condition that the drone beacon is in an abnormal vibration mode, the data in the drone beacon is cleared.

[0007] In a second aspect, an embodiment of the present application provides a beacon data processing device, the device comprising: A vibration data acquisition module, used to acquire vibration data of the UAV during operation collected by a vibration sensor; the vibration sensor is arranged on the UAV; A time domain feature acquisition module, used to process the vibration data to obtain multi-dimensional time domain features corresponding to the vibration data; An abnormal vibration detection module, used to call a pre-trained beacon abnormal vibration detection model to process the multi-dimensional time domain features to detect whether the drone beacon set on the drone is in an abnormal vibration mode; The beacon data clearing module is used to clear the data in the drone beacon when it is determined that the drone beacon is in a physically dismantled state based on the condition that the drone beacon is in an abnormal vibration mode.

[0008] In a third aspect, an embodiment of the present application provides an electronic device, including: A processor, a memory, and a computer program stored in the memory and executable on the processor, wherein the processor implements any of the above-mentioned beacon data processing methods when executing the program.

[0009] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium. When the instructions in the storage medium are executed by a processor of an electronic device, the electronic device can execute any of the beacon data processing methods described above.

[0010] Compared with the prior art, the embodiments of the present application have the following advantages: In an embodiment of the present application, the vibration data of the drone during operation collected by the vibration sensor is obtained, and the vibration sensor is set on the drone. The vibration data is processed to obtain the multi-dimensional time domain features corresponding to the vibration data. The pre-trained beacon abnormal vibration detection model is called to process the multi-dimensional time domain features to detect whether the drone beacon set on the drone is in an abnormal vibration mode. In the case where it is determined that the drone beacon is in a physically dismantled state based on the condition that the detected drone beacon is in an abnormal vibration mode, the data in the drone beacon is cleared. The embodiment of the present application dynamically detects the abnormal vibration of the drone beacon through the beacon abnormal vibration detection model, realizes a dynamic anti-dismantling mechanism, effectively prevents data leakage, and significantly improves the security and adaptability of the drone beacon anti-dismantling mechanism, as well as the security of drone data.

[0011] It should be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] Figure 1 A flowchart of a beacon data processing method provided in an embodiment of the present application; Figure 2 A flowchart of a method for determining an abnormal vibration mode provided in an embodiment of the present application; Figure 3 A flowchart of a method for detecting abnormal vibration patterns provided in an embodiment of the present application; Figure 4 A flowchart of a method for determining a physical dismantling state provided in an embodiment of the present application; Figure 5 A schematic diagram of a drone beacon vibration monitoring and abnormal alarm process provided in an embodiment of the present application; Figure 6 A schematic diagram of a system architecture provided for an embodiment of the present application; Figure 7 A schematic diagram of the structure of a beacon data processing device provided in an embodiment of the present application; Figure 8 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0013] In order to make the above-mentioned objects, features and advantages of the present application more obvious and understandable, further detailed description is given below in conjunction with the accompanying drawings and specific implementation methods.

[0014] The terms used in the embodiments of the present application are only for the purpose of describing specific embodiments, and are not intended to limit the present application. The singular forms "a", "said" and "the" used in the embodiments of the present application are also intended to include plural forms, unless the context clearly indicates other meanings.

[0015] Reference Figure 1 , shows a flow chart of the steps of a beacon data processing method provided by an embodiment of the present application. Figure 1 As shown, the beacon data processing method may include: step 101, step 102, step 103 and step 104.

[0016] Step 101: Obtain vibration data of a UAV during operation collected by a vibration sensor; the vibration sensor is disposed on the UAV.

[0017] The embodiments of the present application can be applied to scenarios where an AI model is combined to determine whether a drone beacon is in a physically dismantled state.

[0018] A drone beacon is a small device mounted on a drone that emits specific signals to help the drone locate, navigate and communicate.

[0019] A vibration sensor is a device used to detect and measure the vibration or movement of an object. It can monitor the vibration state of the equipment in real time and convert the vibration signal into an electrical signal for analysis and processing. Vibration sensors are widely used in industry, machinery, aerospace, automobiles and other fields, mainly for equipment monitoring, fault diagnosis and status evaluation.

[0020] In this embodiment, a vibration sensor is provided on the drone, and the vibration sensor can be used to monitor the vibration data of the drone during operation. Specifically, a high-precision vibration sensor (such as an MSMS (Multi-Stage Mass Spectrometry) sensor, etc.) can be installed at key parts of the drone (such as fuselage connections, etc.), and the vibration data of the drone during flight can be obtained through the sensor. The accuracy of the sensor meets the requirement of being able to distinguish between normal operating vibration and abnormal vibration.

[0021] During the operation of the drone, vibration data can be collected by a vibration sensor installed on the drone. In a specific implementation, a suitable vibration data collection frequency can be set according to factors such as the flight speed of the drone and mission requirements to capture representative vibration information. After the vibration sensor collects the vibration data, the vibration sensor can send the vibration data to the drone beacon.

[0022] After the vibration data of the UAV during operation collected by the vibration sensor is acquired, step 102 is executed.

[0023] Step 102: Process the vibration data to obtain multi-dimensional time domain features corresponding to the vibration data.

