An unmanned aerial vehicle flight state detection method based on spectrum detection technology

Abstract

The present application relates to the technical field of radio orientation, and particularly relates to a UAV flight state detection method based on spectrum detection technology, comprising: obtaining an initial time domain signal of an initial spectrum signal of a UAV in a monitoring period during flight; segmenting the initial time domain signal according to flight states to screen out each complex change period; performing signal decomposition on a time domain sub-signal of each complex change period to obtain a plurality of components; determining an interference frequency tendency index of each component, determining an interference feature comprehensive index of each component according to the interference frequency tendency index, an oscillation intensity characteristic value and a correlation coefficient of each component, and then reconstructing a spectrum sub-signal corresponding to the time domain sub-signal to obtain a target spectrum signal. The data reference value of the target spectrum signal obtained through adaptive filtering is greater, which is conducive to improving the accuracy of UAV flight state detection.

Description

Technical Field

[0001] This invention relates to the field of radio direction finding technology, and more specifically to a method for detecting the flight status of a UAV based on spectrum detection technology. Background Technology

[0002] In recent years, the drone racing sport has flourished, creating an increasingly urgent need for high-precision, real-time detection of its flight status. Currently, radio frequency detection technology based on spectrum analysis is one of the effective ways to achieve this goal.

[0003] However, the electromagnetic environment at drone racing events is extremely complex. Cameras, image transmission equipment, remote controllers, and numerous spectators' electronic devices operate simultaneously, generating a large number of electromagnetic signals of different frequencies and types. These signals superimpose with the drone's radio frequency signals within a confined space, creating severe co-channel interference, adjacent-channel interference, and multipath interference, leading to distortion and attenuation of the radio frequency signals. This complex electromagnetic background noise significantly increases the difficulty of effectively separating and analyzing the drone's true state from the mixed spectrum, resulting in a higher false alarm rate for existing detection systems and insufficient reliability of the acquired flight trajectory, attitude, and other data. Summary of the Invention

[0004] To address the technical problem of low reliability of spectral signals used for flight status detection due to interference from multi-source heterogeneous signals in complex electromagnetic environments, the present invention aims to provide a UAV flight status detection method based on spectrum detection technology. The specific technical solution adopted is as follows: One embodiment of the present invention provides a method for detecting the flight status of a UAV based on spectrum detection technology, the method comprising the following steps: Acquire the initial spectrum signal of the UAV during the monitoring period in flight, and convert the initial spectrum signal into an initial time domain signal; The initial time-domain signal is segmented according to the flight status to filter out various complex change periods; the complex change periods consist of at least multiple interspersed acceleration moments, deceleration moments, unstable acceleration moments, and unstable deceleration moments; For each complex change period, the time-domain sub-signal is decomposed to obtain several components; based on the frequency distribution set corresponding to the target energy set of each component, the interference frequency tendency index of each component is determined. Obtain the oscillation intensity characteristic value of each component and its correlation coefficient with other components in the time domain sub-signal; determine the comprehensive index of interference characteristics for each component based on the interference frequency tendency index, oscillation intensity characteristic value and correlation coefficient of each component; Based on the comprehensive index of interference characteristics of each component, the spectral sub-signal corresponding to the time-domain sub-signal is reconstructed to obtain the target spectral signal.

[0005] Furthermore, the segmentation of the initial time-domain signal based on the flight state to filter out various complex changing time periods includes: The amplitude sequence is obtained based on the initial time-domain signal; Based on the amplitude difference between two adjacent moments in the amplitude sequence, each normal time period is determined; the normal time periods include hovering periods, acceleration periods, and deceleration periods. Within the monitoring period, the time range between any two of the normal periods is considered as the complex change period.

[0006] Further, determining each normal time period based on the amplitude difference between two adjacent moments in the amplitude sequence includes: Calculate the amplitude difference between the next moment and the previous moment in the amplitude sequence, take two moments whose absolute value of the amplitude difference is less than or equal to the amplitude threshold as hovering moments, and form a hovering period by two or more consecutive hovering moments; All moments outside the hovering period within the monitoring period are recorded as undetermined moments, and the two undetermined moments with positive amplitude differences are taken as acceleration moments, and the two undetermined moments with negative amplitude differences are taken as deceleration moments; Based on each acceleration and deceleration moment within the monitoring period, each acceleration period and deceleration period are determined.

[0007] Furthermore, determining the various acceleration and deceleration periods based on the various acceleration and deceleration moments within the monitoring period includes: For each time period of a continuously distributed target time, obtain the amplitude difference sequence corresponding to the time period; the target time is an acceleration time or a deceleration time. Based on the differences between every two adjacent elements in the amplitude difference sequence, determine the stability attribute index for each target time. The target time when the stability attribute index is greater than the stability threshold is determined as the stable target time, and two or more consecutive stable target times are combined to form a target time period; the target time period is an acceleration period or a deceleration period.

