Unmanned aerial vehicle communication signal interference detection method based on big data

Abstract

The application discloses a method for detecting unmanned aerial vehicle communication signal interference based on big data, which comprises data acquisition, data preprocessing, unmanned aerial vehicle communication signal feature selection, construction of signal anomaly detection model, super parameter optimization and unmanned aerial vehicle communication signal interference detection. The application relates to the technical field of signal data processing, in particular to a method for detecting unmanned aerial vehicle communication signal interference based on big data. The method innovatively strengthens and retains the time sequence information of the frequency spectrum features, significantly improves the timeliness and accuracy of unmanned aerial vehicle communication signal interference detection, introduces bidirectional time sequence information fusion and designs a signal anomaly detection activation function to improve the stability of the detection model in various environments, and improves the algorithm for obtaining the optimal parameters of the detection model by obtaining a reverse individual, a dynamic nonlinear convergence factor and an individual search inertia weight update strategy, so as to obtain the optimal parameter combination and improve the accuracy of the communication interference signal detection result.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of signal data processing, and particularly to a UAV communication signal interference detection method based on big data. BACKGROUND

[0002] With the continuous development and application of UAV technology, UAVs are widely used in military, agricultural, logistics and aerial photography fields. The flight and task execution of UAVs rely on stable communication systems to ensure the stable transmission of flight instructions, data transmission and remote control signals. Therefore, a UAV communication signal interference detection method based on big data has emerged. This method collects a large amount of data information, uses big data analysis technology to detect communication signal interference in real time, and provides early warning and response measures in a timely manner, thereby significantly improving the communication stability, anti-interference ability and flight safety of UAVs, and effectively realizing the safety management of UAV groups.

[0003] However, the traditional UAV communication signal interference detection method has the problems of poor signal change adaptability, inability to effectively cope with noise and interference, and signal timing characteristic response lag. The existing signal anomaly detection model applicable to the signal has the technical problems of difficulty in adapting to interference signal changes and activation function response in complex environments, thereby affecting the accuracy of UAV communication interference signal detection. In the existing detection model, the built-in parameters are not properly set, thereby affecting the accuracy of the communication interference signal detection result. SUMMARY

[0004] In view of the above, in order to overcome the defects of the prior art, the present application provides a big data-based unmanned aerial vehicle communication signal interference detection method, in order to overcome the problems of poor signal change adaptability, inability to effectively respond to noise and interference, and signal timing characteristic response lag in the traditional unmanned aerial vehicle communication signal interference detection method, the present application innovatively introduces a dynamic window function and a logarithmic transformation technology, effectively enhances the key signal characteristics in the communication signal, significantly improves the signal detection accuracy in the strong noise and interference scene, further optimizes the identification of important components in the signal through spectrum feature enhancement, and combines the optimal spectrum path extraction technology to effectively filter out the frequency points with the most information, improve the interference signal differentiation and detection accuracy, and through the retention of feature timing information, the detection not only depends on the spectrum feature, but also can effectively capture the dynamic change of the signal in time, thereby significantly improving the timeliness and accuracy of the interference detection, realizing the safe management of the unmanned aerial vehicle; in view of the technical problems that the existing signal anomaly detection model is difficult to adapt to the change of the interference signal and the response of the activation function in the complex environment, thereby affecting the accuracy of the unmanned aerial vehicle communication interference signal detection, the present application innovatively introduces the design of bidirectional timing information fusion and signal anomaly detection activation function, the bidirectional timing information fusion can capture the future timing information, fully master the timing characteristics of the signal change and their mutual relationship, thereby more accurately reflecting the dynamic evolution of the signal, the signal anomaly detection activation function enhances the response in the interference signal area, improves the identification accuracy of the model to the abnormal signal, and at the same time enhances the response sensitivity of the model to different signal types, the method significantly improves the adaptive ability of the model, so that it can more flexibly cope with various complex interference scenes, thereby ensuring the stability and reliability of the detection model in various unmanned aerial vehicle flight environments; in view of the technical problem that the existing detection model has improper built-in parameter setting, thereby affecting the accuracy of the communication interference signal detection result, the present application improves the algorithm for obtaining the optimal parameters of the detection model through the acquisition of reverse individuals, dynamic nonlinear convergence factors and individual search inertia weight update strategy, thereby obtaining the optimal parameter combination, improving the accuracy of the communication interference signal detection result, and further enhancing the safety management of the unmanned aerial vehicle.

[0005] The technical scheme adopted by the present application is as follows: the big data-based unmanned aerial vehicle communication signal interference detection method provided by the present application comprises the following steps:

[0006] Step S1: data acquisition;

[0007] Step S2: data preprocessing;

[0008] Step S3: unmanned aerial vehicle communication signal feature selection;

[0009] Step S4: constructing a signal anomaly detection model;

[0010] Step S5: hyperparameter optimization;

[0011] Step S6: UAV communication signal interference detection.

[0012] Further, in step S1, the data collection, specifically from the UAV management cloud platform, through collection, obtains UAV communication interference detection raw data; the UAV communication interference detection raw data includes historical communication interference detection data and real-time communication interference detection data; the historical communication interference detection data and the real-time communication interference detection data both include UAV communication signal data, UAV flight environment data and UAV flight state data; the historical communication interference detection data further includes UAV communication signal interference mode.

[0013] Further, in step S2, the data preprocessing, specifically, data cleaning, data filtering, data normalization and data completion are performed on the UAV communication interference detection raw data to obtain UAV communication interference detection preliminary data; the data cleaning is used to eliminate invalid and inaccurate data, specifically by processing data missing values, data outliers and data repeated values; the data filtering is used to remove high-frequency noise and interference signals in the data, specifically by using a filtering algorithm to smooth the data; the standardization processing is based on the maximum and minimum normalization method to standardize the data; the data completion is used to complete the missing and uncollected data, specifically using interpolation technology to fill in the missing data points.