[0024] After the drone beacon obtains the vibration data of the drone during operation collected by the vibration sensor, the vibration data can be preprocessed according to the preprocessing method to obtain preprocessed vibration data, and the preprocessed vibration data can be processed to obtain multi-dimensional time domain features. The specific preprocessing method may include: 1. Filtering and noise reduction: Use filtering technology to remove high-frequency noise or low-frequency interference in the original vibration data, and only retain useful vibration information; use wavelet transform technology to perform noise reduction on the original vibration data, effectively removing noise while retaining the detailed characteristics of the signal.

[0025] 2. De-meaning: Subtract the mean of the collected data to make the signal centered on 0 and better extract other features.

[0026] 3. Normalization: Normalize the collected vibration data and map the data to a specific interval of [0,1] so that the vibration data has a unified scale, which is convenient for subsequent feature extraction and model training.

[0027] Of course, in practical applications, other preprocessing methods may be used to preprocess the vibration data, and this embodiment does not limit the preprocessing method for the vibration data.

[0028] After the vibration data is preprocessed, the preprocessed vibration data may be processed to obtain multi-dimensional time domain features corresponding to the vibration data. In this example, the multi-dimensional time domain features may include: mean, variance, root mean square value, peak value, peak-to-peak value, and skewness of the vibration data in the time domain.

[0029] In the field of signal processing, the definition of time domain is a direct description of the change of signal amplitude over time, that is, a sequence of instantaneous values ​​of the signal at different time points. For the vibration signal of drone beacon, time domain analysis focuses on the original fluctuation form of the vibration amplitude on the continuous time axis.

[0030] After obtaining the preprocessed vibration data, the drone beacon can extract time domain features from the preprocessed data, that is, calculate the mean, variance, root mean square value, peak value, peak-to-peak value, and skewness of the vibration data, and reflect the change in the overall vibration level of the drone beacon through the change of features. The following is the feature calculation content: 1. Calculate the average value of vibration data in the time domain to obtain the mean value of vibration data in the time domain. This mean value can reflect the overall level of vibration data. The calculation formula is as follows:

[0031] In the above formula, is the signal length of the vibration data, is the amplitude of the signal at the i-th time point, is the mean value of vibration data in the time domain, is the number of vibration data, that is, the total number of vibration data sample points involved in calculating the mean.

[0032] 2. Calculate the variance of the vibration data in the time domain, which can reflect the degree of fluctuation of the vibration data. The calculation formula is as follows:

[0033] In the above formula, is the variance of vibration data in the time domain, is the signal length of the vibration data, is the amplitude of the signal at the i-th time point, is the mean value of the vibration data in the time domain.

[0034] 3. Calculate the root mean square value of the vibration data in the time domain, which can reflect the energy of the vibration data. The calculation formula is as follows:

[0035] In the above formula, is the root mean square value of the vibration data in the time domain, is the signal length of the vibration data, is the amplitude of the signal at the i-th time point.

[0036] 4. Calculate the maximum value of the vibration data in the time domain, and use the maximum value as the peak value of the vibration data in the time domain, reflecting the maximum amplitude of the vibration data. In this example, the maximum value of the vibration data in the time domain can be obtained, that is, .

[0037] 5. Calculate the difference between the maximum and minimum values ​​of the vibration data in the time domain, and use the difference as the peak-to-peak value of the vibration data in the time domain. The peak-to-peak value can reflect the fluctuation range of the signal. That is, the peak-to-peak value = , It is the minimum value of the vibration data in the time domain, reflecting the minimum amplitude of the vibration data.

[0038] 6. Calculate the skewness of the vibration data in the time domain, which can reflect the symmetry of the vibration data. The calculation formula is as follows:

[0039] In the above formula, is the skewness of the vibration data in the time domain, is the signal length of the vibration data, is the amplitude of the signal at the i-th time point, is the mean value of vibration data in the time domain, is the overall standard deviation of the vibration data, (obtained from the square root of the variance).

[0040] After the vibration data is processed to obtain multi-dimensional time domain features corresponding to the vibration data, step 103 is executed.

[0041] Step 103: calling a pre-trained beacon abnormal vibration detection model to process the multi-dimensional time domain features to detect whether the drone beacon set on the drone is in an abnormal vibration mode.

[0042] The beacon abnormal vibration detection model refers to a model used to detect whether the drone beacon is in an abnormal vibration mode. In this example, the beacon abnormal vibration detection model can be a random forest model.

[0043] In this embodiment, the eigenvalue of the extracted time-domain feature can be sent to the AI inference platform by the UAV beacon. The AI inference platform adopts the random forest algorithm to train the extracted features and constructs an algorithm model that can distinguish normal and abnormal vibration modes. The process of the AI inference platform for model training is as follows: 1. Dataset division: Divide the prepared feature dataset into a training set and a test set. The training set is used to train the model, and the test set is used to evaluate the final performance of the model. The division ratio is 70% for the training set and 30% for the test set.

[0044] 2. Model initialization: Set the forest parameters: 1) Determine the number of decision trees in the forest, such as 100 trees, etc.; 2) The maximum depth (max_depth) of each tree: Limit the growth depth of the decision tree; 3) The minimum number of samples required for node splitting (min_samples_split): The sample number threshold that determines whether the node continues to split.