[0008] Further, determining the interference frequency tendency index for each component based on the frequency distribution set corresponding to the target energy set for each component includes: For each component, determine the maximum energy corresponding to the component, and determine the target energy set based on the maximum energy; Obtain the frequency distribution set corresponding to the target energy set of the component, and determine the interference frequency tendency index of the component based on the similarity between the frequency distribution set and the reference frequency distribution set; the similarity and the interference frequency tendency index are negatively correlated.

[0009] Further, obtaining the oscillation intensity characteristic value of each component includes: For each component, the kurtosis of each peak is calculated in the time-domain sub-signal of the component; the kurtosis is used at least to characterize the state of the peak. The oscillation intensity characteristic value of the component is determined based on the kurtosis of all the peaks corresponding to the component; the kurtosis is positively correlated with the oscillation intensity characteristic value.

[0010] Further, determining the comprehensive interference characteristic index for each component based on the interference frequency tendency index, oscillation intensity characteristic value, and correlation coefficient of each component includes: For each component, the interference characteristic factor of the component is determined based on the correlation coefficient of the component. The product of the interference frequency tendency index, the oscillation intensity characteristic value, and the interference characteristic factor of the component is calculated, and the product is used as the comprehensive interference characteristic index of the component.

[0011] Further, determining the interference characteristic factor of the component based on each correlation coefficient of the component includes: Determine the absolute value of each correlation coefficient for the component, and then calculate the average of the absolute values ​​of all correlation coefficients; The average value is subjected to negative correlation processing, and the resulting negative correlation value is used as the interference characteristic factor of the component.

[0012] Further, the reconstructing of the spectral sub-signal corresponding to the time-domain sub-signal based on the comprehensive index of the interference characteristics of each component to obtain the target spectral signal includes: For each time-domain sub-signal, the interference intensity of the corresponding spectral sub-signal is determined based on the comprehensive index of the interference characteristics of each component of the time-domain sub-signal. The initial step size in the filtering algorithm is adjusted according to the interference intensity of the spectrum sub-signal to determine the adaptive step size when the spectrum sub-signal is filtered. The adaptive step-size filtering algorithm is used to filter the spectrum sub-signal to obtain a new spectrum sub-signal; The target spectrum signal is obtained by replacing the spectrum sub-signals corresponding to the complex changing time periods in the initial spectrum signal with each of the new spectrum sub-signals.

[0013] Furthermore, after obtaining the target spectrum signal, the process also includes: Based on the target spectrum signal, the flight status of the UAV during the monitoring period is detected.

[0014] The present invention has the following beneficial effects: This invention provides a method for detecting the flight status of a UAV based on spectrum detection technology. This method segments the signal according to the flight status, facilitating adaptive denoising of spectral sub-signals during complex and changing time periods. This enhances the targeting of interference suppression, achieving precise denoising in key areas and enabling subsequent denoising strategies to be state-adaptive. The method determines the comprehensive interference characteristic index for each component, performing interference degree analysis and calculation on a component-by-component basis. This reflects the signal's fluctuation characteristics at different frequencies and also reveals the specific frequency and variation characteristics of different types of interference. The comprehensive interference characteristic index integrates interference characteristics from multiple perspectives, facilitating the acquisition of a target spectrum signal with higher denoising reliability. This means filtering out the influence of interference factors on the spectrum signal to a certain extent, clearly presenting the spectral characteristics related to the UAV's flight status and highlighting the key features reflecting the UAV's flight status. Finally, detecting the UAV's flight status through the target spectrum signal helps to more accurately identify UAV flight parameters, improves the accuracy of UAV flight status detection, and provides strong support for UAV design and performance improvement. Attached Figure Description

[0015] To more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 A flowchart illustrating the steps of a method for detecting the flight status of a UAV based on spectrum detection technology, as provided in one embodiment of the present invention; Figure 2 This is a flowchart illustrating the implementation of step S2 in an embodiment of the present invention; Figure 3 This is a flowchart illustrating the implementation of determining the interference frequency tendency index of each component based on the frequency distribution set corresponding to the target energy set of each component in this embodiment of the invention. Figure 4 This is a flowchart illustrating the implementation of step S5 in an embodiment of the present invention. Detailed Implementation

[0017] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the specific implementation methods, structures, features, and effects of the technical solution proposed according to the present invention are described in detail below with reference to the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0018] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0019] To remove interference from other factors in the acquired spectrum signal and enable it to better reflect the flight status of the UAV, one embodiment of the present invention provides a method for detecting the flight status of a UAV based on spectrum detection technology, such as... Figure 1 As shown, it includes the following steps: S1: Acquire the initial spectrum signal of the UAV during the monitoring period during flight, and convert the initial spectrum signal into an initial time domain signal.

[0020] To detect the flight status of a drone using spectrum sensing technology, it is necessary to acquire the drone's spectrum signal, including: The first step is to determine the monitoring frequency range based on the relevant technical data or product manual of the drone.