[0014] Further, in step S3, the UAV communication signal feature selection, specifically, the communication signal feature selection is performed on the UAV communication signal data in the UAV communication interference detection preliminary data, the key signal features of the communication signal are extracted, and the optimized UAV communication signal features are obtained;

[0015] Step S31: communication signal feature extraction, used to convert the UAV communication signal from time domain to frequency domain, and obtain the optimal spectral feature matrix; specifically including the following steps:

[0016] Step S311: convert to frequency domain signal, specifically convert the time domain signal to the frequency domain signal, and adopt a dynamic window function to adapt to the instantaneous change of the signal, the formula used is as follows:

[0017] ;

[0018] ;

[0019] In the formula, represents the distribution value of the communication signal at time t and frequency f, represents the n-th sampling point of the time domain signal, represents the Kaiser window function, denotes a reference window length, denotes a dynamic window length function, denotes a dynamic window length adjustment coefficient, denotes a signal instantaneous frequency derivative, denotes a Fourier transform kernel function;

[0020] Step S312: communication signal enhancement processing, specifically, important spectral features in the communication signal are enhanced through logarithmic transformation and normalization processing, and the formula used is as follows:

[0021] ;

[0022] In the formula, denotes the global spectral features of the enhanced communication signal, i denotes the index of the spectral features, denotes the number of sampling points;

[0023] Step S313: spectral feature strengthening, specifically, the spectral features are further strengthened; the formula used is as follows:

[0024] ;

[0025] In the formula, denotes a spectral feature strengthening function, denotes the amplitude of the spectral point , denotes the rth frequency point in the spectrum, denotes a weight for controlling the spectral amplitude, denotes a weight for controlling the spectral point, is a constant term for preventing the denominator from being zero, denotes a frequency difference function between adjacent spectral points, denotes the rth frequency point in the spectrum, denotes the rth frequency point in the spectrum; denotes the rth frequency point in the spectrum; denotes the rth frequency point in the spectrum;

[0026] Step S314: optimal spectral path extraction, specifically, the most representative spectral features are selected through optimal path extraction, and the formula used is as follows:

[0027] ;

[0028] In the formula, denotes a spectral feature function of the optimal spectral path extraction, denotes the spectral feature at the position , denotes the spectral feature at the position ;

[0029] Step S315: Obtain the high-dimensional optimal spectral feature matrix, using the following formula:

[0030] ;

[0031] In the formula, Represents the high-dimensional optimal spectral feature matrix. express Spectral characteristics at location express Spectral characteristics at location express The spectral characteristics at a location, where R represents the number of frequency points;

[0032] Step S32: Dimensionality reduction of communication signal features, used to reduce redundant information in the communication signal and lower the dimension of the feature matrix, specifically by using principal component analysis. Dimensionality reduction is performed by first calculating the covariance matrix of the feature matrix to measure the correlation between features; then selecting the eigenvector with the largest variance as the principal component, and projecting the signal data into a low-dimensional space to obtain the low-dimensional feature matrix. ;

[0033] Step S33: Preserve the timing characteristics of the communication signal. This preserves the timing information of the signal, enabling interference detection to be analyzed in conjunction with the time dimension, thereby improving detection accuracy. Specifically, a time-delay neural network is used to further enhance the signal features based on the timing information, resulting in optimized UAV communication signal features. The formula used is as follows:

[0034] ;

[0035] ;

[0036] In the formula, This represents the output of the j-th neuron at time t. Representing the low-dimensional feature matrix The first in One characteristic in time The value at time, This represents the bias parameters of the hidden layer. This represents the weight matrix of the hidden layer. This indicates the size of the time delay window for the hidden layer. This indicates the time delay step of the hidden layer. This indicates the size of the time delay window for the output layer. Indicates the time delay step of the output layer. Indicates the number of input features. This represents the temporal feature vector of the output of the r-th output neuron at time t, and represents the optimized UAV communication signal features. Indicates the time of the j-th neuron Output at any moment This indicates the number of neurons in the hidden layer. This represents the bias parameters of the output layer. This represents the weight matrix of the output layer. express function, express function.

[0037] Furthermore, in step S4, the construction of the signal anomaly detection model specifically includes the following steps:

[0038] Step S41: Obtain the input data for the signal anomaly detection model. Specifically, this involves concatenating the optimized UAV communication signal features, UAV flight environment data, and UAV flight status data to obtain the input data for the signal anomaly detection model. The formula used is as follows:

[0039] ;

[0040] In the formula, This represents the input data for the signal anomaly detection model. This indicates the drone flight environment data in the preliminary data for drone communication interference detection. This refers to the drone flight status data in the preliminary data for drone communication interference detection;

[0041] Step S42: GRU bidirectional unit output calculation, used to capture the timing changes of UAV communication signals during interference; the formula used is as follows:

[0042] ;

[0043] ;

[0044] ;

[0045] In the formula, express Unit execution function, This indicates a combination of forward and backward views. The hidden state of the unit, This indicates the current hidden state in the positive direction. This indicates the hidden state in the positive direction of the previous time step. This indicates the hidden state in the opposite direction at the current moment. Indicates the hidden state in the opposite direction at the next moment;

[0046] Step S43: Design the signal anomaly detection activation function, specifically an interference model for accurately identifying UAV communication signals; the formula used is as follows:

[0047] ;

[0048] In the formula, Let represent the activation function for signal anomaly detection, and let x represent the input variable of the activation function for signal anomaly detection. This indicates the amplification factor controlling the negative value region. This indicates the sensitivity of the activation function's response. Indicates adjustment The offset of the activation function. It represents the overall magnitude of the output in the negative value region and is a constant. This indicates the amplification factor controlling the positive value region. This indicates that the output in the positive range is shifted. express function;

[0049] Step S44: Obtain the model output results using the following formula:

[0050] ;

[0051] In the formula, This represents the results of the model. b represents the weight matrix corresponding to the output. y This represents the bias term parameters of the output layer.