[0045] 3. Model training: 1) Sample extraction: Randomly draw B sub-datasets D1, D2,..., DB with the same size as the original dataset from the original training set D with replacement to obtain the training samples for each decision tree.

[0046] 2) Feature selection: When splitting each node, randomly select m features (m << M, where M is the total number of features), and only query the optimal splitting point among these features.

[0047] 3) Construct decision trees: Use the sampled samples and selected features to train each decision tree until the set maximum depth or the number of node samples is reached.

[0048] 4) Combine decision trees: Combine all the trained decision trees into a random forest model.

[0049] 4. Model evaluation: Use the test set to evaluate the performance of the model, and calculate indicators such as accuracy, recall rate, and F1 value.

[0050] 1) The calculation formula for accuracy is as follows:

[0051] 2) The calculation formula for recall rate is as follows:

[0052] 3) The calculation formula for F1 value is as follows:

[0053]

[0054] In the above formula, TP (True Positive) is the number of samples correctly predicted as positive, TN (True Negative) is the number of samples correctly predicted as negative, FP (False Positive) is the number of negative samples wrongly predicted as positive, and FN (False Negative) is the number of positive samples wrongly predicted as negative. Represents the precision rate (i.e., the proportion of samples that are actually positive among samples predicted to be positive).

[0055] After training the random forest model for detecting abnormal vibration of the drone beacon, the model can be deployed on the drone beacon side or on the ground station side. Specifically, the deployment location of the model can be determined according to business needs, which is not limited in this embodiment.

[0056] After obtaining the multi-dimensional time domain features corresponding to the vibration data, the beacon abnormal vibration detection model can be called to process the multi-dimensional time domain features to detect whether the drone beacon set on the drone is in an abnormal vibration mode. Specifically, the abnormal vibration detection model can classify the multi-dimensional time domain features according to the abnormal detection rules to obtain the classified time domain features, and determine whether the drone beacon is in an abnormal vibration mode according to the classified time domain features and the normal feature values ​​corresponding to the classified time domain features according to the abnormal judgment rules. This implementation process can be combined with Figure 2 This is described in detail as follows.

[0057] Reference Figure 2 , shows a flowchart of the steps of a method for determining an abnormal vibration mode provided by an embodiment of the present application. Figure 2 As shown, the abnormal vibration mode determination method may include: step 201 and step 202.

[0058] Step 201: calling the beacon abnormal vibration detection model to classify the multi-dimensional time domain features according to a preset abnormality detection rule to obtain classified time domain features.

[0059] In this embodiment, the anomaly detection rule is used to classify the multi-dimensional time domain features. It divides each time domain feature into different categories based on the statistical analysis and prior knowledge of the time domain features under normal and abnormal vibration modes.

[0060] After obtaining the multi-dimensional time domain features, the beacon abnormal vibration detection model can be called to classify the multi-dimensional time domain features according to its internal rules to obtain classified time domain features. Specifically, the abnormal detection rule can be predetermined by the beacon abnormal vibration detection model, which corresponds to the abnormal determination rule defined in step 202, that is, the classification rule defines which time domain features are grouped together. The classified time domain feature includes a group of time domain features, which can be one or more.

[0061] After the beacon abnormal vibration detection model is called to classify the multi-dimensional time domain features according to the preset abnormality detection rules to obtain the classified time domain features, step 202 is executed.

[0062] Step 202: According to the abnormality determination rule, based on the classification time domain features and the normal feature values ​​corresponding to the classification time domain features, determine whether the drone beacon is in an abnormal vibration mode.

[0063] The abnormality judgment rule is a rule for determining whether the drone beacon is in an abnormal vibration mode based on the classified time domain features and the corresponding normal feature values. The abnormality judgment rule defines the specific way of judging abnormal vibration based on different combinations of time domain features.

[0064] After calling the beacon abnormal vibration detection model to classify the multi-dimensional time domain features according to the pre-set anomaly detection rules to obtain the classified time domain features, it is possible to determine whether the drone beacon is in an abnormal vibration mode according to the classified time domain features and the normal feature values ​​corresponding to the classified time domain features according to the abnormal judgment rules.

[0065] In this example, the normal eigenvalue may be the maximum value when the drone beacon is in a non-physical removal state (i.e., normal vibration mode), which includes normal eigenvalues ​​corresponding to the mean, variance, root mean square value, peak value, peak-to-peak value, and skewness, respectively.

[0066] Specifically, when the abnormal judgment rule is a joint feature judgment rule, it can be determined whether the drone beacon is in an abnormal vibration mode based on the size relationship between the preset characteristic value of the classification time domain feature and the set multiple of the corresponding normal characteristic value. The classification time domain feature includes: at least two of: mean, variance, root mean square value, peak value, peak-to-peak value and skewness.

[0067] Among them, the joint feature judgment rule refers to a rule that combines at least two time domain features among the mean, variance, root mean square value, peak value, peak-to-peak value and skewness to determine whether the drone beacon is in an abnormal vibration mode.

[0068] The setting multiple may be a pre-set multiple corresponding to a normal feature value for abnormal vibration determination. In this example, different setting multiples may be set for different feature combinations. For example, when the joint feature is the six time domain features mentioned above, the setting multiple may be 1. When the joint feature is any two of the six time domain features mentioned above, the setting multiple may be 2.5, etc.

