A near-space lightning double-pulse detection device

Abstract

This invention relates to a near-space lightning double-pulse detection device, belonging to the field of lightning science and technology. The device uses a broadband antenna to acquire electric field signals, a GPS / BeiDou receiver module to collect location and time information, and an embedded software module to control the device, including initializing configuration parameters, setting signal acquisition parameters, setting trigger algorithm parameters, receiving and feeding back commands, storing and exporting data, etc. It employs an interference elimination algorithm to statistically analyze background spectrum energy and narrowband trigger windows in real time; and a lightning double-pulse signal identification algorithm to dynamically adjust the trigger threshold and the number of narrowband trigger windows. The device can communicate with a near-space aerostat platform, receive platform commands, and transmit status data. The near-space lightning double-pulse detection device proposed in this invention can be mounted on a stratospheric near-space balloon platform to detect lightning double-pulse signals, and features high sampling rate, light weight, low power consumption, low computational load, and online identification capability.

Description

Technical Field

[0001] This invention relates to a near-space (20km-30km altitude) lightning double-pulse detection device, belonging to the field of lightning science and technology. The device is mounted on a near-space balloon-borne platform and is used to detect atmospheric lightning double-pulse events in the stratosphere. Background Technology

[0002] In 1995, the American ALEXIS satellite first reported observations of lightning double pulses (TIPP). A lightning double pulse consists of two discrete radio signals, spaced tens of microseconds apart, with peak power reaching the 10 MW range and a spectral range from 25 MHz to 200 MHz, making it the most powerful natural radio source in Earth's environment. Due to the group velocity dispersion effect of the ionosphere, the high-frequency portion of the pulse reaches the satellite first, followed by the low-frequency portion, creating a chirped signal; low-frequency radio waves below 25 MHz are completely reflected by the ionosphere and cannot be detected.

[0003] In 1997, the United States launched the FORTE satellite to conduct extensive and in-depth measurements of lightning double pulse signals. Due to the adjustable trigger threshold of its radio detection payload, it observed nearly 500,000 lightning double pulse signals, demonstrating that this phenomenon is a common occurrence in lightning processes. The FORTE satellite primarily revealed the polarization, coherence, intensity, dual-mode nature of lightning double pulses, and their relationship with lightning radiation. Lightning double pulses are essentially linearly polarized. Dual-mode nature refers to the existence of two events: strong and weak. Strong events correspond to higher radiative power, with pulse widths on the order of microseconds and a relatively flat spectrum over a wide range; weak events have lower power, with pulses shorter than 100 ns and a steeper spectrum. The correlation between lightning double pulses and lightning radiation is not high, and the correlation is even lower for strong events. FORTE satellite data above 100 MHz also showed that the internal structures of the two pulses are completely independent.

[0004] In 2008, the Chemical Defense Research Institute of the former General Armaments Department of my country also detected a lightning double pulse signal using the electromagnetic pulse detector of the Experimental Satellite-3. However, due to the bandwidth limitation of the detector, the time-domain waveform of the signal could not be provided.

[0005] In 2012, Russia launched the Chibis-M scientific microsatellite, which targets double pulses of lightning. It measured radio signals of single events, which consist of only one pulse. It concluded that these events were caused by lightning at a distant near-horizontal horizon, where the two pulses overlapped in time and merged into one pulse.

[0006] Since the discovery of double lightning pulses, a relatively simple explanation has been proposed: a special type of cloud lightning, called a "compact cloud lightning," occurs within thunderclouds. This lightning generates strong radio waves that propagate in all directions. The radio waves traveling along the straight-line distance from the compact cloud lightning to the satellite form the first pulse of the double lightning pulse, while the signal reflected from the ground and delayed before reaching the satellite constitutes the second pulse. This explanation is known as the reflection model.

[0007] In 2015, Professor Wu Huichun of Zhejiang University proposed a new explanation, suggesting that strong radio double pulses originate from the step leader of a ground lightning strike. When the lightning channel contacts the ground, the two leaders closest to the ground generate two clusters of high-energy electrons, whose energy is sufficient to strike the ground. Subsequently, radio waves are radiated through a mechanism called "coherent transit radiation," which occurs when electrons cross the interface between the air and the ground. Due to the relativistic searchlight effect, these transit radio radiations are highly collimated, emitted in the opposite direction to the electrons, i.e., towards space; they penetrate clouds, near-space, the ionosphere, and finally reach the satellite. This theory successfully explained the vast majority of satellite observation data for the first time.