[0021] The second step is to select an antenna that can cover the monitored frequency range. It should be noted that, considering the drone may fly at different distances and in different directions, an antenna with appropriate gain and directionality must be selected.

[0022] The third step is to select a spectrum analyzer with sufficient frequency coverage, high resolution, adjustable video bandwidth, and a suitable scan time range to meet the analysis needs of different signals.

[0023] The fourth step is to connect the selected antenna to the spectrum analyzer via an RF cable, and set the relevant parameters of the spectrum analyzer to meet the signal acquisition requirements.

[0024] The fifth step involves scanning the signals within the determined monitoring frequency band based on the spectrum analyzer after setup, in order to obtain the initial spectrum signal of the monitoring period during the drone's flight.

[0025] The acquisition of the initial spectrum signal can be customized by the implementer according to the specific scenario requirements and needs, and no specific limitation is made here. As for the duration of the monitoring period, it should at least include one complete state change cycle of the UAV. Those skilled in the art can select and adjust it within the range of 10 to 60 seconds according to the actual application scenario, such as racing, aerial photography or logistics transportation.

[0026] Time-domain analysis is more suitable for capturing the dynamic behavior and transient state changes of UAVs during flight. In order to facilitate subsequent data analysis, the initial spectral signal needs to be converted into an initial time-domain signal.

[0027] In one embodiment, the initial spectral signal is converted into an initial time-domain signal using the discrete Fourier transform.

[0028] The implementation process of the Discrete Fourier Transform is an existing technology and will not be described in detail here.

[0029] S2 segments the initial time-domain signal according to the flight status to filter out various complex changing time periods.

[0030] Here, the complex change period consists of at least multiple interspersed acceleration moments, deceleration moments, unstable acceleration moments, and unstable deceleration moments, such as unstable acceleration or deceleration, sudden maneuvers, and specific modes of electromagnetic interference.

[0031] During periods of complex change, drastic changes in motor power may generate broadband electromagnetic noise, or rapid attitude adjustments may lead to significant multipath effects. Therefore, the spectral sub-signals corresponding to these periods of complex change are precisely the data with the most concentrated electromagnetic interference and the worst signal quality. Adaptive denoising of the spectral sub-signals during these complex periods enhances the targeting of interference suppression, achieving precise denoising in key areas and enabling subsequent denoising strategies to possess state adaptability.

[0032] The flight state of a drone changes during flight, and the spectral signal exhibits different characteristics in several aspects under different flight states. For example, when hovering, the amplitude of its time-domain signal fluctuates very little, remaining at a relatively stable level; during flight acceleration, the amplitude increases significantly, and the rate of change increases markedly; when turning, in order to maintain balance and change direction, the power output of the motors is constantly adjusted, causing the amplitude of the relevant signal to fluctuate again. Therefore, based on the amplitude changes of the collected time-domain signal, the entire flight period can be initially segmented, that is, the time periods of different flight states can be initially determined.

[0033] The basic flight states—horizontal cruise, hovering, acceleration, deceleration, and turning—cover the fundamental actions of a drone in changing its position, speed, and direction in the air. Other complex flight states are combinations and superpositions of these basic states. Since racing drones need to frequently change speed, direction, and altitude during competitions, they rarely have the opportunity for prolonged horizontal cruise. Furthermore, turning also involves continuous changes in direction and speed. Therefore, this embodiment defines the time periods corresponding to hovering, acceleration, and deceleration as normal time periods, and determines the various complex change time periods based on these normal time periods.

[0034] As an exemplary implementation, step S2 described above can be achieved through... Figure 2 The steps shown are to be implemented as follows: S21, obtain the amplitude sequence based on the initial time-domain signal.

[0035] In one embodiment, the amplitude sequence corresponding to the initial time-domain signal can be obtained by using the direct time-domain extraction method based on the initial time-domain signal.

[0036] S22, determine each normal time period based on the amplitude difference between two adjacent moments in the amplitude sequence.

[0037] Here, the normal period includes the hovering period, the acceleration period, and the deceleration period, meaning that its flight process is predictable and the resulting spectrum is continuous and smooth.

[0038] Since the spectral variations during normal flight periods are part of the UAV's normal flight state and are signals that need to be accurately measured, noise that does not need to be removed should be preserved to avoid excessive smoothing or distortion that might result from forceful denoising. However, for periods of complex variation, deep denoising is required to isolate abnormal and strong interference.

[0039] As an exemplary implementation, step S22 described above can be achieved through the following steps: The first step is to calculate the amplitude difference between the next moment and the previous moment in the amplitude sequence, and to take the two moments whose absolute value of the amplitude difference is less than or equal to the amplitude threshold as hovering moments, and to form a hovering period by taking two or more consecutive hovering moments.