[0052] Furthermore, in step S5, the hyperparameter optimization specifically includes the following steps:

[0053] Step S51: Initialize parameters, specifically by constructing initial algorithm parameters; the initial algorithm parameters include population size N and maximum number of iterations. ;

[0054] Step S52: Initialize the population, specifically by using the chaotic perturbation mapping initialization method to initialize the search individuals, including the following steps:

[0055] Step S521: Generate N chaotic sequence values ​​and initialize the random value z0 using the following formula:

[0056] ;

[0057] In the formula, This represents the (i+1)th chaotic sequence value in the d-th dimension. This represents the i-th chaotic sequence value in the d-th dimension. Indicates being between A random number uniformly distributed within a range;

[0058] Step S522: Chaotic sequence value transformation search individual, using the following formula:

[0059] ;

[0060] In the formula, This represents the initial position of the i-th individual in the d-th dimension. and Let represent the lower and upper bounds of the search space for dimension d in the t-th iteration, respectively;

[0061] Step S53: Obtain the reverse individual position using the following formula:

[0062] ;

[0063] In the formula, This represents the position of the i-th reverse individual in the d-th dimension of the t-th generation population. Indicates being between Random numbers uniformly distributed within a range This represents the position of the i-th individual in the d-th dimension within the t-th generation of the population.

[0064] Step S54: Calculate the fitness value, specifically by calculating the fitness value f of each individual in the population. i The performance of the signal anomaly detection model based on individual location is used as the fitness value of the individual.

[0065] Step S55: Reselect the search individual, specifically selecting an individual with better fitness as the new search individual; the formula used is as follows:

[0066] ;

[0067] In the formula, Represents the fitness function;

[0068] Step S56: Set individual search parameters, specifically by balancing global and local search capabilities using a dynamic nonlinear convergence factor; the formula used is as follows:

[0069] ;

[0070] ;

[0071] ;

[0072] ;

[0073] In the formula, t represents the current iteration number. Indicates the convergence factor in the current iteration. Indicates the search control parameters. This represents the distance between the currently searched individual and a randomly selected search individual. This represents the distance between the current searched individual and the optimal searched individual. Indicates a random search for individuals. This represents the position of the individual with the best fitness value in the population during the t-th iteration on dimension d. and Indicates being between Random numbers uniformly distributed within a range This represents the decay parameter that controls the linear convergence factor;

[0074] Step S57: Update individual search inertia weights, using the following formula:

[0075] ;

[0076] In the formula, This represents the individual search inertia weight in the current iteration. This represents the parameter for adjusting the individual search inertia weight;

[0077] Step S58: Individual location update, using the following formula:

[0078] ;

[0079] In the formula, This represents the position of the i-th individual in the d-th dimension within the (t+1)-th generation of the population. Indicates being between Random numbers uniformly distributed within a range This represents the parameters that control the shape of the spiral search. This indicates the growth rate of the spiral search;

[0080] Step S59: Search determination, specifically, by constructing search termination conditions, the optimal individual position is determined, and the optimal individual position data setting is obtained;

[0081] The search termination conditions include threshold termination and iteration termination;

[0082] The threshold termination specifically involves setting a fitness threshold, whereby the individual fitness value f... i When the fitness threshold is exceeded, the hyperparameter search is complete.

[0083] The term "iteration termination" specifically refers to terminating the iteration and obtaining the optimal individual position when the maximum number of iterations is reached.

[0084] The optimal individual location includes the combination of parameters for the signal anomaly detection model.

[0085] Further, in step S6, the UAV communication signal interference detection specifically involves taking the UAV flight environment data, UAV flight status data, and optimized UAV communication signal characteristics from the preliminary UAV communication interference detection data as input data and passing them to the hyperparameter-optimized signal anomaly detection model to obtain the UAV communication signal interference detection result. If the UAV communication signal interference detection result is an abnormal signal, the system automatically switches to the backup communication link to ensure uninterrupted communication and sends an interference alarm to the operator to report the UAV communication interference status.

[0086] The beneficial effects achieved by the present invention using the above solution are as follows:

[0087] (1) In view of the problems of poor adaptability to signal changes, inability to effectively deal with noise and interference, and lag in response of signal timing characteristics in traditional UAV communication signal interference detection methods, this solution innovatively introduces dynamic window function and logarithmic transformation technology, which effectively enhances the key signal features in communication signals and significantly improves the signal detection accuracy in scenarios with strong noise and interference. Through spectral feature enhancement, the identification of important components in the signal is further optimized. Combined with the optimal spectral path extraction technology, the most informative frequency points are effectively screened, improving the discrimination and detection accuracy of interference signals. By retaining the feature timing information, the detection not only depends on spectral features, but can also effectively capture the dynamic changes of the signal in time, thereby significantly improving the timeliness and accuracy of interference detection and realizing the safe management of UAVs.

[0088] (2) In view of the technical problem that existing signal anomaly detection models are difficult to adapt to changes in interference signals and activation function responses in complex environments, which leads to the inaccuracy of UAV communication interference signal detection, this solution innovatively introduces the design of bidirectional temporal information fusion and signal anomaly detection activation function. Bidirectional temporal information fusion can capture future temporal information, fully grasp the temporal characteristics of signal changes and their interrelationships, and thus more accurately reflect the dynamic evolution of signals. The signal anomaly detection activation function enhances the model's recognition accuracy of abnormal signals by enhancing the response in the interference signal region, and at the same time enhances the model's response sensitivity to different signal types. This method significantly improves the model's adaptability, enabling it to more flexibly cope with various complex interference scenarios, thereby ensuring the stability and reliability of the detection model in various UAV flight environments.