[0008] To verify the physical mechanism of lightning double pulse generation, it is necessary to detect the lightning double pulse signal without distortion. However, an inherent limitation of satellite detection is that radio signals must pass through the ionosphere, and ionospheric dispersion severely distorts the waveform of the radio signal, making it impossible to infer the characteristics of high-energy electrons in lightning. Balloon-based aerial detection perfectly avoids this limitation, measuring a true original lightning double pulse waveform, allowing for the inference of high-energy electron characteristics and further revealing the physical mechanism of lightning double pulse generation. According to the novelty search results, there is currently no equipment or device in China capable of detecting lightning double pulses from the air.

[0009] The invention patent "Regional Lightning Warning Method" (patent number 200310116065.4) discloses a lightning warning method that simultaneously monitors the static electricity and VHF radiation waves of thunderclouds. It issues a lightning alarm when both the electric field generated by thunderclouds and VHF radiation waves are detected simultaneously, which can provide relatively accurate lightning warnings and avoid false alarms when the atmospheric electric field meter and VHF receiver are monitored alone. This can effectively reduce the false alarm rate of the system and improve monitoring efficiency.

[0010] The utility model patent "A Very High Frequency Lightning Monitoring Device" (patent number 201720857009.3) discloses a lightning location device, which uses a biconical antenna, a high-frequency amplification circuit, a high-frequency circuit and an intermediate-frequency amplification circuit to form a very high frequency lightning electromagnetic pulse signal receiving device, which can improve the accuracy of lightning location coefficient.

[0011] The invention patent "A three-dimensional positioning method and system for multi-station lightning VHF radiation sources" (patent number 201310108397.1) discloses a three-dimensional positioning system for lightning VHF radiation sources, including at least three single-station sites and a data processing center, which can detect the three-dimensional spatial evolution characteristics of lightning over a wide area and effectively extend the detection distance of a single station.

[0012] The invention patent "Method and System for Monitoring the Entire Process of Lightning by Integrating VHF and Thunder Sound Detection" (Patent No. 201810772012.4) discloses a method for detecting and locating lightning by integrating VHF and thunder sound detection. Through VHF and thunder sound three-dimensional location algorithms and data clustering fusion processing methods, it can provide three-dimensional monitoring results of the entire process of lightning occurrence, development and breakdown, which facilitates a multi-angle understanding of the lightning discharge mechanism.

[0013] The aforementioned patents focus on lightning location research, but they lack waveform acquisition capabilities and cannot infer the characteristics of high-energy electrons from the waveform.

[0014] The invention patent "Combined Very Low Frequency and Very High Frequency Lightning Location System" (patent number 200910090116.8) discloses a multi-station location system that can monitor and locate lightning discharges in real time and in three dimensions. It includes at least four lightning location monitoring stations. Each station is equipped with a very low frequency lightning radiation receiver, a very high frequency lightning radiation receiver, a GPS receiver, and an industrial control computer. It can determine the number, type, characteristic parameters, and statistical characteristics of lightning discharge events occurring in a specified area, and monitor and warn of lightning activity.

[0015] The utility model patent "Lightning Multi-Band Detection Observation Station and System" (patent number 201420488037.9) discloses an observation station capable of comprehensively detecting lightning across multiple frequency bands, including very low frequency, low frequency, high frequency, and very high frequency. Each observation station consists of a broadband slow electric field change detector, a broadband fast electric field change detector, a very high frequency radiation source detector, and a high-speed data acquisition board. By setting up multiple observation stations, the discharge process and development and transmission characteristics of lightning can be studied.

[0016] The invention patent "Short Baseline Time Difference Method Ultra-High Frequency Lightning Radiation Source Detection and Positioning System" (Patent No. 200510041997.6) discloses a short baseline ultra-high frequency lightning radiation source detection and positioning system that integrates electric field measurement and time difference method. It can realize the accurate detection and positioning of lightning radiation sources at a single station and overcome some shortcomings of existing electric field measurement method and time difference method lightning positioning systems.