[0040] In one embodiment, the amplitude difference between the next and previous moments in the amplitude sequence is calculated as the amplitude difference to represent the change in amplitude between two adjacent moments. The amplitude threshold is taken as an empirical value of 3. When the absolute value of a certain amplitude difference is less than or equal to the amplitude threshold, it indicates that the amplitude change of the corresponding two adjacent moments is relatively stable and the UAV is in a hovering state. The corresponding two adjacent moments can be taken as hovering moments. All hovering moments are obtained, and two or more consecutive hovering moments are combined into a hovering period.

[0041] The magnitude threshold can be set by the implementer based on the specific drone model and hovering flight standards; no specific limit is set here.

[0042] The second step is to record all moments outside the hovering period within the monitoring period as undetermined moments, and to record the two undetermined moments with positive amplitude differences as acceleration moments, and the two undetermined moments with negative amplitude differences as deceleration moments.

[0043] In one embodiment, after acquiring each hovering period, other times in the monitoring period other than the hovering period are taken as undetermined times; for two adjacent undetermined times, if the amplitude difference between the two adjacent undetermined times is positive, the two adjacent undetermined times are determined to be acceleration times, and if the amplitude difference between the two adjacent undetermined times is negative, the two adjacent undetermined times are determined to be deceleration times.

[0044] The third step is to determine the acceleration and deceleration periods based on the acceleration and deceleration times within the monitoring period.

[0045] Because the amplitude fluctuates when a drone turns, sometimes increasing and sometimes decreasing, two or more consecutive acceleration or deceleration moments may also be part of the turning process. Therefore, acceleration and deceleration periods cannot be divided solely based on the continuity of similar moments; the stability of amplitude changes must be considered.

[0046] As an exemplary implementation, the various acceleration and deceleration periods are determined, including: The first sub-step is to obtain the amplitude difference sequence corresponding to each time period of the target time period with continuous distribution.

[0047] Here, the process for determining the acceleration period and deceleration time is the same. To reduce unnecessary descriptions, both acceleration and deceleration times are referred to as target times. It is worth noting that the target period is determined by analyzing times of the same attribute.

[0048] During the acceleration or deceleration of the drone, the motor power changes relatively steadily. Although there may be some small fluctuations, the overall amplitude fluctuation frequency is not high, so the amplitude difference between adjacent target moments should be close.

[0049] In one embodiment, for each time period of a continuously distributed target time, the amplitude difference between every two adjacent target times within the time period is first obtained to form an amplitude difference sequence.

[0050] The second sub-step involves determining the stability attribute index for each target time based on the differences between every two adjacent elements in the amplitude difference sequence.

[0051] In one embodiment, the absolute value of the difference between every two adjacent elements in the amplitude difference sequence is calculated to represent the difference between two adjacent amplitude differences; the absolute values ​​of all differences are processed by the maximum-minimum normalization to obtain normalized values ​​limited to between 0 and 1; then the obtained normalized values ​​are subjected to negative correlation processing, i.e., taking the reciprocal, to obtain the stable attribute index corresponding to the three target times, and thus obtain the stable attribute index for each target time.

[0052] Of course, an exponential function with the natural constant as the base, i.e., exp(-), can also be used to normalize the negative correlation of the data in order to obtain stable attribute indicators.

[0053] The third sub-step is to determine the target time when the stable attribute index is greater than the stable threshold as the stable target time, and to form a target time period by combining two or more consecutive stable target times.

[0054] Here, the target time period is either the acceleration period or the deceleration period.

[0055] In one embodiment, the stability threshold can be an empirical value of 0.6 or 0.7. If a certain stability attribute index is greater than the stability threshold, it indicates that the amplitude difference between adjacent target times is close, the amplitude fluctuation is small, and the acceleration or deceleration is stable. In this case, each target time corresponding to the calculation of the stability attribute index is taken as a stable target time, and two or more consecutive stable target times are combined into a target time period. The size of the stability threshold can be set by the implementer according to the specific actual situation, and is not specifically limited here.

[0056] Based on the above process of determining the target time period, the various acceleration and deceleration periods within the monitoring period can be obtained.

[0057] S23, within the monitoring period, the time range between any two normal periods is taken as the complex change period.

[0058] In one embodiment, the time range between any two adjacent normal time periods is defined as a complex change period. These complex change periods do not overlap.

[0059] Thus, this embodiment has obtained all the complex change periods within the monitoring period.

[0060] It should be noted that by selecting complex and variable periods for targeted processing, the processing efficiency has been significantly improved. It is understood that by performing in-depth processing only on a small proportion of complex periods, the computational overhead can be significantly reduced, enabling the system to have real-time processing capabilities.

[0061] After obtaining the complex change periods, the process also includes: if some periods belong to neither normal periods nor complex change periods, then a backup processing strategy is activated. This backup processing strategy includes, but is not limited to: treating these partial periods as complex change periods or recording the data for subsequent model optimization. For isolated moments, they can be assigned to periods of any adjacent attributes.