[0069] In the case where the abnormality determination rule is a joint feature determination rule, it is possible to determine whether the drone beacon is in an abnormal vibration mode based on the size relationship between the preset eigenvalue of the classification time domain feature and the set multiple of the corresponding normal eigenvalue. For example, when the mean and variance increase at the same time and are both greater than 2.5 times the corresponding normal eigenvalue, it may indicate that the drone beacon is subjected to an unstable external force, or that the internal vibration system fails, resulting in both a deviation in the average level of vibration and a deterioration in stability, and the drone beacon may be in an abnormal vibration mode, etc. Alternatively, when the peak-to-peak value, skewness, and root mean square value increase and are all greater than 2 times the corresponding normal eigenvalue, it may mean that the vibration amplitude range of the beacon has expanded, the distribution of the vibration signal has changed asymmetrically, the vibration energy has changed, and the drone beacon may be in an abnormal vibration mode, etc.

[0070] It can be understood that the above examples are merely examples listed for a better understanding of the technical solutions of the embodiments of the present application, and are not intended to be the sole limitation to the embodiments.

[0071] In practical applications, other methods can also be used to determine the abnormal vibration mode of the drone beacon. For example, the majority feature abnormality determination: when more than a certain number (such as more than half) of the time domain features are classified as abnormal categories, the drone beacon is determined to be in an abnormal vibration mode, etc. Or it can be a comprehensive weight determination: assign different weights to each time domain feature, calculate the weighted score of the abnormal feature, and when the score exceeds a certain threshold, the drone beacon is determined to be in an abnormal vibration mode, etc.

[0072] In this embodiment, the characteristic values ​​of the multi-dimensional time domain characteristics can be weighted based on the Gini importance coefficient to determine the abnormal vibration mode. Figure 3 This is described in detail as follows.

[0073] Reference Figure 3 , shows a flowchart of the steps of an abnormal vibration mode detection method provided by an embodiment of the present application. Figure 3 As shown, the abnormal vibration mode detection method may include: step 301, step 302, step 303 and step 304.

[0074] Step 301: calling the beacon abnormal vibration detection model to process the multi-dimensional time domain features to obtain the Gini importance coefficient of the multi-dimensional time domain features.

[0075] In this embodiment, the beacon abnormal vibration detection model can be a random forest model. The Gini importance coefficient is a method to measure the importance of a feature in a random forest model. The Gini importance coefficient reflects the importance of the feature in the model decision-making process. The larger the coefficient, the greater the influence of the feature on the model's judgment of abnormal vibration. The dynamic optimization judgment logic can be executed through the Gini importance coefficient in the random forest model. During the training process, the AI ​​(Artificial Intelligence) model quantifies the contribution of each feature to the abnormal judgment (such as the weight distribution of the mean, variance, etc.), and adjusts the feature weight according to the actual data distribution, thereby reducing the false alarm rate and enhancing the adaptability to complex vibration patterns. If the peak feature importance of the vibration signal is significantly higher than other features (such as variance, etc.), the model will prioritize the peak change to trigger the alarm, and combine the auxiliary judgment of other features, thereby reducing misjudgment in complex flight environments (such as sudden strong winds, etc.) and accurately identifying physical demolition behaviors.

[0076] After obtaining the multi-dimensional time domain features, the beacon abnormal vibration detection model can be called to process the multi-dimensional time domain features to obtain the Gini importance coefficient of the multi-dimensional time domain features. The calculation formula can be as follows: (1) (2) In the above formulas (1) and (2), For each time domain feature The Gini importance coefficient of is the Gini impurity (used to measure the probability of randomly selecting a sample from a set and labeling it incorrectly), is the number of feature categories, For the The tree belongs to the category The sample proportion of is the category index (the value range is 1 to , assuming that In the tree categories, then from 1 to ), is the total number of trees in the random forest model, Time domain features In the The reduction in Gini impurity caused by splitting on a tree, For the The sum is calculated for all nodes in the tree.

[0077] After the beacon abnormal vibration detection model is called to process the multi-dimensional time domain features and obtain the Gini importance coefficient of the multi-dimensional time domain features, step 302 is executed.

[0078] Step 302: Determine the feature weight corresponding to the multi-dimensional time domain feature according to the Gini importance coefficient.

[0079] After calling the beacon abnormal vibration detection model to process the multi-dimensional time domain features and obtaining the Gini importance coefficient of the multi-dimensional time domain features, the feature weights corresponding to the multi-dimensional time domain features can be determined according to the Gini importance coefficient. Specifically, in order to obtain reasonable feature weights, the calculated Gini importance coefficients can be normalized. Because the value ranges of the Gini importance coefficients of different features may be different, directly using these coefficients as weights may cause the influence of certain features to be too large or too small. The normalization method can be to divide the Gini importance coefficient of each feature by the sum of the Gini importance coefficients of all features. The value obtained after normalization is the feature weight corresponding to each multi-dimensional time domain feature. These weights reflect the proportion of each feature in the comprehensive judgment of whether the beacon is vibrating abnormally.

[0080] After determining the feature weights corresponding to the multi-dimensional time-domain features according to the Gini importance coefficient, step 303 is performed.

[0081] Step 303: Perform weighted processing on the multi-dimensional time domain features based on the feature weights to obtain time domain feature values.