[0017] The invention patent "A Real-Time Feedback Electromagnetic Field Direction Finding Antenna System" (patent number 201910348999.1) discloses a real-time feedback electromagnetic field direction finding antenna system for lightning signal detection. Through the design of the magnetic loop antenna, it achieves accurate angle identification under high sensitivity conditions, overcoming the problems of the existing technology where the direction finding of lightning signals is greatly affected by background electromagnetic signals and the detection sensitivity is insufficient.

[0018] The drawback of the aforementioned patent is that it lacks a triggering algorithm for lightning double pulses, making it unable to identify double pulse signals from the numerous types of lightning signals.

[0019] The paper "Automatic Identification and Statistical Analysis of Bipolar Narrow Pulse" (Electric Porcelain Surge Arresters, Vol. 2, 2019) introduces an automatic identification method for bipolar narrow pulses in lightning. By statistically analyzing multiple time-domain parameters in the waveform of bipolar narrow pulses in lightning, nine parameters, including the high-frequency amplitude ratio, are selected as identification criteria to automatically identify bipolar narrow pulses mixed in with conventional lightning discharges, achieving an identification rate of 87.01%.

[0020] The literature “Time-Domain Feature Analysis of Lightning Electric Field Variation Waveform and Discharge Type Identification” (Meteorology, Vol. 35, No. 3) also proposes a method to identify negative ground flash return stroke, positive ground flash return stroke, and bipolar narrow pulse by utilizing the time-domain features of lightning electric field variation waveform.

[0021] The aforementioned literature has conducted relevant research on lightning electromagnetic pulse waveform identification, but the identification algorithm is relatively complex and computationally intensive, and can only be used for offline analysis, not for online identification.

[0022] The paper "Application of Time-Frequency Joint Analysis in Lightning Location" (Journal of Radio Science, Vol. 34, No. 4) proposes a new method for VHF lightning location. By utilizing the joint time-frequency distribution characteristics of the signal and combining the STDC function with WVD, the development process of the VHF radiation source of lightning can be clearly located.

[0023] The literature “VHF radiation characteristics analysis of ground flash discharge process” (Proceedings of the Chinese Society for Electrical Engineering, Vol. 25, No. 9) analyzed the pre-breakdown process and the step leader process of ground flash. By observing the VHF radiation characteristics of the radiated electric field at the center frequency of 280MHz of ground flash, the radiation pulses of ground flash in this frequency band were divided into three categories: isolated pulses, continuous pulses and multiple isolated pulses.

[0024] The aforementioned literature studied the radiation characteristics of VHF electromagnetic pulses during lightning processes, but it also lacked triggering algorithms and interference signal elimination methods for lightning double pulses, and could not effectively suppress signals such as interference from complex environments and artificial communication interference. Summary of the Invention

[0025] The purpose of this invention is to solve the problems of existing equipment being unable to conduct lightning detection experiments in near space and being unable to identify lightning double pulse signals from a wide variety of lightning signals during the detection process. This invention provides an experimental device suitable for lightning double pulse detection in near space, which can communicate with a near space floating platform, receive platform commands, and transmit status data. It has timing and positioning information, can collect and store measurement data, has triggering algorithms for lightning double pulses and interference signal elimination algorithms, and can identify target signals online and effectively suppress interference from complex environments.

[0026] The technical solution adopted by this invention to solve the above problems is as follows: A near-space lightning double pulse detection device uses a broadband antenna 1 to acquire electric field signals, a GPS / BeiDou receiving module 2 to collect position and time information, and an embedded software module 11 to control the device, including initializing configuration parameters, setting signal acquisition parameters, setting trigger algorithm parameters, receiving and feeding back commands, storing and exporting data, etc.; an interference elimination algorithm module 12 is responsible for real-time statistics of background spectrum energy and average power; a lightning double pulse signal identification algorithm module 13 is responsible for dynamically adjusting the number of triggering frequency points and triggering thresholds; a communication interface is connected to a floating platform payload compartment management computer 14, which is responsible for receiving commands from the ground command and control system and controlling the operation of the device; and a network interface is connected to a host computer 15 for device development and debugging and scientific data export.