[0062] It should also be noted that complex and changing time periods contain the most discriminative features of UAV flight. Applying uniform and excessive filtering to these periods could obscure important state characteristics. This application, through targeted processing, removes interference while preserving, to the maximum extent possible, the key spectral features reflecting the true flight state. This provides a cleaner and more reliable data foundation for subsequent flight state analysis, namely, acquiring the target spectral signal. This helps reduce the false positive and false negative rates in state identification, improving the accuracy and reliability of the entire detection system. Specifically, this corresponds to steps S3 to S5 below.

[0063] S3 decomposes the time-domain sub-signal for each complex change period to obtain several components; and determines the interference frequency tendency index of each component based on the frequency distribution set corresponding to the target energy set of each component.

[0064] In real-world drone racing scenarios, numerous electronic and communication devices are present, such as wireless communication base stations, walkie-talkies used by event organizers, on-site wireless cameras, and mobile phones carried by spectators and staff. These devices occupy specific frequency bands and transmit signals. When spectrum analyzers collect data, the signals from these electronic and communication devices are collected along with the target drone's spectrum signal. Furthermore, when multiple drones fly simultaneously, their spectrum signals are superimposed on the collected data. These other devices not only exist as independent signal components in the collected data but also interfere with the target drone's spectrum signal. Therefore, complex and fluctuating periods may be caused by the drone constantly changing its flight state during flight, or by various external interferences leading to continuous changes in the spectrum signal.

[0065] Each component obtained after signal decomposition can characterize an oscillation mode with a specific time scale in the original signal, reflecting the signal's fluctuation characteristics at different frequencies. For the initial spectral signal of the monitoring period, some components after decomposition may be related to the UAV's own flight state. Different flight states of the UAV can cause changes in the operating states of other electronic systems, thus affecting signal quality. Simultaneously, it is also related to external environmental factors, such as electromagnetic interference from nearby electronic equipment and environmental noise. Different types of interference often have specific frequencies and variation characteristics, which are reflected in the corresponding components. Therefore, it is necessary to obtain the components for each complex and changing time period to analyze the signal interference characteristics of different components.

[0066] For each complex time-domain sub-signal, signal decomposition is performed to obtain several components, including: In one embodiment, principal component analysis is used to decompose the time-domain sub-signal of each complex period of change into several components.

[0067] The implementation process of principal component analysis is existing technology and will not be elaborated here. Furthermore, each complex period of change has its corresponding multiple components.

[0068] As an exemplary implementation, the interference frequency tendency index for each component is determined based on the frequency distribution set corresponding to the target energy set for each component. This can be achieved by... Figure 3 The steps shown are to be implemented as follows: S31, for each component, determine the maximum energy corresponding to the component, and determine the target energy set based on the maximum energy.

[0069] For components related to the flight status of drones, the center frequency of the components is relatively stable and has a specific range. However, the sources of interference signals are diverse, and the frequency range of their components may be scattered across a wide frequency band, and there may be abrupt changes, with large amplitude fluctuations.

[0070] In one embodiment, for each component within the same complex time period, the maximum energy is determined. The maximum energy at a preset multiple is taken as the minimum value of the target energy set, and the maximum energy value is taken as the maximum value of the target energy set, thereby obtaining energy data located between the two extreme values. The preset multiple can be 0.2 or 0.3, and the implementer can set the size of the preset multiple according to specific practical needs; no specific limitation is made here.

[0071] S32, obtain the frequency distribution set corresponding to the target energy set of the component, and determine the interference frequency tendency index of the component based on the similarity between the frequency distribution set and the reference frequency distribution set.

[0072] Here, the similarity is negatively correlated with the interference frequency tendency index; that is, the greater the similarity, the smaller the interference frequency tendency index.

[0073] In one embodiment, the intersection between the frequency distribution set and the reference frequency distribution set is determined, the length of the intersection is taken as the similarity between the frequency distribution set and the reference frequency distribution set, and the similarity is negatively correlated to obtain a negative correlation value, such as the reciprocal, which is used as the interference frequency tendency index of the corresponding component.

[0074] As an example, the formula for calculating the interference frequency tendency index of a component can be: In the formula, A represents the interference frequency tendency index of the component. This represents a non-zero constant, and its empirical value can be 0.01. It is used to avoid the case where the denominator of a fraction is zero. It represents the intersection between the frequency distribution set and the reference frequency distribution set, that is, the degree of similarity between the frequency distribution set and the reference frequency distribution set.

[0075] The reference frequency distribution set can be the intersection of the frequency distribution sets of all components. The higher the overlap between the frequency distribution set of a certain component and the reference frequency distribution set, the more consistent the frequency distribution of that component is with the frequency distributions of other components, the less affected it is by interference, and the less likely that the component is an interfering component.

[0076] In another embodiment, based on a frequency distribution set and a reference frequency distribution set, the similarity between the frequency presence, amplitude distribution, and energy concentration of the two sets is analyzed to determine multiple indicators of different dimensions. A comprehensive analysis of all indicators yields an interference frequency tendency index with higher numerical accuracy. The calculation processes for frequency presence, amplitude distribution, and energy concentration are existing technologies and will not be elaborated here.