[0089] (3) In view of the technical problem that the existing detection model has inappropriate built-in parameter settings, which leads to inaccurate detection results of communication interference signals, this solution improves the algorithm for obtaining the optimal parameters of the detection model by obtaining reverse individuals, dynamic nonlinear convergence factors and individual search inertia weight update strategies, thereby obtaining the optimal parameter combination, improving the accuracy of communication interference signal detection results, and further enhancing the safety management of UAVs. Attached Figure Description

[0090] Figure 1 This is a flowchart illustrating the UAV communication signal interference detection method based on big data provided by the present invention.

[0091] Figure 2 This is a flowchart illustrating step S3;

[0092] Figure 3 This is a flowchart illustrating step S4;

[0093] Figure 4 This is a flowchart illustrating step S5;

[0094] Figure 5 This is a flowchart illustrating step S31;

[0095] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used together with the embodiments of the invention to explain the invention and do not constitute a limitation thereof. Detailed Implementation

[0096] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0097] In the description of this invention, it should be understood that the terms "upper", "lower", "front", "rear", "left", "right", "top", "bottom", "inner", "outer", etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the system or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0098] Example 1, see Figure 1 The technical solution adopted by this invention is as follows: The UAV communication signal interference detection method based on big data provided by this invention includes the following steps:

[0099] Step S1: Data acquisition. Raw data for UAV communication interference detection is obtained by collecting historical and real-time communication interference detection data from the UAV management cloud platform.

[0100] Step S2: Data preprocessing. By cleaning, filtering, normalizing, and completing the raw data of UAV communication interference detection, preliminary data of UAV communication interference detection is obtained.

[0101] Step S3: UAV communication signal feature selection. By performing communication signal feature selection on the UAV communication signal data in the preliminary data for UAV communication interference detection, key signal features of the communication signals are extracted to obtain optimized UAV communication signal features.

[0102] Step S4: Construct a signal anomaly detection model. The signal anomaly detection model is established by using the output of the GRU bidirectional unit and designing a signal anomaly detection activation function.

[0103] Step S5: Hyperparameter optimization. By improving the algorithm for obtaining the optimal parameters of the detection model, the optimal parameter combination of the signal anomaly detection model is obtained.

[0104] Step S6: UAV communication signal interference detection. By passing the data to the hyperparameter-optimized signal anomaly detection model, the UAV communication signal interference detection results are obtained.

[0105] Example 2, see Figure 1 This embodiment is based on the above embodiment. In step S1, the data acquisition specifically involves collecting raw data on UAV communication interference detection from the UAV management cloud platform. The raw data on UAV communication interference detection includes historical communication interference detection data and real-time communication interference detection data. Both the historical and real-time communication interference detection data include UAV communication signal data, UAV flight environment data, and UAV flight status data. The historical communication interference detection data also includes UAV communication signal interference modes. The UAV communication signal interference modes include normal signal, mild interference, moderate interference, and severe interference. The UAV communication signal data includes communication signal strength, communication signal-to-noise ratio, communication signal bit error rate, communication signal frequency, communication signal transmission rate, and communication signal modulation method. The UAV flight environment data includes flight temperature data, flight humidity data, flight air pressure data, flight wind speed data, and flight background noise data. The UAV flight status data includes UAV flight altitude, UAV flight position, and UAV flight speed.

[0106] Example 3, see Figure 1This embodiment is based on the above embodiment. In step S2, the data preprocessing specifically involves data cleaning, data filtering, data normalization, and data completion of the raw data for UAV communication interference detection to obtain preliminary data for UAV communication interference detection. The data cleaning is used to eliminate invalid and inaccurate data, specifically by processing missing data values, outliers, and duplicate data values. The data filtering is used to remove high-frequency noise and interference signals from the data, specifically by smoothing the data using a filtering algorithm. The normalization process is based on the minimum-maximum normalization method to standardize the data. The data completion is used to complete missing and uncollected data, specifically by using interpolation techniques to fill in missing data points.

[0107] Example 4, see Figure 1 , Figure 2 and Figure 5 This embodiment is based on the above embodiment. In step S3, the selection of UAV communication signal features specifically involves selecting communication signal features from the UAV communication signal data in the preliminary data of UAV communication interference detection, extracting key signal features of the communication signals, and obtaining optimized UAV communication signal features.

[0108] Step S31: Communication signal feature extraction, used to convert the UAV communication signal from the time domain to the frequency domain and obtain the optimal spectral feature matrix; specifically including the following steps:

[0109] Step S311: Convert to frequency domain signal. Specifically, convert the time domain signal to a frequency domain signal and use a dynamic window function to adapt to the instantaneous changes in the signal. The formula used is as follows:

[0110] ;

[0111] ;

[0112] In the formula, This represents the distribution of the communication signal over time t and frequency f. This represents the nth sampling point of the time-domain signal. This represents the Kaiser window function. Indicates the reference window length, This represents the dynamic window length function. This represents the dynamic window length adjustment coefficient. The instantaneous frequency derivative of the signal. Represents the Fourier transform kernel function;

[0113] Step S312: Communication signal enhancement processing, specifically, enhances important spectral features in the communication signal through logarithmic transformation and normalization, using the following formula:

[0114] ;

[0115] In the formula, This represents the global spectral characteristics of the enhanced communication signal, where i represents the index of the spectral characteristic. Indicates the number of sampling points;

[0116] Step S313: Spectral feature enhancement, specifically, further enhancing the spectral features; the formula used is as follows:

