Unmanned aerial vehicle thunderstorm monitoring device based on electric field electromagnetic detection and flight path planning

Abstract

The application discloses a thunderstorm monitoring device and path planning method of unmanned aerial vehicle based on electric field and magnetic field detection. The device comprises a quadrature magnetic antenna array, a vertical differential electric field sensor, a signal processing module, a positioning algorithm module, a distance estimation module, an autonomous decision module, a GNSS module and an inertial measurement module. The quadrature magnetic antenna array is composed of two orthogonally arranged magnetic antennas, which are at a 45° angle with the flight direction of the unmanned aerial vehicle, and is used for receiving the horizontal magnetic field signal of lightning radiation. The vertical differential electric field sensor comprises two groups of upper and lower electrode plates and measures the differential value of the vertical electric field intensity. Through the cooperative measurement of the magnetic antenna array and the electric field sensor, the lightning azimuth (θ) is calculated by combining the magnetic field intensity ratio, and the quadrant is corrected by using the electric field polarity. Based on the magnetic field amplitude attenuation model and the waveform rise time correction factor, a distance estimation model is constructed. The autonomous decision module generates an approach path or an obstacle avoidance path according to the lightning azimuth and distance, and drives the flight control system to adjust the heading.

Description

[0001] The present application relates to the technical field of atmospheric science detection, and particularly relates to a UAV thunderstorm position monitoring device based on electric field and magnetic field detection and a flight path planning method, which realizes real-time positioning of a thunderstorm position by a UAV, distance estimation, and autonomous planning of a flight path for atmospheric data collection. BACKGROUND

[0002] A thunderstorm is a severe weather phenomenon, often accompanied by lightning, thunder, short-term heavy rain, strong wind, and even hail, etc. Through thunderstorm monitoring, the occurrence and development trend of a thunderstorm can be discovered in advance, and timely warnings can be issued to the public to remind people to take protective measures, such as avoiding activities in open areas, staying away from tall trees and metal objects, etc., thereby effectively reducing the threat to human life caused by thunderstorm-related events such as lightning strikes. Thunderstorm monitoring is of great significance to aviation, navigation, railway and highway transportation. Most of the existing thunderstorm monitoring technologies are based on atmospheric electric field monitoring, ground-based lightning positioning, radar monitoring, satellite remote sensing monitoring, and manual observation.

[0003] With the continuous expansion of the application field of UAVs, a new approach is opened up for thunderstorm detection by providing an air-based carrier to realize air-based thunderstorm monitoring following the movement of a thunderstorm cloud. In a complex meteorological environment, a thunderstorm cloud moves constantly, and it is difficult to rely on manual control for air-based thunderstorm monitoring. In the field of UAV logistics and power line inspection, it is inevitable to encounter a thunderstorm cloud according to the planned path, and lightning activities pose a serious threat to UAVs, and the problem of UAV flight safety is increasingly prominent. Some existing technologies equip UAVs with simple lightning detectors, but the detection accuracy is limited, and it is difficult to accurately determine the direction and distance of lightning, which cannot provide effective path planning decision support for UAVs. In addition, traditional detection methods are greatly disturbed by environmental noise, and lack effective means to suppress the electromagnetic interference of UAVs, resulting in poor signal reliability. In terms of positioning accuracy, traditional methods do not fully consider the changes in the propagation characteristics of electromagnetic waves under different atmospheric conditions and the influence of lightning waveform characteristics on distance estimation, resulting in a large error in the positioning result, which cannot meet the needs of fine path planning for UAVs. At the same time, existing technologies fail to deeply integrate lightning detection with UAV flight control, making it difficult to realize real-time dynamic planning and autonomous decision-making of the path. SUMMARY

[0004] The application aims to provide a UAV thunderstorm position monitoring device and path planning method based on electric field and magnetic field detection. The monitoring device comprises a quadrature magnetic antenna array, a vertical differential electric field sensor, a signal processing module, a positioning algorithm module, a distance estimation unit, an autonomous decision module, a GNSS module and an inertial measurement module. The quadrature magnetic antenna array is composed of two quadrature arranged magnetic antennas, which are at a 45° angle with the flight direction of the UAV, and is used to receive horizontal magnetic field signals radiated by lightning. The vertical differential electric field sensor comprises two groups of upper and lower electrode plates, which measure the differential value of the vertical electric field intensity. The signal processing module performs band-pass filtering, logarithmic amplification and ratio calculation on the magnetic field signals, and outputs the magnetic field intensity ratio of two quadrature magnetic field components. The electric field signal is differentially amplified, the differential mode signal is extracted, and common mode interference is suppressed. The positioning algorithm module calculates the lightning incidence azimuth according to the magnetic field intensity ratio, and judges whether the lightning is located in front of or behind the heading of the UAV in combination with the electric field polarity.

[0005] The application is realized by the following technical scheme: the UAV thunderstorm position monitoring device based on electric field and magnetic field detection comprises a quadrature magnetic antenna array 1, a vertical differential electric field sensor 2, a signal processing module 3, a positioning algorithm module 4, a distance estimation module 5, an autonomous decision module 6, a GNSS module 7 and an inertial measurement module 8.

[0006] The quadrature magnetic antenna array 1 is composed of two quadrature arranged magnetic antennas 1a and 1b, which are at a 45° angle with the flight direction of the UAV, and is used to receive horizontal magnetic field signals radiated by lightning. The vertical differential electric field sensor 2 comprises two groups of upper and lower electrode plates 2a and 2b, which measure the differential value of the vertical electric field intensity. The signal processing module 3 performs band-pass filtering, logarithmic amplification and ratio calculation on the magnetic field signals, and outputs the magnetic field intensity ratio (H1 / H2) of two quadrature magnetic field components. The electric field signal is differentially amplified, and signal enhancement is realized by extracting the differential mode signal between the two electrodes and suppressing common mode interference.