[0077] By referring to the calculation process of the interference frequency tendency index of a single component, the interference frequency tendency index of each component can be obtained.

[0078] Thus, this embodiment has determined the interference frequency tendency index for each component.

[0079] S4. Obtain the oscillation intensity characteristic value of each component and its correlation coefficient with other components in the time domain sub-signal; determine the comprehensive index of interference characteristics for each component based on the interference frequency tendency index, oscillation intensity characteristic value and correlation coefficient of each component.

[0080] Here, the interference feature comprehensive index can analyze the degree of interference of a single component from multiple perspectives, which is beneficial for subsequent adaptive filtering processing.

[0081] As an exemplary implementation, obtaining the oscillation intensity characteristic value of each component and its correlation coefficient with other components in the time-domain sub-signal includes: The first step is to obtain the oscillation intensity characteristic value of each component.

[0082] Interference signals typically appear instantaneously, randomly, and unstable. Therefore, in the time-domain signal of each component, the kurtosis of each peak in the component is calculated to represent the state of each peak. The larger the kurtosis, the sharper the corresponding peak, and thus the more sudden the peak appears, and the more likely it is an instantaneous interference.

[0083] The first sub-step involves calculating the kurtosis of each peak in the time-domain sub-signal of each component.

[0084] Here, kurtosis is used at least to characterize the state of a wave peak.

[0085] The second sub-step is to determine the oscillation intensity characteristic value of the component based on the kurtosis of all the peaks corresponding to the component.

[0086] Here, kurtosis is positively correlated with the oscillation intensity eigenvalue. The larger the oscillation intensity eigenvalue, the more likely the peak in the component is caused by instantaneous interference, and the greater the probability that the corresponding component is an interference component.

[0087] In one embodiment, the average or cumulative value of the kurtosis of all peaks corresponding to a component is calculated as the oscillation intensity characteristic value of the corresponding component.

[0088] The calculation process for the kurtosis of the wave peaks in the component is existing technology and will not be elaborated here.

[0089] The second step is to obtain the correlation coefficient between each component and other components in the time-domain sub-signal.

[0090] As an exemplary implementation, a comprehensive index of interference characteristics for each component is determined based on the interference frequency tendency index, oscillation intensity characteristic value, and correlation coefficient of each component, including: The first step is to determine the interference characteristic factor of each component based on its correlation coefficients.

[0091] Because each component of the signal is related to the drone's motion and system characteristics, there may be some correlation between the components reflecting the drone's flight state under normal circumstances. However, due to their randomness and independence, interference signals typically have low correlation with other normal components. Therefore, the lower the correlation between a component and other components, the more likely the corresponding component is to be an interference component.

[0092] The first sub-step is to determine the absolute value of each correlation coefficient of the component, and then calculate the average of the absolute values ​​of all correlation coefficients.

[0093] In one embodiment, the correlation coefficient can be the Pearson correlation coefficient, which ranges from -1 to 1. The calculation process for the Pearson correlation coefficient is existing technology and will not be described in detail here.

[0094] The second sub-step involves performing negative correlation processing on the average value, and using the resulting negative correlation value as the interference characteristic factor of the component.

[0095] In one embodiment, negative correlation processing can be implemented using the inverted method or exp(-).

[0096] In another embodiment, the mean DTW distance between a single component and other components in the same signal is calculated as an interference characteristic factor of the component.

[0097] The mean DTW distance is used to represent the correlation coefficient. The larger the mean DTW distance, the weaker the correlation, indicating that the component is less similar to other components, and the greater the likelihood that the corresponding component is an interference component. The calculation process of the DTW distance is existing technology and will not be described in detail here.

[0098] The second step is to calculate the product of the interference frequency tendency index, oscillation intensity eigenvalue, and interference characteristic factor of the component, and use the product as the comprehensive index of interference characteristics of the component.

[0099] For each component, the larger the interference frequency tendency index of the component, the greater the amplitude fluctuation of the component, and the lower the correlation with other components, the more likely the corresponding component is to be an interference component, and the larger the corresponding interference characteristic comprehensive index will be.

[0100] As an example, the formula for calculating the comprehensive index of interference characteristics of components can be: In the formula, B represents the comprehensive index of interference characteristics of the component, A represents the interference frequency tendency index of the component, j represents the oscillation intensity characteristic value of the component, and q represents the interference characteristic factor of the component.

[0101] By referring to the calculation process of the comprehensive index of interference characteristics of the above components, the comprehensive index of interference characteristics of each component can be obtained.

[0102] Thus, this embodiment has determined the comprehensive index of interference characteristics for each component.

[0103] S5. Based on the comprehensive index of interference characteristics of each component, the spectral sub-signal corresponding to the time-domain sub-signal is reconstructed to obtain the target spectral signal.