[0082] After determining the feature weights corresponding to the multi-dimensional time-domain features according to the Gini importance coefficient, the multi-dimensional Haas features can be weighted based on the feature weights to obtain the time-domain feature values. Specifically, for each multi-dimensional time-domain feature, its feature value is multiplied by the corresponding feature weight. Such weighted calculations are performed on all multi-dimensional time-domain features, and all weighted feature values ​​are added together to obtain a comprehensive time-domain feature value. This time-domain feature value comprehensively considers the time-domain features of each dimension and their importance, and can more accurately reflect the vibration state of the drone beacon.

[0083] After weighted processing is performed on the multi-dimensional time-domain features based on the feature weights to obtain the time-domain feature values, step 304 is executed.

[0084] Step 304: Detect whether the drone beacon is in an abnormal vibration mode based on the time domain eigenvalue.

[0085] After weighting the multi-dimensional time domain features based on the feature weights to obtain the time domain feature values, it is possible to detect whether the drone beacon is in an abnormal vibration mode based on the time domain feature values. Specifically, during the model training process, a suitable threshold is usually determined based on a large amount of training data (it can be understood that the threshold can be dynamically adjusted according to the continuous adjustment process of the model). This threshold is used to distinguish between the beacon in a normal vibration mode and an abnormal vibration mode. Furthermore, the calculated time domain feature value can be compared with the set threshold. If the time domain feature value exceeds the threshold, it is determined that the drone beacon is in an abnormal vibration mode. If the time domain feature value is lower than or equal to the threshold, it is determined that the drone beacon is in a normal vibration mode.

[0086] The embodiment of the present application assigns feature weights to multi-dimensional time domain features through the Gini importance coefficient of time domain feature data (such as mean, variance, etc.) to judge the abnormal vibration mode of the drone beacon, which can improve the accuracy and real-time performance of the dynamic anti-dismantling mechanism.

[0087] Step 104: When it is determined that the drone beacon is in a physically dismantled state based on the condition that the drone beacon is in an abnormal vibration mode, clear the data in the drone beacon.

[0088] After detecting that the drone beacon is in an abnormal vibration mode, the condition that the detected drone beacon is in an abnormal vibration mode can be obtained, and based on the condition, it can be determined whether the drone beacon is in a physically removed state. The process of determining that the drone beacon is in a physically removed state based on the condition that the detected drone beacon is in an abnormal vibration mode can be combined with Figure 4 This is described in detail as follows.

[0089] Reference Figure 4 , shows a flow chart of the steps of a method for determining a physical removal state provided by an embodiment of the present application. Figure 4 As shown, the method for determining the physical dismantling status may include: step 401 and step 402.

[0090] Step 401: When the number of detections of the abnormal vibration mode reaches a threshold number, it is determined that the drone beacon is in a physically removed state.

[0091] In this embodiment, when the number of detections of the abnormal vibration pattern reaches a number threshold, it can be determined that the drone beacon is in a physically removed state.

[0092] In this example, a time period can be preset, such as 3 hours, 2 hours, etc. If the drone beacon is detected to be in an abnormal vibration mode multiple times within the time period (i.e., the number threshold is reached), it can be determined that the drone beacon is in a physically removed state.

[0093] The embodiment of the present application determines that the drone beacon is in a physically dismantled state through the detection results of multiple abnormal vibration patterns, which can avoid some external factors from causing wrong judgments on the state of the drone beacon and improve the accuracy of judging the physical dismantling state of the drone beacon.

[0094] Step 402: When the abnormal vibration pattern indicates that the variation amplitude of the time domain characteristics of the vibration data reaches a variation amplitude threshold, it is determined that the drone beacon is in a physically removed state.

[0095] When the abnormal vibration pattern indicates that the amplitude of the change of the time domain characteristics of the vibration data reaches the amplitude of change threshold, it can be determined that the drone beacon is in a physically removed state. Specifically, when the amplitude of the change of multiple or all of the six time domain characteristics corresponding to the vibration data collected by the abnormal vibration pattern is very large (i.e., reaches the amplitude of change threshold), it can be determined that the drone beacon is in a physically removed state.

[0096] When it is determined that the drone beacon is in a physically dismantled state, the data in the drone beacon can be cleared to avoid data leakage.

[0097] The embodiment of the present application dynamically detects abnormal vibrations of drone beacons through a beacon abnormal vibration detection model, implements a dynamic anti-dismantling mechanism, effectively prevents data leakage, and significantly improves the security and adaptability of the drone beacon anti-dismantling mechanism, as well as the security of drone data.

[0098] Next, combine Figure 5 and Figure 6 The process and system of model training and anomaly detection are described in detail.

[0099] like Figure 6 As shown, the abnormal vibration detection model (i.e., random forest model) can be set up in the ground station to build an AI reasoning platform.

[0100] like Figure 5 As shown, the process of model training and anomaly detection may include: 1. Data collection: The vibration sensor collects the vibration data of the drone in real time during operation and sends the real-time data to the drone beacon.

[0101] 2. Data preprocessing: The drone beacon preprocesses the collected vibration data through filtering and noise reduction, de-averaging, and normalization to facilitate subsequent feature extraction and model training.