[0007] The positioning algorithm module 4 calculates the lightning incidence azimuth according to the magnetic field intensity ratio (H1 / H2). θ The arctangent function is calculated as follows: , If the lightning is a negative ground flash (current direction downward), the electric field signal E is negative polarity, and the θ angle needs to be corrected: π = θ + θ + π , in combination with the electric field polarity to judge whether the lightning is located in front of or behind the heading of the UAV.

[0008] The distance estimation module 5 estimates the distance of lightning based on the magnetic field amplitude decay model and the waveform polarity correlation D .

[0009] The autonomous decision module 6 issues a close-range warning according to the lightning direction (±5°) and distance (±500m), generates a task path plan, and drives the flight control system to adjust the heading.

[0010] The monitoring device realizes single-station solution of lightning azimuth (θ) θ ) and distance (R) D ) through the cooperative measurement of the orthogonal magnetic antenna array and the vertical differential electric field sensor. The magnetic antenna is arranged at an angle of 45° to improve the directional resolution, and the electric field polarity is combined to determine the lightning azimuth quadrant. Based on the magnetic field amplitude decay model, a distance estimation model is constructed, and the waveform rise time correction is introduced to improve the close-range detection accuracy, forming a small unmanned aerial vehicle thunderstorm monitoring device.

[0011] Further, the magnetic field signal is band-pass filtered, and a Butterworth band-pass filter (1kHz-1MHz) is used, and its transfer function is: , Where s is a complex frequency variable; ω 0 is the center angular frequency; Q is the quality factor, which measures the frequency selectivity of the filter, Q The higher the passband is, the narrower the passband is, and the steeper the transition band is, Q The lower the passband is, the wider the passband is, K is the gain coefficient.

[0012] Further, the magnetic field strength ratio calculation is temperature drift compensated: , Where, V H1 、 V H2 is the voltage value output by the logarithmic amplification circuit after the magnetic antenna 1a, 1b senses the magnetic field, which is proportional to the magnetic field strength, K cal1 ( T )、 K cal2 ( T ) is the temperature calibration coefficient, which compensates for the gain drift of the magnetic antenna and the amplifier caused by temperature change (T) T is the temperature) to ensure the accuracy of the ratio H 1 / H 2.

[0013] Further, the distance estimation module 5 uses the lightning waveform rise timet r <10μs as a distance correction factor, the corrected distance is D D' When t r <2μs triggers a close-range alarm.

[0014] The absolute position information provided by the GNSS module 7 and the relative motion information provided by the inertial measurement module 8 are combined to calculate the heading angle of the UAV through a data fusion algorithm (such as Kalman filtering) β .

[0015] A path planning method for a UAV thunderstorm monitoring device based on electric field and magnetic field detection, comprising the following steps: S1. Electromagnetic signal acquisition and preprocessing: Real-time acquisition of horizontal magnetic field component of lightning through orthogonal magnetic antenna array (45° angle with the flight direction of the UAV) H 1、 H 2, synchronously acquire vertical electric field intensity difference value through vertical differential electric field sensor ΔE z ; Band-pass filter (1kHz~1MHz) and logarithmic amplification processing are performed on the magnetic field signal, and the magnetic field intensity ratio R H =H1 / H2; The signal processing module 3 automatically adjusts the filtering parameters according to the environmental noise characteristics through an adaptive noise suppression circuit, uses wavelet transform to eliminate the 20~50kHz frequency band noise generated by the UAV motor, and improves the signal-to-noise ratio.

[0016] S2. Attitude solution and position update: GNSS data acquisition obtains the latitude, longitude, altitude, speed, timestamp and other information of the UAV, and IMU inertial measurement data acquisition obtains the acceleration, angular velocity, magnetic field intensity and other information of the UAV. The GNSS data is preliminarily processed to remove outliers and smooth the filter data preprocessing; the IMU data is preliminarily processed, and the accelerometer and gyroscope data are filtered and denoised, and the temperature compensation data preprocessing. According to the IMU data, the attitude information (pitch angle, roll angle, yaw angle) of the UAV is calculated by using Kalman filter data fusion algorithm; according to the attitude information of the GNSS data and the IMU data, the position information of the UAV is updated, and the positioning result of the UAV is output, including latitude, longitude, height, speed, attitude and other information.

[0017] S3. Lightning azimuth solution: according to the magnetic field intensity ratio R H calculate the lightning incidence azimuth angle θ ​=arctan( R H If the lightning is a negative ground lightning, the electric field signal... E It is negative polarity and needs to be... θ conduct π Angle correction, θ '= θ + π Based on the vertical electric field polarity, it was determined that the lightning was located in front of the drone's flight path (Δ). E z >0) or behind (Δ) E z <0); GNSS module 7 provides the UAV's geodetic coordinates and heading angle β, with an accuracy ≤0.1°. The UAV's heading angle β is set to 0° at true north and increases clockwise. Correction The lightning azimuth angle θ' is taken as the origin of the unmanned aerial vehicle, The same coordinate system as β, Combined with the drone's heading angle β With the corrected lightning azimuth θ' Calculate the angle between the lightning and its heading. γ , γ =∣ β-θ' |, | β-θ' |≤90∘andΔ E z >0→Lightning is directly ahead (high-risk area); | β-θ' |>90∘andΔ E z >0→Lightning is to the side / front (medium risk zone), Δ E z <0→ Lightning is behind (low-risk area).

[0018] S4. A distance model is constructed based on the electric and magnetic field amplitude attenuation model. The lightning distance D is estimated using the amplitude ratio method based on the measurements from the orthogonal magnetic antenna array 1 and the vertical differential electric field sensor 2, combined with the waveform rise time. t r Corrected distance D Error, corrected distance is D' ,when t r Distance correction is performed when the time is less than 10 μs. t r <2 μ A near-field warning is triggered at time s.