[0104] Here, the target spectrum signal refers to the initial spectrum signal after adaptive filtering.

[0105] As an exemplary implementation, step S5 described above can be achieved through... Figure 4 The steps shown are to be implemented as follows: S51, for each time-domain sub-signal, determines the interference intensity of the corresponding spectral sub-signal based on the comprehensive index of the interference characteristics of each component of the time-domain sub-signal.

[0106] In one embodiment, the average value of the comprehensive index of interference characteristics of all components of the time-domain sub-signal is calculated as the interference intensity of the spectral sub-signal corresponding to the time-domain sub-signal.

[0107] It should be noted that the greater the interference intensity, the more external interference the spectrum sub-signal of the corresponding time period is subjected to, and the lower the reliability of the drone's flight status presented by the corresponding spectrum sub-signal; the smaller the interference intensity, the less external interference the drone's spectrum is subjected to in the corresponding time period, and the more accurately it can reflect the drone's flight status in the corresponding time period.

[0108] S52, adjust the initial step size in the filtering algorithm according to the interference intensity of the spectrum sub-signal, and determine the adaptive step size when the spectrum sub-signal is filtered.

[0109] In one embodiment, the initial step size is set to 0.01, and the interference intensity is normalized to the range of [-0.5, 0.5]. Based on the normalized interference intensity, the initial step size is adjusted, and the adjusted step size is used as the adaptive step size when filtering the spectral sub-signal.

[0110] When the interference intensity is high, it means that the interference on the spectrum sub-signal is stronger. The step size factor should be reduced. A smaller step size factor can slow down the convergence speed of the filter, but it can improve the stability of the filter and better suppress strong interference. When the interference intensity is low, it means that the interference on the spectrum sub-signal is weaker. The step size factor can be appropriately increased to speed up the convergence speed of the filter and enable the filter to adapt to changes in the signal more quickly.

[0111] As an example, the formula for calculating the adaptive step size when filtering a spectral sub-signal can be: ; Let C represent the initial step size, and C represent the normalized interference intensity of the spectral sub-signal. This represents the adaptive step size when filtering the spectral sub-signal.

[0112] S53 uses an adaptive step-size filtering algorithm to filter the spectrum sub-signal and obtain a new spectrum sub-signal.

[0113] In one embodiment, the implementation process of the filtering algorithm is existing technology and will not be described in detail here.

[0114] It should be noted that by adjusting the step size factor, the adaptive filter can adaptively remove external interference according to the interference intensity, highlight the spectral signal characteristics related to the UAV's flight status, and enable the detected spectral signal to better reflect the UAV's flight status.

[0115] S54, replace the corresponding complex change period spectrum sub-signals in the initial spectrum signal with each new spectrum sub-signal to obtain the target spectrum signal.

[0116] Thus, this embodiment has obtained the target spectrum signal for the monitoring period.

[0117] S6 detects the flight status of the UAV during the monitoring period based on the target spectrum signal.

[0118] In one embodiment, the target spectrum signal is converted into a two-dimensional format suitable for input to a convolutional neural network, and the two-dimensional target spectrum signal is input into a trained convolutional neural network model. The model outputs the predicted flight state, thereby obtaining more accurate UAV flight state information.

[0119] The convolutional neural network model used to predict flight status can be trained using historical flight spectrum data after removing interference and corresponding flight statuses to ensure that the network can learn the characteristic patterns of spectral signals under different flight statuses. The training process of the convolutional neural network model is existing technology and will not be described in detail here.

[0120] Thus, this embodiment completes the detection of UAV flight status with high accuracy.

[0121] In summary, this invention provides a method for detecting the flight status of unmanned aerial vehicles (UAVs) based on spectrum detection technology. This method adjusts the detected UAV spectrum signal through an adaptive filtering algorithm, which can filter out some external interference during the spectrum detection process, so that the spectrum features related to the UAV flight status can be clearly presented, highlighting the key features of the UAV flight status, which helps to more accurately identify UAV flight parameters and improve the accuracy of UAV flight status detection.

[0122] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A method for detecting the flight status of a UAV based on spectrum detection technology, characterized in that, Includes the following steps: Acquire the initial spectrum signal of the UAV during the monitoring period in flight, and convert the initial spectrum signal into an initial time domain signal; The initial time-domain signal is segmented according to the flight status to filter out various complex change periods; the complex change periods consist of at least multiple interspersed acceleration moments, deceleration moments, unstable acceleration moments, and unstable deceleration moments; For each complex time-domain sub-signal, signal decomposition is performed to obtain several components; Based on the frequency distribution set corresponding to the target energy set of each component, determine the interference frequency tendency index of each component; Obtain the oscillation intensity characteristic value of each component and its correlation coefficient with other components in the time domain sub-signal; Based on the interference frequency tendency index, oscillation intensity characteristic value and correlation coefficient of each component, the comprehensive index of interference characteristics of each component is determined; Based on the comprehensive index of interference characteristics of each component, the spectral sub-signal corresponding to the time-domain sub-signal is reconstructed to obtain the target spectral signal; The process of reconstructing the spectral sub-signal corresponding to the time-domain sub-signal based on the comprehensive index of the interference characteristics of each component to obtain the target spectral signal includes: For each time-domain sub-signal, the interference intensity of the corresponding spectral sub-signal is determined based on the comprehensive index of the interference characteristics of each component of the time-domain sub-signal. The initial step size in the filtering algorithm is adjusted according to the interference intensity of the spectrum sub-signal to determine the adaptive step size when the spectrum sub-signal is filtered. The adaptive step-size filtering algorithm is used to filter the spectrum sub-signal to obtain a new spectrum sub-signal; The target spectrum signal is obtained by replacing the spectrum sub-signals corresponding to the complex changing time periods in the initial spectrum signal with each of the new spectrum sub-signals.