[0102] 3. Feature extraction: The drone beacon extracts time domain features from the preprocessed data, calculates the mean, variance, root mean square value, peak value and other time domain features of the vibration signal, and reflects the changes in the overall vibration level of the drone through changes in the features.

[0103] 4. Model training: The drone beacon sends the extracted feature values ​​to the AI ​​reasoning platform. The AI ​​reasoning platform uses the random forest algorithm to train the extracted features and build an algorithm model that can distinguish between normal and abnormal vibration patterns.

[0104] 5. UAV vibration monitoring: After the drone and the drone beacon are intelligently bound, the drone vibration monitoring and abnormal alarm process during the operation of the drone can be: 1) Data collection and feature extraction: The drone beacon continuously receives vibration data sent from the drone vibration sensor, and extracts time domain features of the vibration data to generate a set of eigenvalues. 2) Eigenvalue transmission: The drone beacon sends the extracted eigenvalues ​​to the AI ​​reasoning platform every minute. 3) AI reasoning platform abnormality judgment: After receiving the eigenvalues, the AI ​​reasoning platform analyzes the data through the pre-trained algorithm model to determine whether an abnormality occurs. 4) Abnormal feedback and status judgment: Result return: The AI ​​reasoning platform returns the abnormal result to the drone beacon in real time; Beacon removal judgment: The beacon judges whether it has been physically removed based on the returned abnormal information and the preset logic (such as continuous abnormalities or sudden drops in eigenvalues).

[0105] 6. Abnormal alarm: If it is determined that the drone beacon has been physically removed, the drone beacon will immediately send an alarm message to the ground station and clear the data in the beacon.

[0106] The embodiments of the present application have the following technical effects: 1. Method for binding drone beacons to drones based on AI technology: By collecting vibration sensor data and combining it with AI model training technology, the intelligent and safe binding of drone beacons to drones can be achieved, significantly improving safety and reliability.

[0107] 2. Dynamic anti-dismantling mechanism: The AI ​​model is used to dynamically detect abnormal vibrations of drone beacons and implement a dynamic anti-dismantling mechanism to effectively prevent data leakage. This significantly improves the security and adaptability of the drone beacon anti-dismantling mechanism and the security of drone data.

[0108] 3. Application of AI technology in drones: AI algorithms are used to train the extracted feature values ​​and build algorithm models, which improves the accuracy and generalization ability of the models.

[0109] The beacon data processing method provided in the embodiment of the present application obtains the vibration data of the drone during operation collected by a vibration sensor, and the vibration sensor is set on the drone. The vibration data is processed to obtain the multi-dimensional time domain features corresponding to the vibration data. The pre-trained beacon abnormal vibration detection model is called to process the multi-dimensional time domain features to detect whether the drone beacon set on the drone is in an abnormal vibration mode. When it is determined that the drone beacon is in a physically dismantled state based on the condition that the detected drone beacon is in an abnormal vibration mode, the data in the drone beacon is cleared. The embodiment of the present application dynamically detects the abnormal vibration of the drone beacon through the beacon abnormal vibration detection model, realizes a dynamic anti-dismantling mechanism, effectively prevents data leakage, and significantly improves the security and adaptability of the drone beacon anti-dismantling mechanism, as well as the security of drone data.

[0110] Reference Figure 7 , shows a schematic diagram of the structure of a beacon data processing device provided in an embodiment of the present application. Figure 7 As shown, the beacon data processing device 700 may include the following modules: A vibration data acquisition module 710 is used to acquire vibration data of the UAV during operation collected by a vibration sensor; the vibration sensor is disposed on the UAV; A time domain feature acquisition module 720 is used to process the vibration data to obtain multi-dimensional time domain features corresponding to the vibration data; The abnormal vibration detection module 730 is used to call a pre-trained beacon abnormal vibration detection model to process the multi-dimensional time domain features to detect whether the drone beacon set on the drone is in an abnormal vibration mode; The beacon data clearing module 740 is used to clear the data in the drone beacon when it is determined that the drone beacon is in a physically dismantled state based on the condition that the drone beacon is in an abnormal vibration mode.

[0111] Optionally, the time domain feature acquisition module includes: A mean value calculation unit, used to calculate the average value of the vibration data in the time domain to obtain the mean value of the vibration data in the time domain; A variance calculation unit, used for calculating the variance of the vibration data in the time domain; A root mean square value calculation unit, used to calculate the root mean square value of the vibration data in the time domain; A peak value calculation unit, used to calculate the maximum value of the vibration data in the time domain, and use the maximum value as the peak value of the vibration data in the time domain; A peak-to-peak value calculation unit, used to calculate the difference between the maximum value and the minimum value of the vibration data in the time domain, and use the difference as the peak-to-peak value of the vibration data in the time domain; A skewness calculation unit, used for calculating the skewness of the vibration data in the time domain; The time domain feature acquisition unit is used to use the mean, variance, root mean square value, peak value, peak-to-peak value and skewness of the vibration data in the time domain as multi-dimensional time domain features corresponding to the vibration data.

[0112] Optionally, the abnormal vibration detection module includes: A feature classification unit, used for calling the beacon abnormal vibration detection model to classify the multi-dimensional time domain features according to a preset abnormality detection rule to obtain classified time domain features; The abnormal vibration determination unit is used to determine whether the drone beacon is in an abnormal vibration mode according to the classification time domain features and the normal feature values ​​corresponding to the classification time domain features in accordance with the abnormal judgment rules.