[0019] S5. Dynamic Programming of Flight Path: Location of lightning θ ±5° and distance D' Input reinforcement learning model at ±500m to generate ellipsoidal thunderstorm region envelope centered on lightning region; Based on the GNSS satellite navigation system, this paper integrates the global path A* algorithm generation with local path requirements and uses the dynamic window method (DWA) to calculate the next velocity v of the UAV. next And the three-dimensional position offsets Δx, Δy, Δz, satisfying the dynamic constraints (maximum acceleration ≤ 5m / s², pitch angle ≤ 30°). Based on the real-time linkage mechanism between rise time correction distance and UAV flight path, when D' When the distance is ≤1km, the emergency flight mode is triggered, and a flight path is generated based on the C² continuous B-spline curve, with an avoidance radius ≥300m, and the spiral path approaches the center of the thunderstorm.

[0020] S6. Flight Control and Feedback Optimization: Adjusting the drone's attitude angle via a PID controller ( α , φ ) and thrust, execute the planned path; The system monitors the flight in real time. If the lightning location is updated or the flight path is blocked, it returns to step S1 to recalculate and updates the weights of the reinforcement learning model. The autonomous decision-making module 6 integrates reinforcement learning algorithms to optimize path planning based on historical lightning distribution data.

[0021] Furthermore, in step S5, the reinforcement learning model employs a deep Q-network (DQN), with the input being the modified lightning azimuth angle. θ' ,distance D' The current speed of the drone, v current The remaining battery power is output as a confidence score for the obstacle avoidance path (0.8~1.0 indicates a safe path); the major axis direction of the ellipsoid of the thunderstorm region envelope is related to the modified lightning azimuth angle. θ' Consistent, major axis length L =2 D' minor axis length W =0.5 L .

[0022] Furthermore, in step S6, the parameters of the PID controller are dynamically adjusted according to the mass m and moment of inertia I of the UAV, satisfying a response time ≤ 200ms and an overshoot ≤ 10%.

[0023] The path planning method for UAV thunderstorm monitoring devices based on electric and magnetic field detection also includes: a pre-set lightning activity heat map database, which automatically switches to active detection mode when the UAV enters a historical high lightning frequency area, with a spiral radius R = (0.2~1.0). D' Atmospheric data are collected by approaching the lightning center using a spiral path with a climb rate of ≥5m / s.

[0024] The beneficial effects of the present application are: the orthogonal magnetic antenna array is used for cooperative measurement with the vertical differential electric field sensor, the magnetic antenna is arranged at an angle of 45°, the direction resolution is improved, the lightning direction quadrant is determined combined with the electric field polarity, the lightning direction positioning accuracy is improved, more accurate direction information is provided for subsequent path planning decision. The distance estimation unit constructs a distance estimation model based on a magnetic field amplitude attenuation model, and introduces a waveform rise time correction factor, which effectively improves the near distance detection accuracy, enables the unmanned aerial vehicle to more accurately judge the lightning distance, and timely takes measures to avoid obstacles, thereby enhancing the flight safety. The signal processing module uses a Butterworth band-pass filter to filter the magnetic field signal, and uses a temperature drift compensation technology to ensure the accuracy of the magnetic field strength ratio calculation and improve the reliability of the system under different environmental temperatures. The autonomous decision module generates a path planning combined with the lightning direction and distance information, dynamically plans the path using a reinforcement learning model, integrates global path planning and local obstacle avoidance requirements, considers the unmanned aerial vehicle dynamics constraints, generates a safe and feasible path planning, and meets the real-time landing or obstacle avoidance requirements of the unmanned aerial vehicle. The flight control and feedback optimization mechanism accurately adjusts the unmanned aerial vehicle attitude and thrust through a PID controller, executes the planned path, and monitors the flight trajectory in real time, updates and optimizes the path strategy according to the lightning position and environmental changes, and improves the flight safety of the unmanned aerial vehicle in complex meteorological environments. The pre-set lightning activity thermal map database enables the unmanned aerial vehicle to automatically switch to the active detection mode when entering a historical high lightning frequency area, and to approach the thunderstorm center along a specific spiral path to collect atmospheric data, which not only enriches the lightning detection data, but also provides more detailed information for meteorological research, and expands the application function of the unmanned aerial vehicle in the meteorological field. BRIEF DESCRIPTION OF DRAWINGS

[0025] Figure 1 System block diagram of the present application; Figure 2 Unmanned aerial vehicle top view; Figure 3 Unmanned aerial vehicle side view; Figure 4 Unmanned aerial vehicle heading and lightning incidence angle relationship plan view; Figure 5 Unmanned aerial vehicle positioning algorithm flowchart; Figure 6 Path planning method block diagram.

[0026] In the figure: 1-orthogonal magnetic antenna array, 1a-left magnetic antenna, 1b-right magnetic antenna, 2-vertical differential electric field sensor, 2a-upper electrode plate, 2b-lower electrode plate, 3-signal processing module, 4-positioning algorithm module, 5-distance estimation module, 6-autonomous decision module, 7-GNSS module, 8-inertial measurement module. DETAILED DESCRIPTION

[0027] In order for those skilled in the art to better understand the present application, the present application will be described in conjunction withFigures 1-6 Further description of the present application, in the description, the terms "upper", "lower", "left", "right", "inner", "outer" and the like indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, or the orientation or positional relationship commonly understood by those skilled in the art, which is only for the convenience of describing the present application and simplifying the description, and does not indicate or imply that the indicated member or component must have a particular orientation, be constructed and operated in a particular orientation, and the content mentioned in the embodiments is not a limitation of the present application.