[0117] ;

[0118] In the formula, Represents the spectral feature enhancement function. Represents the spectrum points amplitude, This represents the r-th frequency point in the spectrum. The weights that control the amplitude of the spectrum This represents the weight of the control spectrum points. It is a constant term used to prevent the denominator from being zero. A function representing the frequency difference between adjacent spectral points. Indicates the first in the spectrum Frequency point, Indicates the first in the spectrum Frequency point;

[0119] Step S314: Optimal Spectrum Path Extraction. Specifically, this involves extracting the most representative spectral features using the optimal path extraction formula, as follows:

[0120] ;

[0121] In the formula, This represents the spectral feature function extracted by the optimal spectral path. express Spectral characteristics at location express Spectral characteristics at location;

[0122] Step S315: Obtain the high-dimensional optimal spectral feature matrix, using the following formula:

[0123] ;

[0124] In the formula, Represents the high-dimensional optimal spectral feature matrix. express Spectral characteristics at location express Spectral characteristics at location express The spectral characteristics at a location, where R represents the number of frequency points;

[0125] Step S32: Dimensionality reduction of communication signal features, used to reduce redundant information in the communication signal and lower the dimension of the feature matrix, specifically by using principal component analysis. Dimensionality reduction is performed by first calculating the covariance matrix of the feature matrix to measure the correlation between features; then selecting the eigenvector with the largest variance as the principal component, and projecting the signal data into a low-dimensional space to obtain the low-dimensional feature matrix. ;

[0126] Step S33: Preserve the timing characteristics of the communication signal. This preserves the timing information of the signal, enabling interference detection to be analyzed in conjunction with the time dimension, thereby improving detection accuracy. Specifically, a time-delay neural network is used to further enhance the signal features based on the timing information, resulting in optimized UAV communication signal features. The formula used is as follows:

[0127] ;

[0128] ;

[0129] In the formula, This represents the output of the j-th neuron at time t. Representing the low-dimensional feature matrix The first in One characteristic in time The value at time, This represents the bias parameters of the hidden layer. This represents the weight matrix of the hidden layer. This indicates the size of the time delay window for the hidden layer. This indicates the time delay step of the hidden layer. This indicates the size of the time delay window for the output layer. Indicates the time delay step of the output layer. Indicates the number of input features. This represents the temporal feature vector of the output of the r-th output neuron at time t, and represents the optimized UAV communication signal features. Indicates the time of the j-th neuron Output at any moment This indicates the number of neurons in the hidden layer. This represents the bias parameters of the output layer. This represents the weight matrix of the output layer. express function, express function.

[0130] By performing the above operations, this solution addresses the problems of poor signal change adaptability, inability to effectively cope with noise and interference, and lag in response to signal timing characteristics in traditional UAV communication signal interference detection methods. It innovatively introduces dynamic window functions and logarithmic transform techniques to effectively enhance key signal features in communication signals, significantly improving signal detection accuracy in scenarios with strong noise and interference. Through spectral feature enhancement, the identification of important components in the signal is further optimized. Combined with optimal spectral path extraction technology, the most informative frequency points are effectively selected, improving the discriminability and detection accuracy of interference signals. By preserving characteristic timing information, detection not only relies on spectral features but also effectively captures the dynamic changes of the signal over time, thereby significantly improving the timeliness and accuracy of interference detection and achieving safe management of UAVs.

[0131] Example 5, see Figure 1 and Figure 3 This embodiment is based on the above embodiment. In step S4, the construction of the signal anomaly detection model specifically includes the following steps:

[0132] Step S41: Obtain the input data for the signal anomaly detection model. Specifically, this involves concatenating the optimized UAV communication signal features, UAV flight environment data, and UAV flight status data to obtain the input data for the signal anomaly detection model. The formula used is as follows:

[0133] ;

[0134] In the formula, This represents the input data for the signal anomaly detection model. This indicates the drone flight environment data in the preliminary data for drone communication interference detection. This refers to the drone flight status data in the preliminary data for drone communication interference detection;

[0135] Step S42: GRU bidirectional unit output calculation, used to capture the timing changes of UAV communication signals during interference; the formula used is as follows:

[0136] ;

[0137] ;

[0138] ;

[0139] In the formula, express Unit execution function, This indicates a combination of forward and backward views. The hidden state of the unit, This indicates the current hidden state in the positive direction. This indicates the hidden state in the positive direction of the previous time step. This indicates the hidden state in the opposite direction at the current moment. Indicates the hidden state in the opposite direction at the next moment;

[0140] Step S43: Design the signal anomaly detection activation function, specifically an interference model for accurately identifying UAV communication signals; the formula used is as follows:

[0141] ;

[0142] In the formula, Let represent the activation function for signal anomaly detection, and let x represent the input variable of the activation function for signal anomaly detection. This indicates the amplification factor controlling the negative value region. This indicates the sensitivity of the activation function's response. Indicates adjustment The offset of the activation function. It represents the overall magnitude of the output in the negative value region and is a constant. This indicates the amplification factor controlling the positive value region. This indicates that the output in the positive range is shifted. express function;

[0143] Step S44: Obtain the model output results using the following formula:

[0144] ;

[0145] In the formula, This represents the results of the model. b represents the weight matrix corresponding to the output. y This represents the bias term parameters of the output layer.