[0113] Optionally, the abnormal vibration determination unit includes: The abnormal mode determination subunit is used to determine whether the drone beacon is in an abnormal vibration mode according to the size relationship between the preset characteristic value of the classification time domain feature and the set multiple of the corresponding normal characteristic value when the abnormal judgment rule is a joint feature judgment rule, and the classification time domain feature includes: at least two of: mean, variance, root mean square value, peak value, peak-to-peak value and skewness.

[0114] Optionally, the abnormal vibration detection module includes: A coefficient acquisition unit, used for calling the beacon abnormal vibration detection model to process the multi-dimensional time domain features to obtain the Gini importance coefficient of the multi-dimensional time domain features; A feature weight determination unit, used to determine the feature weight corresponding to the multi-dimensional time domain feature according to the Gini importance coefficient; A time domain feature value acquisition unit, configured to perform weighted processing on the multi-dimensional time domain features based on the feature weights to obtain time domain feature values; The abnormal mode detection subunit is used to detect whether the drone beacon is in an abnormal vibration mode according to the time domain characteristic value.

[0115] Optionally, the beacon data clearing module includes: A first state determination unit is configured to determine that the drone beacon is in a physical removal state when the number of detections of the abnormal vibration mode reaches a number threshold; The second state determination unit is used to determine that the drone beacon is in a physical removal state when the abnormal vibration pattern indicates that the change amplitude of the time domain characteristic of the vibration data reaches a change amplitude threshold.

[0116] Optionally, the time domain feature acquisition module includes: A data preprocessing unit, used to preprocess the vibration data according to a preprocessing method to obtain preprocessed vibration data; The time domain feature acquisition unit is used to process the pre-processed vibration data to obtain the multi-dimensional time domain features.

[0117] The beacon data processing device provided in the embodiment of the present application obtains the vibration data of the drone during operation collected by a vibration sensor, and the vibration sensor is arranged on the drone. The vibration data is processed to obtain the multi-dimensional time domain features corresponding to the vibration data. The pre-trained beacon abnormal vibration detection model is called to process the multi-dimensional time domain features to detect whether the drone beacon set on the drone is in an abnormal vibration mode. When it is determined that the drone beacon is in a physically dismantled state based on the condition that the detected drone beacon is in an abnormal vibration mode, the data in the drone beacon is cleared. The embodiment of the present application dynamically detects the abnormal vibration of the drone beacon through the beacon abnormal vibration detection model, realizes the dynamic anti-dismantling mechanism, effectively prevents data leakage, and significantly improves the security and adaptability of the drone beacon anti-dismantling mechanism, as well as the security of drone data.

[0118] An embodiment of the present application also provides an electronic device, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program implements the above-mentioned beacon data processing method when executed by the processor.

[0119] Figure 8 FIG. 8 is a schematic diagram showing the structure of an electronic device 800 according to an embodiment of the present invention. Figure 8 As shown, the electronic device 800 includes a central processing unit (CPU) 801, which can perform various appropriate actions and processes according to computer program instructions stored in a read-only memory (ROM) 802 or computer program instructions loaded from a storage unit 808 to a random access memory (RAM) 803. In the RAM 803, various programs and data required for the operation of the electronic device 800 can also be stored. The CPU 801, the ROM 802, and the RAM 803 are connected to each other via a bus 804. An input / output (I / O) interface 805 is also connected to the bus 804.

[0120] Multiple components in the electronic device 800 are connected to the I / O interface 805, including: an input unit 806, such as a keyboard, a mouse, a microphone, etc.; an output unit 807, such as various types of displays, speakers, etc.; a storage unit 808, such as a disk, an optical disk, etc.; and a communication unit 809, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 809 allows the electronic device 800 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0121] The various processes and processing described above may be performed by the processing unit 801. For example, the method of any of the above embodiments may be implemented as a computer software program, which is tangibly contained in a computer-readable medium, such as the storage unit 808. In some embodiments, part or all of the computer program may be loaded and / or installed on the electronic device 800 via the ROM 802 and / or the communication unit 809. When the computer program is loaded into the RAM 803 and executed by the CPU 801, one or more actions in the method described above may be performed.

[0122] In addition, an embodiment of the present application further provides a computer-readable storage medium on which a computer program is stored, and when the program is executed by a processor, the above-mentioned beacon data processing method is implemented.

[0123] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0124] Those skilled in the art will appreciate that the embodiments of the present application can be provided as methods, devices, or computer program products. Therefore, the present application can adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0125] The embodiments of the present application are described with reference to the flowcharts and / or block diagrams of the methods, terminals (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing terminal to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing terminal generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0126] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing terminal to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture including an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0127] These computer program instructions can also be loaded onto a computer or other programmable data processing terminal so that a series of operating steps are executed on the computer or other programmable terminal to produce a computer-implemented process, thereby providing instructions for implementing the process in the computer or other programmable terminal. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0128] Finally, it should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or terminal including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or terminal. In the absence of further restrictions, the elements defined by the statement "comprise a ..." do not exclude the presence of other identical elements in the process, method, article or terminal including the elements.