[0028] The unmanned aerial vehicle thunderstorm monitoring device based on electric field and magnetic field detection of the present application comprises a quadrature magnetic antenna array 1, a vertical differential electric field sensor 2, a signal processing module 3, a positioning algorithm module 4, a distance estimation module 5, an autonomous decision module 6, a GNSS module 7, an inertial measurement module 8, and the system block diagram is shown in Figure 1 The quadrature magnetic antenna array 1 is composed of two orthogonally arranged magnetic antennas 1a and 1b, which are installed on both sides of the central axis of the unmanned aerial vehicle and form a 45° angle with the flight direction of the unmanned aerial vehicle, as shown in Figure 2 The top view of the unmanned aerial vehicle, the unmanned aerial vehicle in the figure is a fixed-wing unmanned aerial vehicle, only for illustration, and a rotary-wing unmanned aerial vehicle can also be used, the arrow in the center of the unmanned aerial vehicle points forward, which is the heading, and the magnetic antennas 1a and 1b are arranged as shown in Figure 2 The enlarged view of the upper left corner of the figure. The quadrature magnetic antenna array 1 is used to receive the horizontal magnetic field signal of lightning radiation, and its working frequency range covers the typical frequency band of lightning radiation, and it is connected to the signal processing module 3 through a coaxial cable. The quadrature magnetic antennas 1a and 1b use nanocrystalline alloy magnetic cores, and the magnetic permeability can reach 10 4 The high magnetic permeability can make the antenna more effectively convert the magnetic field energy into an electric signal when receiving a weak magnetic field signal, improve the sensitivity of the antenna, and help detect more distant or weaker signal sources.

[0029] The vertical differential electric field sensor 2 comprises upper and lower electrode plates 2a and 2b, the upper electrode plate 2a is installed on the upper part of the unmanned aerial vehicle, and the lower electrode plate 2b is installed on the lower part of the unmanned aerial vehicle, the electrode plates are fixed by insulating supports, as shown in Figure 3 The side view of the unmanned aerial vehicle. The electrode plates 2a and 2b are metal plates with a diameter of 100-200 mm to improve the electric field sensing sensitivity, and the output signal is transmitted to the signal processing module 3 through a differential cable. The electrode plates 2a and 2b measure the differential value of the vertical electric field intensity.

[0030] The signal processing module 3 is installed in the cabin of the unmanned aerial vehicle and comprises a band-pass filter circuit, an adaptive noise suppression circuit, a logarithmic amplification circuit, a differential amplification circuit and an analog-to-digital converter (ADC). The signal processing module 3 performs band-pass filtering and logarithmic amplification processing on the magnetic field signal, calculates the magnetic field intensity ratio, and outputs two orthogonal magnetic field components magnetic field intensity ratio R H=H1 / H2. The magnetic field signal is filtered using a Butterworth bandpass filter (1kHz~1MHz), and its transfer function is: , Where s is a complex frequency variable; ω 0 is the center angular frequency; Q The quality factor measures the frequency selectivity of a filter. Q The higher the passband, the narrower it becomes, and the steeper the transition zone. Q The lower the passband, the wider and gentler it is; K This is the gain coefficient; Temperature changes directly affect the performance of magnetic materials, the output stability of sensors, and the overall reliability of the measurement system through various physical mechanisms. Therefore, temperature drift compensation is needed for the calculation of the magnetic field strength ratio. The compensation formula is as follows: , in, V H1 , V H2 The voltage value output by the logarithmic amplifier circuit after the magnetic fields induced by the magnetic antennas 1a and 1b are proportional to the magnetic field strength. K cal1 ( T ), K cal2 ( T ) is the temperature calibration coefficient, used to compensate for temperature changes in the magnetic antenna and amplifier. T To prevent gain drift caused by temperature, ensure the ratio H 1 / H Accuracy of 2. Temperature calibration coefficient. K cal1 ( T ), K cal2 ( T (), through dynamic temperature linear regression or quadratic curve fitting, R-T Relationship, extract slope K cal Temperature drift compensation is not only necessary for correcting the inherent properties of materials, but also crucial for ensuring the reliability of system-level measurements.

[0031] The signal processing module 3 differentially amplifies the electric field signal, and signal enhancement is achieved by extracting the differential mode signal between the two electrodes and suppressing common mode interference. The electric field signal source (such as an electrode) usually presents a high output impedance (> 1 MΩ), which is easily affected by distributed capacitance and electromagnetic interference, and a high input impedance front end is required to avoid signal attenuation. The electric field interference in the environment is mostly in-phase noise, and differential amplification extracts the effective differential mode component by canceling the common mode signal. Differential amplification can use a two-way operational amplifier to achieve "amplify the difference and suppress the common mode" through symmetric design and high CMRR devices, while considering bandwidth, noise and stability.

[0032] The signal processed by the signal processing module 3 is transmitted to the positioning algorithm module 4. The positioning algorithm module 4 is based on an embedded processor (such as ARMCortex-A series), which stores and runs the lightning positioning algorithm. The positioning algorithm module 4 calculates the lightning incidence azimuth θ according to the magnetic field strength ratio (H1 / H2) through the arctangent function , The azimuth θ represents the horizontal direction angle of the lightning relative to the north direction. The process of discharging from the negative charge in the cloud to the ground (the lower part of the cloud is usually negatively charged) has a downward current direction, resulting in a negative polarity of the electric field signal (i.e. the electric field intensity instantaneous value is negative), which is called negative ground flash. If the lightning is a negative ground flash (the current direction is downward), the electric field signal E is negative in polarity, and the negative polarity signal will cause the actual direction to deviate by 180°. The θ needs to be π angle corrected: the corrected lightning incidence azimuth θ = θ + π , and the corrected azimuth θ' restores the true lightning position. In combination with the vertical electric field polarity, it is determined whether the lightning is located in front of or behind the heading of the unmanned aerial vehicle. At the same time, the unmanned aerial vehicle geodetic coordinates and heading angle β provided by the GNSS module 7 have an accuracy of ≤0.1°, and the unmanned aerial vehicle heading angle β increases clockwise with the north as 0°. The latitude and longitude information provided by the GNSS module 7 can be used to determine the absolute position and general direction of motion of the unmanned aerial vehicle, and the inertial measurement module 8 can provide acceleration and angular velocity information of the unmanned aerial vehicle, reflecting the short-term motion state of the unmanned aerial vehicle. The absolute position information provided by the GNSS module 7 and the relative motion information provided by the inertial measurement module 8 are combined to calculate the heading angle β of the unmanned aerial vehicle through a data fusion algorithm (such as Kalman filtering). Through the fusion of the two kinds of information, the heading angle of the unmanned aerial vehicle can be updated in real time, overcoming the limitations of a single sensor. The lightning azimuth θ' is corrected with the unmanned aerial vehicle as the origin and the same coordinate system as β , in combination with the heading angle of the unmanned aerial vehicleβ with the modified lightning azimuth angle θ' , the included angle of the relative course of lightning is calculated γ , γ =∣ β - θ' ∣, and the danger level of the area where the lightning is located is judged accordingly. The relationship between the heading of the unmanned aerial vehicle and the incidence angle of the lightning is shown in Figure 4 , where N is north, E is east, and P is the position of the lightning.