[0146] By performing the above operations, this solution addresses the technical problem of existing signal anomaly detection models struggling to adapt to changes in interference signals and activation function responses in complex environments, thus affecting the accuracy of UAV communication interference signal detection. This innovative approach introduces bidirectional temporal information fusion and a signal anomaly detection activation function. Bidirectional temporal information fusion captures future temporal information, comprehensively grasping the temporal characteristics of signal changes and their interrelationships, thereby more accurately reflecting the dynamic evolution of signals. The signal anomaly detection activation function enhances the response in interference signal regions, improving the model's accuracy in identifying anomaly signals and enhancing its sensitivity to different signal types. This method significantly improves the model's adaptability, enabling it to more flexibly cope with various complex interference scenarios, thus ensuring the stability and reliability of the detection model in various UAV flight environments.

[0147] Example 6, see Figure 1 and Figure 4 This embodiment is based on the above embodiment. In step S5, the hyperparameter optimization specifically includes the following steps:

[0148] Step S51: Initialize parameters, specifically by constructing initial algorithm parameters; the initial algorithm parameters include population size N and maximum number of iterations. ;

[0149] Step S52: Initialize the population, specifically by using the chaotic perturbation mapping initialization method to initialize the search individuals, including the following steps:

[0150] Step S521: Generate N chaotic sequence values ​​and initialize the random value z0 using the following formula:

[0151] ;

[0152] In the formula, This represents the (i+1)th chaotic sequence value in the d-th dimension. This represents the i-th chaotic sequence value in the d-th dimension. Indicates being between A random number uniformly distributed within a range;

[0153] Step S522: Chaotic sequence value transformation search individual, using the following formula:

[0154] ;

[0155] In the formula, This represents the initial position of the i-th individual in the d-th dimension. and Let represent the lower and upper bounds of the search space for dimension d in the t-th iteration, respectively;

[0156] Step S53: Obtain the reverse individual position using the following formula:

[0157] ;

[0158] In the formula, This represents the position of the i-th reverse individual in the d-th dimension of the t-th generation population. Indicates being between Random numbers uniformly distributed within a range This represents the position of the i-th individual in the d-th dimension within the t-th generation of the population.

[0159] Step S54: Calculate the fitness value, specifically by calculating the fitness value f of each individual in the population. iThe performance of the signal anomaly detection model based on individual location is used as the fitness value of the individual.

[0160] Step S55: Reselect the search individual, specifically selecting an individual with better fitness as the new search individual; the formula used is as follows:

[0161] ;

[0162] In the formula, Represents the fitness function;

[0163] Step S56: Set individual search parameters, specifically by balancing global and local search capabilities using a dynamic nonlinear convergence factor; the formula used is as follows:

[0164] ;

[0165] ;

[0166] ;

[0167] ;

[0168] In the formula, t represents the current iteration number. Indicates the convergence factor in the current iteration. Indicates the search control parameters. This represents the distance between the currently searched individual and a randomly selected search individual. This represents the distance between the current searched individual and the optimal searched individual. Indicates a random search for individuals. This represents the position of the individual with the best fitness value in the population during the t-th iteration on dimension d. and Indicates being between Random numbers uniformly distributed within a range This represents the decay parameter that controls the linear convergence factor;

[0169] Step S57: Update individual search inertia weights, using the following formula:

[0170] ;

[0171] In the formula, This represents the individual search inertia weight in the current iteration. This represents the parameter for adjusting the individual search inertia weight;

[0172] Step S58: Individual location update, using the following formula:

[0173] ;

[0174] In the formula, This represents the position of the i-th individual in the d-th dimension within the (t+1)-th generation of the population. Indicates being between Random numbers uniformly distributed within a range This represents the parameters that control the shape of the spiral search. This indicates the growth rate of the spiral search;

[0175] Step S59: Search determination, specifically, by constructing search termination conditions, the optimal individual position is determined, and the optimal individual position data setting is obtained;

[0176] The search termination conditions include threshold termination and iteration termination;

[0177] The threshold termination specifically involves setting a fitness threshold, whereby the individual fitness value f... i When the fitness threshold is exceeded, the hyperparameter search is complete.

[0178] The term "iteration termination" specifically refers to terminating the iteration and obtaining the optimal individual position when the maximum number of iterations is reached.

[0179] The optimal individual location includes the combination of parameters for the signal anomaly detection model.

[0180] By performing the above operations, this solution addresses the technical problem of inaccurate communication interference signal detection results due to inappropriate built-in parameter settings in existing detection models. It improves the algorithm for obtaining optimal parameters of the detection model by acquiring reverse individuals, dynamic nonlinear convergence factors, and individual search inertia weight update strategies. This results in the optimal parameter combination, improving the accuracy of communication interference signal detection results and further enhancing the safety management of UAVs.

[0181] Example 7, see Figure 1 This embodiment is based on the above embodiment. In step S6, the UAV communication signal interference detection specifically involves taking the UAV flight environment data, UAV flight status data, and optimized UAV communication signal characteristics from the preliminary UAV communication interference detection data as input data and passing them to the hyperparameter-optimized signal anomaly detection model to obtain the UAV communication signal interference detection result. If the UAV communication signal interference detection result is an abnormal signal, the system automatically switches to the backup communication link to ensure uninterrupted communication and sends an interference alarm to the operator to report the UAV communication interference status.

[0182] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0183] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention.

[0184] The present invention and its embodiments have been described above. This description is not restrictive, and the accompanying drawings are only one embodiment of the present invention; the actual structure is not limited thereto. In conclusion, if those skilled in the art are inspired by this description and design similar structures and embodiments without departing from the spirit of the invention, such designs should fall within the protection scope of the present invention.