2. The method for detecting the flight status of a UAV based on spectrum detection technology according to claim 1, characterized in that, The segmentation of the initial time-domain signal based on flight status to filter out various complex changing time periods includes: The amplitude sequence is obtained based on the initial time-domain signal; Based on the amplitude difference between two adjacent moments in the amplitude sequence, each normal time period is determined; the normal time periods include hovering periods, acceleration periods, and deceleration periods. Within the monitoring period, the time range between any two of the normal periods is considered as the complex change period.

3. The method for detecting the flight status of a UAV based on spectrum detection technology according to claim 2, characterized in that, The step of determining each normal time period based on the amplitude difference between two adjacent moments in the amplitude sequence includes: Calculate the amplitude difference between the next moment and the previous moment in the amplitude sequence, take two moments whose absolute value of the amplitude difference is less than or equal to the amplitude threshold as hovering moments, and form a hovering period by two or more consecutive hovering moments; All moments outside the hovering period within the monitoring period are recorded as undetermined moments, and the two undetermined moments with positive amplitude differences are taken as acceleration moments, and the two undetermined moments with negative amplitude differences are taken as deceleration moments; Based on each acceleration and deceleration moment within the monitoring period, each acceleration period and deceleration period are determined.

4. The method for detecting the flight status of a UAV based on spectrum detection technology according to claim 3, characterized in that, The step of determining each acceleration period and deceleration period based on each acceleration and deceleration moment within the monitoring period includes: For each time period of a continuously distributed target time, obtain the amplitude difference sequence corresponding to the time period; the target time is an acceleration time or a deceleration time. Based on the differences between every two adjacent elements in the amplitude difference sequence, determine the stability attribute index for each target time. The target time when the stability attribute index is greater than the stability threshold is determined as the stable target time, and two or more consecutive stable target times are combined to form a target time period; the target time period is an acceleration period or a deceleration period.

5. The method for detecting the flight status of a UAV based on spectrum detection technology according to claim 1, characterized in that, The step of determining the interference frequency tendency index for each component based on the frequency distribution set corresponding to the target energy set for each component includes: For each component, determine the maximum energy corresponding to the component, and determine the target energy set based on the maximum energy; Obtain the frequency distribution set corresponding to the target energy set of the component, and determine the interference frequency tendency index of the component based on the similarity between the frequency distribution set and the reference frequency distribution set; the similarity and the interference frequency tendency index are negatively correlated.

6. The method for detecting the flight status of a UAV based on spectrum detection technology according to claim 1, characterized in that, The process of obtaining the oscillation intensity characteristic value of each component includes: For each component, the kurtosis of each peak is calculated in the time-domain sub-signal of the component; the kurtosis is used at least to characterize the state of the peak. The oscillation intensity characteristic value of the component is determined based on the kurtosis of all the peaks corresponding to the component; the kurtosis is positively correlated with the oscillation intensity characteristic value.

7. The method for detecting the flight status of a UAV based on spectrum detection technology according to claim 1, characterized in that, The determination of the comprehensive interference characteristic index for each component based on the interference frequency tendency index, oscillation intensity characteristic value, and correlation coefficient of each component includes: For each component, the interference characteristic factor of the component is determined based on the correlation coefficient of the component. The product of the interference frequency tendency index, the oscillation intensity characteristic value, and the interference characteristic factor of the component is calculated, and the product is used as the comprehensive interference characteristic index of the component.

8. The method for detecting the flight status of a UAV based on spectrum detection technology according to claim 7, characterized in that, The step of determining the interference characteristic factor of a component based on its correlation coefficients includes: Determine the absolute value of each correlation coefficient for the component, and then calculate the average of the absolute values ​​of all correlation coefficients; The average value is subjected to negative correlation processing, and the resulting negative correlation value is used as the interference characteristic factor of the component.

9. The method for detecting the flight status of a UAV based on spectrum detection technology according to claim 1, characterized in that, After obtaining the target spectrum signal, the process also includes: Based on the target spectrum signal, the flight status of the UAV during the monitoring period is detected.

Images