[0033] The distance estimation module 5 is integrated in an embedded processor, and based on the amplitude decay model of electric and magnetic fields and the correlation of wave polarity, the distance of lightning is estimated by using the amplitude ratio method based on the measurement values of the orthogonal magnetic antenna array 1 and the vertical differential electric field sensor 2 D . In the far field area, the ratio of the electric field intensity to the magnetic field intensity is a constant, i.e. the free space wave impedance (~377Ω), but in the near field area, the ratio is no longer a constant but a function related to the distance D . The present application estimates the distance of lightning by using this principle D . The specific method is as follows: first, the correspondence between the measurement value of the orthogonal magnetic antenna array 1 and the magnetic field intensity H is calibrated through experiments, and the correspondence between the measurement value of the vertical differential electric field sensor 2 and the electric field intensity E is calibrated through experiments; the functional relationship between the ratio R and the distance D is established through experiments: , wherein k is a constant related to the energy of the lightning itself and the local environment, p is an index (usually close to 1), and the typical values of k and p can be obtained through a large amount of data statistical analysis, and the distance estimation formula is: , This method offsets the uncertainty of the source intensity of lightning, and the ratio of E and H to the distance is relatively stable whether it is a weak lightning or a strong lightning, which is equivalent to normalizing the “energy” of the lightning. Although the accuracy of the method for estimating the distance of lightning is not high, it is only used for flight path planning of the unmanned aerial vehicle, and does not need to have a high accuracy like lightning positioning.

[0034] The rise time (τ) of the lightning waveform t r is a key parameter for describing the time required for the electric field or magnetic field intensity to rise from 10% to 90% of the pulse amplitude, which reflects the steepness of the lightning discharge process and the speed of energy release.t r There is a significant positive correlation between rise time and the distance between the observation point and the lightning. This phenomenon is mainly caused by the dispersion effect and medium attenuation during electromagnetic wave propagation. In UAV lightning location systems, rise time is used... t r The correction for near-range calculations is primarily based on the dual requirements of electromagnetic wave propagation characteristics and sensor physical limitations. Lightning-generated electromagnetic waves contain abundant high-frequency components (MHz to GHz levels), and these high-frequency components attenuate much more than low-frequency components when propagating through the atmosphere, resulting in a longer rise time (...). t r It directly reflects the intensity of the high-frequency components of the signal. t r The shorter the distance, the more high-frequency components the signal contains, and the more severe the attenuation occurs at close range. When lightning is too close (e.g., D At depths <500m, strong electromagnetic fields can cause magnetic field sensors to saturate, compressing the magnetic field peak value and leading to... H peak The distance calculation fails, but the linear response is still maintained even when the rising edge slope saturates. The magnetic field sensor can still maintain its linear response in the saturation region. dt r / dH The linear response becomes a more reliable physical quantity. Using rise time (… t r The method corrects the near-range calculation by using a coupling model of the high-frequency attenuation characteristics of electromagnetic waves and the physical response of the sensor to solve the problem of distance calculation distortion caused by near-range magnetic field saturation.

[0035] Distance estimation module 5 constructs a distance model based on the electric and magnetic field amplitude attenuation model, using the lightning waveform rise time. t r <10μs was used as a distance correction factor, combined with the waveform rise time. t r Distance D Error correction, the corrected distance is D' Data acquisition and modeling are conducted at different known distances within the target application environment (or a series of representative environments). D i Above, a large number of data samples were collected, and for each sample, three key features were collected simultaneously: E peak The peak value of the electric field intensity H peak The peak value of the magnetic field strength, t r Waveform rise time. A hybrid model is constructed based on physics to establish the rise time. t r With distance DA single model is used to analyze the relationship between the amplitude ratio (such as E / H ) and the model residual (the D actual - D tr ) by taking the amplitude information as a correction term, and a correction function is established: , The corrected distance formula is: , When t r <10us, the distance is corrected, t r <2 μ s, the near distance early warning is triggered.

[0036] The autonomous decision module 6 integrates a reinforcement learning algorithm, optimizes the path planning according to historical lightning distribution data, issues a near distance early warning according to the lightning direction (±5°) and distance (±500m), generates a task path planning, and drives the flight control system to adjust the heading.

[0037] The monitoring device realizes single-station solution of lightning azimuth ( θ ) and distance ( D ) through the cooperative measurement of the orthogonal magnetic antenna array and the vertical differential electric field sensor. The magnetic antenna is arranged at an angle of 45° to improve the directional resolution, and the lightning azimuth quadrant is determined in combination with the electric field polarity. A distance estimation model is constructed based on the magnetic field amplitude attenuation model, the near distance detection accuracy is improved by introducing the waveform rise time correction, and a small unmanned aerial vehicle thunderstorm monitoring device is formed.

[0038] The path planning method of the unmanned aerial vehicle thunderstorm monitoring device based on electric field and magnetic field detection of the application, the path planning method block diagram is seen Figure 6 , comprising the following steps: S1. Electromagnetic signal acquisition and pretreatment: The horizontal magnetic field components of lightning are collected in real time through the orthogonal magnetic antenna array (at an angle of 45° to the flight direction of the unmanned aerial vehicle) H 1、 H 2, and the vertical electric field intensity difference value is obtained synchronously through the vertical differential electric field sensor ΔE z ; The magnetic field signal is band-pass filtered (1kHz~1MHz) and logarithmically amplified, and the magnetic field intensity ratio is calculated R H =H1 / H2; The signal processing module 3 automatically adjusts the filtering parameters according to the environmental noise characteristics through an adaptive noise suppression circuit, and uses wavelet transform to eliminate the 20-50 kHz frequency band noise generated by the unmanned aerial vehicle motor, thereby improving the signal-to-noise ratio.