Claims

1. A method for detecting interference in UAV communication signals based on big data, characterized in that: The method includes the following steps: Step S1: Data acquisition. By acquiring data, the raw data for UAV communication interference detection is obtained. Step S2: Data preprocessing, by cleaning, filtering, normalizing and completing the raw data of UAV communication interference detection, preliminary data of UAV communication interference detection is obtained; Step S3: UAV communication signal feature selection. By using spectral feature enhancement and optimal spectral path selection, a high-dimensional optimal spectral feature matrix is ​​obtained. The high-dimensional optimal spectral feature matrix is ​​then subjected to feature dimensionality reduction. Finally, a time-delay neural network is used to perform time-series correlation modeling on the dimensionality-reduced features to obtain the optimized UAV communication signal features. Step S4: Construct a signal anomaly detection model. By using the optimized UAV communication signal characteristics and preliminary UAV communication interference detection data as input data, and introducing bidirectional time series information fusion and a signal anomaly detection activation function, a signal anomaly detection model is established, thus obtaining the signal anomaly detection model. Step S5: Hyperparameter optimization. The algorithm for obtaining the optimal parameters of the detection model is improved by obtaining reverse individuals, dynamic nonlinear convergence factors, and individual search inertia weight update strategies, thereby obtaining the optimal parameter combination of the signal anomaly detection model. Step S6: UAV communication signal interference detection. By passing real-time data to the hyperparameter-optimized signal anomaly detection model, the UAV communication signal interference detection results are obtained.

2. The method for detecting UAV communication signal interference based on big data according to claim 1, characterized in that: In step S3, the UAV communication signal feature selection specifically involves selecting communication signal features from the UAV communication signal data in the preliminary data of UAV communication interference detection, extracting key signal features of the communication signals, and obtaining optimized UAV communication signal features. Step S31: Communication signal feature extraction, used to convert the UAV communication signal from the time domain to the frequency domain and obtain the optimal spectral feature matrix; Specifically, the following steps are included: Step S311: Convert to frequency domain signal. Specifically, convert the time domain signal to a frequency domain signal and use a dynamic window function to adapt to the instantaneous changes in the signal. The formula used is as follows: ； ； In the formula, This represents the distribution of the communication signal over time t and frequency f. This represents the nth sampling point of the time-domain signal. This represents the Kaiser window function. Indicates the reference window length, This represents the dynamic window length function. This represents the dynamic window length adjustment coefficient. The instantaneous frequency derivative of the signal. Represents the Fourier transform kernel function; Step S312: Communication signal enhancement processing, specifically, enhances important spectral features in the communication signal through logarithmic transformation and normalization, using the following formula: ； In the formula, This represents the global spectral characteristics of the enhanced communication signal, where i represents the index of the spectral characteristic. Indicates the number of sampling points; Step S313: Spectral feature enhancement, specifically, further enhancing the spectral features; the formula used is as follows: ； In the formula, Represents the spectral feature enhancement function. Represents the spectrum points amplitude, This represents the r-th frequency point in the spectrum. The weights that control the amplitude of the spectrum This represents the weight of the control spectrum points. It is a constant term used to prevent the denominator from being zero. A function representing the frequency difference between adjacent spectral points. Indicates the first in the spectrum Frequency point, Indicates the first in the spectrum Frequency point; Step S314: Optimal Spectrum Path Extraction. Specifically, this involves extracting the most representative spectral features using the optimal path extraction formula, as follows: ； In the formula, This represents the spectral feature function extracted by the optimal spectral path. express Spectral characteristics at location express Spectral characteristics at location; Step S315: Obtain the high-dimensional optimal spectral feature matrix, using the following formula: ； In the formula, Represents the high-dimensional optimal spectral feature matrix. express Spectral characteristics at location express Spectral characteristics at location express The spectral characteristics at a location, where R represents the number of frequency points; Step S32: Dimensionality reduction of communication signal features, used to reduce redundant information in the communication signal and lower the dimension of the feature matrix, specifically by using principal component analysis. Dimensionality reduction is performed; first, the covariance matrix of the feature matrix is ​​calculated to measure the correlation between features. Then, the eigenvector with the largest variance is selected as the principal component, and the signal data is projected into a low-dimensional space to obtain the low-dimensional feature matrix. ; Step S33: Preserve the timing characteristics of the communication signal. This preserves the timing information of the signal, enabling interference detection to be analyzed in conjunction with the time dimension, thereby improving detection accuracy. Specifically, a time-delay neural network is used to further enhance the signal features based on the timing information, resulting in optimized UAV communication signal features. The formula used is as follows: ； ； In the formula, This represents the output of the j-th neuron at time t. Representing the low-dimensional feature matrix The first in One characteristic in time The value at time, This represents the bias parameters of the hidden layer. This represents the weight matrix of the hidden layer. This indicates the size of the time delay window for the hidden layer. This indicates the time delay step of the hidden layer. This indicates the size of the time delay window for the output layer. Indicates the time delay step of the output layer. Indicates the number of input features. This represents the temporal feature vector of the output of the r-th output neuron at time t, and represents the optimized UAV communication signal features. Indicates the time of the j-th neuron Output at any moment This indicates the number of neurons in the hidden layer. This represents the bias parameters of the output layer. This represents the weight matrix of the output layer. express function, express function.