[0129] The beacon data processing method, device, electronic device and computer-readable storage medium provided by the present application are introduced in detail above. Specific examples are used in this article to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method of the present application and its core idea; at the same time, for general technical personnel in this field, according to the idea of ​​the present application, there will be changes in the specific implementation method and application scope. In summary, the content of this specification should not be understood as a limitation on the present application.

Claims

1. A beacon data processing method, characterized in that: The method comprises: Acquiring vibration data of the UAV during operation collected by a vibration sensor; the vibration sensor is arranged on the UAV; Processing the vibration data to obtain multi-dimensional time domain features corresponding to the vibration data; Calling a pre-trained beacon abnormal vibration detection model to process the multi-dimensional time domain features to detect whether the drone beacon set on the drone is in an abnormal vibration mode; When it is determined that the drone beacon is in a physically dismantled state based on the condition that the drone beacon is in an abnormal vibration mode, the data in the drone beacon is cleared.

2. The method according to claim 1, characterized in that The processing of the vibration data to obtain multi-dimensional time domain features corresponding to the vibration data includes: Calculating the average value of the vibration data in the time domain to obtain the mean value of the vibration data in the time domain; Calculating the variance of the vibration data in the time domain; Calculate the root mean square value of the vibration data in the time domain; Calculating the maximum value of the vibration data in the time domain, and taking the maximum value as the peak value of the vibration data in the time domain; Calculating the difference between the maximum value and the minimum value of the vibration data in the time domain, and taking the difference as the peak-to-peak value of the vibration data in the time domain; Calculating the skewness of the vibration data in the time domain; The mean, variance, root mean square value, peak value, peak-to-peak value and skewness of the vibration data in the time domain are used as multi-dimensional time domain features corresponding to the vibration data.

3. The method according to claim 2, characterized in that The calling of the pre-trained beacon abnormal vibration detection model to process the multi-dimensional time domain features to detect whether the drone beacon set on the drone is in an abnormal vibration mode includes: Calling the beacon abnormal vibration detection model to classify the multi-dimensional time domain features according to a preset abnormality detection rule to obtain classified time domain features; According to the abnormality determination rule, it is determined whether the drone beacon is in an abnormal vibration mode based on the classification time domain features and the normal feature values ​​corresponding to the classification time domain features.

4. The method according to claim 3, characterized in that The determining whether the drone beacon is in an abnormal vibration mode according to the abnormal determination rule and the classified time domain feature and the normal feature value corresponding to the classified time domain feature includes: In the case where the abnormality determination rule is a joint feature determination rule, whether the drone beacon is in an abnormal vibration mode is determined based on the size relationship between a preset feature value of the classification time domain feature and a set multiple of the corresponding normal feature value, and the classification time domain feature includes: at least two of: mean, variance, root mean square value, peak value, peak-to-peak value and skewness.

5. The method according to claim 1, characterized in that The calling of the pre-trained beacon abnormal vibration detection model to process the multi-dimensional time domain features to detect whether the drone beacon set on the drone is in an abnormal vibration mode includes: Calling the beacon abnormal vibration detection model to process the multi-dimensional time domain features to obtain the Gini importance coefficient of the multi-dimensional time domain features; Determining the feature weights corresponding to the multi-dimensional time domain features according to the Gini importance coefficient; Performing weighted processing on the multi-dimensional time domain features based on the feature weights to obtain time domain feature values; According to the time domain characteristic value, it is detected whether the drone beacon is in an abnormal vibration mode.

6. The method according to claim 1, characterized in that The determining that the drone beacon is in a physical removal state according to the detected condition that the drone beacon is in an abnormal vibration mode includes: When the number of detections of the abnormal vibration mode reaches a number threshold, determining that the drone beacon is in a physically removed state; or When the abnormal vibration pattern indicates that the variation amplitude of the time domain characteristic of the vibration data reaches a variation amplitude threshold, it is determined that the drone beacon is in a physically dismantled state.

7. The method according to claim 1, characterized in that The processing of the vibration data to obtain multi-dimensional time domain features corresponding to the vibration data includes: Preprocessing the vibration data according to a preprocessing method to obtain preprocessed vibration data; The pre-processed vibration data is processed to obtain the multi-dimensional time domain features.

8. A beacon data processing device, characterized in that: The device comprises: A vibration data acquisition module, used to acquire vibration data of the UAV during operation collected by a vibration sensor; the vibration sensor is arranged on the UAV; A time domain feature acquisition module, used to process the vibration data to obtain multi-dimensional time domain features corresponding to the vibration data; An abnormal vibration detection module, used to call a pre-trained beacon abnormal vibration detection model to process the multi-dimensional time domain features to detect whether the drone beacon set on the drone is in an abnormal vibration mode; The beacon data clearing module is used to clear the data in the drone beacon when it is determined that the drone beacon is in a physically dismantled state based on the condition that the drone beacon is in an abnormal vibration mode.

9. An electronic device, characterized in that: include: A processor, a memory, and a computer program stored in the memory and executable on the processor, wherein the processor implements the beacon data processing method according to any one of claims 1 to 7 when executing the program.

10. A computer-readable storage medium, characterized in that: When the instructions in the storage medium are executed by a processor of an electronic device, the electronic device is enabled to execute the beacon data processing method according to any one of claims 1 to 7.

Images