[0039] S2. Attitude solution and position update: The GNSS data acquisition obtains the longitude, latitude, altitude, speed, timestamp and other information of the unmanned aerial vehicle, and the IMU inertial measurement data acquisition obtains the acceleration, angular velocity, magnetic field intensity and other information of the unmanned aerial vehicle. The GNSS and IMU data are collected in parallel, and the hardware errors are eliminated through preprocessing. The GNSS data is preliminarily processed, including removing outliers, smoothing filtering data preprocessing; the IMU data is preliminarily processed, including filtering and denoising the accelerometer and gyroscope data, and temperature compensation data preprocessing. According to the IMU data, the Kalman filter data fusion algorithm is used to calculate the attitude information (pitch angle, roll angle, yaw angle) of the unmanned aerial vehicle; according to the GNSS data and the attitude information of the IMU data, the position information of the unmanned aerial vehicle is updated. The Kalman filter is used to fuse the dynamic response of the gyroscope and the static stability of the accelerometer, solving the problem of attitude angle drift. When the GNSS signal is temporarily lost or the precision is insufficient, the IMU data and the known initial position information are used for inertial navigation calculation to calculate the current position of the unmanned aerial vehicle. The inertial navigation calculation formula is: , wherein, v k is the speed update, representing the current time speed v k is the speed of the previous time v k-1 plus the acceleration after attitude transformation a b ( t ) is the integral in the time interval[ t k -1, t k ]; p k is the position update, representing the current time position p k is the position of the previous time p k-1 plus the speed after attitude transformation v b ( t ) is the integral in the time interval[ t k-1 , t k ]. The GNSS position information and the inertial navigation calculation result are fused to obtain more accurate unmanned aerial vehicle position information, including longitude, latitude, height, speed, attitude and other information. The unmanned aerial vehicle positioning flow chart is shown in Figure 5 .

[0040] S3. Lightning azimuth solution: according to the magnetic field intensity ratio RH Calculate the incident azimuth of lightning θ =arctan( R H If the lightning is a negative ground lightning, the electric field signal... E It is negative polarity and needs to be... θ conduct π Angle correction, θ '= θ + π Based on the vertical electric field polarity, it was determined that the lightning was located in front of the drone's flight path (Δ). E z >0) or behind (Δ) E z <0); GNSS module 7 provides the UAV's geodetic coordinates and heading angle β, with an accuracy of ≤0.1°. The UAV's heading angle β is set to 0° with true north as the reference point and increases clockwise. The corrected lightning azimuth angle θ′ is set to the UAV as the origin and is in the same coordinate system as β. β With the corrected lightning azimuth θ' Calculate the angle between the lightning and its heading. γ , γ =∣ β-θ' |, | β-θ' |≤90∘andΔ E z >0→Lightning is directly ahead (high-risk area), | β-θ' |>90∘andΔ E z >0→Lightning is to the side / front (medium risk zone), Δ E z <0→ Lightning is behind (low-risk area).

[0041] S4. A distance model is constructed based on the electric and magnetic field amplitude attenuation model. The lightning distance is estimated using the amplitude ratio method based on the measurements from the orthogonal magnetic antenna array 1 and the vertical differential electric field sensor 2. D Combined with waveform rise time t r Corrected distance D Error, corrected distance is D'。 when t r Distance correction is performed when the time is less than 10 μs. t r <2 μ Trigger a close-range warning when s; when t r Distance correction is performed when the time is less than 10 μs. t r <2 μ A near-field warning is triggered at time s.

[0042] S5. Dynamic Programming of Flight Path: Lightning azimuth θ ±5° and distance D' ±500m are input into a reinforcement learning model to generate an ellipsoidal thunderstorm region envelope centered on the lightning area. The reinforcement learning model uses a deep Q network (DQN) with inputs of the corrected lightning azimuth θ' , distance D' , current speed v current of the unmanned aerial vehicle, and remaining power, and outputs a confidence score of the obstacle avoidance path (0.8-1.0 for a safe path); the ellipsoidal long axis direction of the thunderstorm region envelope is consistent with the corrected lightning azimuth θ' , the long axis length L =2 D' , and the short axis length W =0.5 L ; Based on the GNSS satellite navigation system, the global path A* algorithm is fused to generate a local path requirement, and the dynamic window method (DWA) is used to calculate the next speed v next and three-dimensional position offset Δx, Δy, Δz of the unmanned aerial vehicle, which satisfies the dynamic constraint (maximum acceleration ≤5m / s², pitch angle ≤30°); Based on the real-time linkage mechanism of the rise time correction distance and the unmanned aerial vehicle flight path, when D' ≤1km, the emergency flight mode is triggered, a fly-around path is generated based on the C² continuous B-spline curve, the avoidance radius is ≥300m, and the spiral path approaches the thunderstorm center.

[0043] S6. Flight control and feedback optimization: The PID controller is used to adjust the attitude angle (θ α , φ ) and thrust of the unmanned aerial vehicle to execute the planned path, and the PID controller parameters are dynamically adjusted according to the mass m and moment of inertia I of the unmanned aerial vehicle, with a response time ≤200ms and an overshoot ≤10%. The flight path of the unmanned aerial vehicle is monitored in real time during flight, and if the lightning position is updated or the flight path is blocked, step S1 is returned to recalculate and update the reinforcement learning model weight, and the autonomous decision module 6 integrates the reinforcement learning algorithm to optimize the path planning according to historical lightning distribution data.

[0044] The path planning method of the unmanned aerial vehicle thunderstorm monitoring device based on electric field and magnetic field detection further includes: a lightning activity thermal map database is preset, when the unmanned aerial vehicle enters a historical high lightning frequency area, it is automatically switched to an active detection mode, and a spiral path with a spiral radius R=(0.2-1.0) D' , and a climb rate ≥5m / s is used to approach the lightning center to collect atmospheric data.