3. The method for detecting UAV communication signal interference based on big data according to claim 1, characterized in that: In step S4, the construction of the signal anomaly detection model specifically includes the following steps: Step S41: Obtain the input data for the signal anomaly detection model. Specifically, this involves concatenating the optimized UAV communication signal features, UAV flight environment data, and UAV flight status data to obtain the input data for the signal anomaly detection model. The formula used is as follows: ； In the formula, This represents the input data for the signal anomaly detection model. This indicates the drone flight environment data in the preliminary data for drone communication interference detection. This refers to the drone flight status data in the preliminary data for drone communication interference detection; Step S42: GRU bidirectional unit output calculation, used to capture the timing changes of UAV communication signals during interference; the formula used is as follows: ； ； ； In the formula, express Unit execution function, This indicates a combination of forward and backward inputs. The hidden state of the unit, This indicates the current hidden state in the positive direction. This indicates the hidden state in the positive direction of the previous time step. This indicates the hidden state in the opposite direction at the current moment. Indicates the hidden state in the opposite direction at the next moment; Step S43: Design the signal anomaly detection activation function, specifically an interference model for accurately identifying UAV communication signals; the formula used is as follows: ； In the formula, Let represent the activation function for signal anomaly detection, and let x represent the input variable of the activation function for signal anomaly detection. This indicates the amplification factor controlling the negative value region. This indicates the sensitivity of the activation function's response. Indicates adjustment The offset of the activation function. It represents the overall magnitude of the output in the negative value region and is a constant. This indicates the amplification factor controlling the positive value region. This indicates that the output in the positive range is shifted. express function; Step S44: Obtain the model output results using the following formula: ； In the formula, This represents the results of the model. b represents the weight matrix corresponding to the output. y This represents the bias term parameters of the output layer.

4. The method for detecting UAV communication signal interference based on big data according to claim 1, characterized in that: In step S5, the hyperparameter optimization specifically includes the following steps: Step S51: Initialize parameters, specifically construct the initial parameters of the algorithm; the initial parameters of the algorithm include the population size N and the maximum number of iterations. ; Step S52: Initialize the population, specifically by using the chaotic perturbation mapping initialization method to initialize the search individuals, including the following steps: Step S521: Generate N chaotic sequence values ​​and initialize the random value z0 using the following formula: ； In the formula, This represents the (i+1)th chaotic sequence value in the d-th dimension. This represents the i-th chaotic sequence value in the d-th dimension. Indicates being between A random number uniformly distributed within a range; Step S522: Chaotic sequence value transformation search individual, using the following formula: ； In the formula, This represents the initial position of the i-th individual in the d-th dimension. and Let represent the lower and upper bounds of the search space for dimension d in the t-th iteration, respectively; Step S53: Obtain the reverse individual position using the following formula: ； In the formula, This represents the position of the i-th reverse individual in the d-th dimension of the t-th generation population. Indicates being between Random numbers uniformly distributed within a range This represents the position of the i-th individual in the d-th dimension within the t-th generation of the population. Step S54: Calculate the fitness value, specifically by calculating the fitness value f of each individual in the population. i The performance of the signal anomaly detection model based on individual location is used as the fitness value of the individual. Step S55: Reselect the search individual, specifically selecting an individual with better fitness as the new search individual; the formula used is as follows: ； In the formula, Represents the fitness function; Step S56: Set individual search parameters, specifically by balancing global and local search capabilities using a dynamic nonlinear convergence factor; the formula used is as follows: ； ； ； ； In the formula, t represents the current iteration number. Indicates the convergence factor in the current iteration. Indicates the search control parameters. This represents the distance between the currently searched individual and a randomly selected search individual. This represents the distance between the current searched individual and the optimal searched individual. Indicates a random search for individuals. This represents the position of the individual with the best fitness value in the population during the t-th iteration on dimension d. and Indicates being between Random numbers uniformly distributed within a range This represents the decay parameter that controls the linear convergence factor; Step S57: Update individual search inertia weights, using the following formula: ； In the formula, This represents the individual search inertia weight in the current iteration. This represents the parameter for adjusting the individual search inertia weight; Step S58: Individual location update, using the following formula: ； In the formula, This represents the position of the i-th individual in the d-th dimension within the (t+1)-th generation of the population. Indicates being between Random numbers uniformly distributed within a range This represents the parameters that control the shape of the spiral search. This indicates the growth rate of the spiral search; Step S59: Search determination, specifically, by constructing search termination conditions, the optimal individual position is determined, and the optimal individual position data setting is obtained; The search termination conditions include threshold termination and iteration termination; The threshold termination specifically involves setting a fitness threshold, whereby the individual fitness value f... i When the fitness threshold is exceeded, the hyperparameter search is complete. The term "iteration termination" specifically refers to terminating the iteration and obtaining the optimal individual position when the maximum number of iterations is reached. The optimal individual location includes the combination of parameters for the signal anomaly detection model.

5. The method for detecting UAV communication signal interference based on big data according to claim 1, characterized in that: In step S6, the UAV communication signal interference detection specifically involves taking the UAV flight environment data, UAV flight status data, and optimized UAV communication signal characteristics from the preliminary UAV communication interference detection data as input data and passing them to the hyperparameter-optimized signal anomaly detection model to obtain the UAV communication signal interference detection result. If the UAV communication signal interference detection result is an abnormal signal, the system automatically switches to the backup communication link to ensure uninterrupted communication and sends an interference alarm to the operator to report the UAV communication interference status.

6. The method for detecting UAV communication signal interference based on big data according to claim 1, characterized in that: In step S1, the data acquisition specifically involves collecting raw data on UAV communication interference detection from the UAV management cloud platform. The raw data on UAV communication interference detection includes historical communication interference detection data and real-time communication interference detection data. Both the historical and real-time communication interference detection data include UAV communication signal data, UAV flight environment data, and UAV flight status data. The historical communication interference detection data also includes UAV communication signal interference patterns.

7. The method for detecting UAV communication signal interference based on big data according to claim 1, characterized in that: In step S2, the data preprocessing specifically involves data cleaning, data filtering, data normalization, and data completion of the raw data for UAV communication interference detection to obtain preliminary data for UAV communication interference detection. Data cleaning is used to eliminate invalid and inaccurate data, specifically by processing missing values, outliers, and duplicate values. Data filtering is used to remove high-frequency noise and interference signals from the data, specifically by smoothing the data using filtering algorithms. Data normalization is based on the minimum-max normalization method to standardize the data. Data completion is used to fill in missing and uncollected data, specifically by using interpolation techniques to fill in missing data points.

Images