[0045] In embodiment one, the lightning incidence azimuth θ and the heading angle of the unmanned aerial vehicleβ Estimate the electric field difference value Δ. E z =-1.8 kV / m， The voltage values ​​output by the logarithmic amplifier circuit after the magnetic antennas 1a and 1b induce magnetic fields are... V H1 =2.35V V H2 =1.88V, temperature T=30℃; calibration coefficient K cal1 ( T =1.02-0.003( T- 25), K cal2 ( T =0.98 + 0.002 T- 25), ( T (Unit: °C) Temperature drift compensation: , , Substitute the parameters, ; magnetic field ratio R H = H 1 / H 2 = 1.269 → θ =arctan(1.269)≈51.83°, Electric field difference Δ E z =-1.8 kV / m (Negative polarity) → Negative ground flash correction θ ′=51.83°+180°=231.83°, GNSS heading angle β =120.5° (north reference, clockwise), relative heading angle γ =∣120.5°-231.83°∣=111.33°>90°, and Δ E z <0→ Lightning is behind, low-risk zone.

[0046] This invention is merely an embodiment of an unmanned aerial vehicle (UAV) carrying an electromagnetic field detection device as an airborne carrier, autonomously planning its flight path according to mission requirements. In practical applications, it can carry atmospheric detection equipment to detect approaching thunderstorm clouds and wirelessly transmit the data back to the ground center. During application, note... D Electric field difference Δ at <300m E zNonlinear, resulting in azimuth angle solution invalid, therefore, when compiling flight path planning algorithm, should fly ≥300m away from the center of the thunderstorm, atmospheric data acquisition.

[0047] The application can also be used in the fields of unmanned aerial vehicle logistics, power line inspection and the like, and a flight path is autonomously planned to avoid thunderstorm areas by detecting thunderstorm positions.

[0048] The specification and drawings of the present application are only specific to one embodiment, rather than limiting, and those skilled in the art can make many forms without departing from the purpose of the present application and the scope protected by the claims under the inspiration of the present application, which are all within the protection scope of the present application.

Claims

1. An unmanned aerial vehicle thunderstorm monitoring device based on electric and magnetic field detection, characterized by: The application relates to a lightning detection system for unmanned aerial vehicles, which comprises a quadrature magnetic antenna array (1), a vertical differential electric field sensor (2), a signal processing module (3), a positioning algorithm module (4), a distance estimation module (5), an autonomous decision-making module (6), a GNSS module (7) and an inertial measurement module (8). The quadrature magnetic antenna array (1) is composed of two magnetic antennas (1a and 1b) arranged in quadrature and forms an angle of 45 degrees with the flight direction of the unmanned aerial vehicle, and is used for receiving horizontal magnetic field signals radiated by lightning. The vertical differential electric field sensor (2) comprises upper and lower electrode plates (2a and 2b) and is used for measuring the differential value of the vertical electric field intensity. The signal processing module (3) is used for performing band-pass filtering, logarithmic amplification and ratio calculation on the magnetic field signals and outputs the magnetic field intensity ratio (H1 / H2) of two quadrature magnetic field components; the signal processing module (3) is used for performing differential amplification on the electric field signals, and signal enhancement is realized by extracting the differential mode signal between two electrodes and suppressing common mode interference. The positioning algorithm module (4) calculates the lightning incidence azimuth angle The autonomous decision-making module (6) sends a close-range early warning according to the lightning direction (plus or minus 5 degrees) and the distance (plus or minus 500 m), generates a task path plan and drives the flight control system to adjust the heading. By the arctangent function: ， If the lightning is negative (current direction downward), the electric field signal E is negative polarity, and the The magnetic field signals are subjected to band-pass filtering, a Butterworth band-pass filter (1kHz-1MHz) is adopted, and the transfer function is as follows: π angle correction needs to be made: The magnetic field intensity ratio calculation is subjected to temperature drift compensation. The application further comprises the following steps: + π , combined with the electric field polarity to determine whether the lightning is in front of or behind the UAV heading.​​ The distance estimation module (5) estimates the lightning distance based on an electric field and magnetic field amplitude attenuation model and waveform polarity correlation D ; S1. Electromagnetic signal acquisition and pretreatment: The monitoring device realizes single-station solution of lightning azimuth ( The signal processing module (3) automatically adjusts the filtering parameters according to the environmental noise characteristics through an adaptive noise suppression circuit, adopts wavelet transform to remove the 20-50kHz frequency band noise generated by the unmanned aerial vehicle motor and improves the signal-to-noise ratio. ) and distance ( D ) through the cooperative measurement of the orthogonal magnetic antenna array and the vertical differential electric field sensor, the magnetic antenna is arranged at an angle of 45° to improve the directional resolution, and the lightning azimuth quadrant is determined in combination with the electric field polarity; a distance estimation model is constructed based on a magnetic field amplitude attenuation model, the waveform rise time correction is introduced to improve the near-distance detection accuracy, and a small unmanned aerial vehicle storm monitoring device is formed.

2. The unmanned aerial vehicle thunderstorm monitoring device based on electric and magnetic field detection according to claim 1, characterized in that: S2. Attitude solution and position update: ， where s is a complex frequency variable; GNSS data acquisition is used to acquire the longitude, latitude, altitude, speed and time stamp of the unmanned aerial vehicle, and IMU data acquisition is used to acquire the acceleration, angular velocity and magnetic field intensity of the unmanned aerial vehicle. 0 is the center angular frequency; Q Q is the quality factor, which measures the filter frequency selectivity, Q The higher the passband is narrower, the transition band is steeper, Q The lower the passband is wider, K G is the gain coefficient.

3. The unmanned aerial vehicle thunderstorm monitoring device based on electric and magnetic field detection according to claim 1, characterized in that: The GNSS data is subjected to preliminary processing, and the abnormal values are removed and the data is subjected to smoothing filtering pretreatment. ， in, V H1 , V H2 The voltage value output by the logarithmic amplifier circuit after the magnetic field induced by the magnetic antennas (1a, 1b) is proportional to the magnetic field strength. K cal1 ( T ), K cal2 ( T ) is the temperature calibration coefficient, used to compensate for temperature changes in the magnetic antenna and amplifier. T To prevent gain drift caused by temperature, ensure the ratio H 1 / H The precision is 2.

4. The unmanned aerial vehicle thunderstorm monitoring device based on electric and magnetic field detection according to claim 1, characterized in that: The distance estimation module (5) uses the lightning waveform rise time t r <10 μs as a distance correction factor to correct the distance D , and the corrected distance is D' When t r <2 μs, a close distance alarm is triggered.

5. The unmanned aerial vehicle thunderstorm monitoring device based on electric and magnetic field detection according to claim 1, characterized in that: The absolute position information provided by the GNSS module (7) and the relative motion information provided by the inertial measurement module (8) are combined through a data fusion algorithm (such as Kalman filtering) to calculate the heading angle of the UAV β .

6. A path planning method for the unmanned aerial vehicle storm monitoring device based on electric field and magnetic field detection according to claim 1, characterized in that: The IMU data is subjected to preliminary processing, and the accelerometer and gyroscope data are subjected to filtering denoising and temperature compensation data pretreatment. The attitude information (pitch angle, roll angle and yaw angle) of the unmanned aerial vehicle is calculated by adopting a Kalman filtering data fusion algorithm according to the IMU data; and the position information of the unmanned aerial vehicle is updated according to the attitude information of the GNSS data and the IMU data. Real-time acquisition of horizontal magnetic field component of lightning by orthogonal magnetic antenna array H 1、 H 2, synchronously acquire vertical electric field intensity differential value by vertical differential electric field sensor S3. Lightning direction solution: z ; The magnetic field signal is band-pass filtered (1 kHz~1 MHz) and logarithmically amplified, and the magnetic field intensity ratio is calculated R H =H1 / H2; The autonomous decision-making module (6) sends a close-range early warning according to the lightning direction (plus or minus 5 degrees) and the distance (plus or minus 500 m), generates a task path plan and drives the flight control system to adjust the heading. The PID controller parameters are dynamically adjusted according to the mass m and the moment of inertia I of the unmanned aerial vehicle in step S6, and the response time is less than or equal to 200 ms and the overshoot is less than or equal to 10%. The application further comprises the following steps: ​ ​ ​ ​ According to the ratio of the magnetic field strength R H The lightning incidence azimuth angle is calculated ​ =arctan( R H ), if the lightning is a negative ground flash, the electric field signal E is negative polarity, and the ​ angle correction needs to be made π ; Combining the vertical electric field polarity determines that the lightning is located in front of the UAV heading (Δ E z >0) or behind (Δ E z <0); the GNSS module (7) provides the UAV geodetic coordinates and heading angle β, with an accuracy of ≤0.1°, and the UAV heading angle β takes the north as 0°, increasing clockwise, the corrected lightning azimuth θ' is taken as the origin with the UAV, and is in the same coordinate system as β, combining the UAV heading angle β with the corrected lightning azimuth ​ , the included angle of the relative heading of the lightning is calculated ​ , ​ =∣ ​ ∣, ∣ ​ ∣≤90∘ and Δ E z >0→ the lightning is in front (high-risk area), ∣ ​ ∣>90∘ and Δ E z >0→ the lightning is in front (medium-risk area), Δ E z <0→ the lightning is behind (low-risk area); S4. Constructing the distance model based on the electric field and magnetic field amplitude attenuation model, estimating the lightning distance by the amplitude ratio method through the measurement values of the orthogonal magnetic antenna array (1) and the vertical difference electric field sensor (2) D , combining the waveform rise time t r Correct the distance D error, the corrected distance is D' When t r When 10us, correct the distance t r <2 ​ When 2s, trigger the near distance early warning; ​ Lightning azimuth ​ ±5° and distance D' ±500m input reinforcement learning model to generate an ellipsoidal thunderstorm region envelope centered on the lightning region; Based on GNSS satellite navigation system, fusion global path A* algorithm generation and local path demand, using dynamic window method (DWA) to calculate the next speed v of UAV next And three-dimensional position offset Δx, Δy, Δz, meet the dynamics constraint (maximum acceleration ≤5 m / s², pitch angle ≤30°); Based on the real-time linkage mechanism of the rising time correcting distance and the UAV flight path, when D' ≤1km, trigger the emergency flight mode, generate the circling flight path based on the C2 continuous B-spline curve, avoid the radius ≥300m, and spiral path into the thunderstorm center; ​ Adjusting the drone's attitude angle via a PID controller ( α , ​ ) and thrust, execute the planned path; ​ 7. The path planning method of the unmanned aerial vehicle thunderstorm monitoring device based on electric field and magnetic field detection according to claim 6, characterized in that: In step S5, the reinforcement learning model adopts a deep Q network (DQN), and the input is the modified lightning azimuth ​ , distance D' , current speed v current of the unmanned aerial vehicle, and remaining power, and the output is a confidence score of a flight path (0.8-1.0 is a safe path). The ellipsoid long axis direction of the thunderstorm area envelope and the modified lightning azimuth ​ The consistent long axis length L = 2 D' The short axis length W = 0.5 L .

8. A path planning method for the unmanned aerial vehicle thunderstorm monitoring device based on electric and magnetic field detection according to claim 6, characterized in that: ​ 9. A path planning method for the unmanned vehicle storm monitoring device based on electric and magnetic field detection according to any one of claims 6-8, characterized in that: ​ The preset lightning activity thermal map database is switched to the active detection mode automatically when the unmanned aerial vehicle enters a historical high lightning frequency area, and a spiral path with a spiral radius R=(0.2~1.0) D' , and a climbing rate greater than or equal to 5 m / s is used to approach the thunderstorm center to collect atmospheric data.

Images