[0027] The near-space lightning double pulse detection device of the present invention includes a broadband antenna 1, a GPS / BeiDou receiver module 2, an analog front-end conditioning circuit 3, an analog-to-digital converter module (ADC) 4, a preprocessor FPGA 5, a data buffer 6, a postprocessor FPGA 7, a data storage card 8, a network module 9, and a communication module 10; the postprocessor FPGA 7 includes an embedded software module 11; the embedded software module 11 includes an interference elimination algorithm module 12 and a lightning double pulse signal recognition algorithm module 13; auxiliary equipment includes a floating platform payload compartment management computer 14 and a host computer 15.

[0028] The broadband antenna 1 has an effective frequency range of 25MHz to 1GHz, and its output is connected to the analog front-end conditioning circuit 3.

[0029] The GPS / BeiDou receiver module 2 adopts a high-precision, low-power commercial standard module with a timing accuracy of 30ns and a positioning accuracy of 3m. Its output is connected to the post-processor FPGA7.

[0030] The analog front-end conditioning circuit 3 receives the radio frequency signal input from the broadband antenna 1 and performs circuit impedance transformation, signal gain transformation and low-pass filtering. Its output is connected to the analog-to-digital converter module ADC4.

[0031] The analog-to-digital converter module ADC4 uses a high-speed, dual-channel standard chip with a precision of 14 bits and a sampling rate of 3GSps. Its output is connected to the preprocessor FPGA5, and the transmitted signal is a high-speed serial signal using the 8-lane JESD204B protocol.

[0032] The preprocessor FPGA5 uses a commercially available standard chip to receive input data from the analog-to-digital converter module ADC4 and buffer it in data buffer 6. Data buffer 6 has a precision of 32-bit to 64-bit and a capacity of 4GB to 40GB, and is used to temporarily store sampled data. The output of the preprocessor FPGA5 is connected to the inter-chip LVDS communication interface 71 of the postprocessor FPGA7 via a high-speed LVDS bus.

[0033] The post-processor FPGA7 uses a commercial standard chip. It is the control center of this device and is responsible for initializing the device, managing local storage, communicating with the payload compartment management computer 14 of the floating platform, receiving time and location information from the GPS / BeiDou receiver module 2, and exporting data.

[0034] The data storage card 8 uses an industrial-grade SD card with a capacity of 32GB to store the probe data. It is connected to the SD storage interface 72 of the post-processor FPGA7. The processor FPGA can perform operations such as formatting, reading, writing, and querying on the data storage card 8.

[0035] Network module 9 uses a commercial standard PHY chip, with one end connected to the RGMII network interface 73 of the post-processor FPGA7 and the other end connected to the TCP / IP network interface of the host computer 15 to realize network communication with the host computer 15. The host computer 15 is mainly used for device development and debugging, scientific data export and subsequent processing.

[0036] One end of the communication module 10 is connected to the SPI communication interface 74 of the post-processor FPGA7, and the other end is connected to the airborne platform payload compartment management computer 14 via an RS422 bus.

[0037] The post-processor FPGA7 is connected to the analog front-end conditioning circuit 3 through the LVCMOS control interface 76, which can control the gain of the analog front-end conditioning circuit 3.

[0038] The post-processor FPGA7 is connected to the analog-to-digital converter module ADC4 via the JESD204B & SPI interface 77, which can control the sampling rate of the analog-to-digital converter module ADC4.

[0039] After the device is powered on, the embedded software module 11 executes the setting program in sequence, including: start → initialize analog front-end conditioning circuit 3 → initialize data buffer 6 → load firmware into the configuration space of post-processor FPGA7 → initialize peripherals → configure analog-to-digital converter module ADC4 → set parameters of pre-processor FPGA5 → receive information from GPS / BeiDou receiver module 2 → enter background survey mode or scientific exploration mode → end.

[0040] Interference elimination algorithm module 12 operates in background survey mode. Through background survey, it removes various narrowband and broadband interferences, including those from broadcasting, television, radio, radar, electrostatic discharge, and mounted platform instruments and equipment. Interference elimination algorithm module 12 includes the following operation steps:

[0041] Step 1: Select the broadband detection frequency band

[0042] The selection of broadband detection frequency bands needs to cover the characteristic frequency band of lightning double pulses and avoid common interference electromagnetic frequency bands. The broadband detection frequency bands are 25MHz~75MHz and 110MHz~300MHz.

[0043] Step 2: Statistical analysis of narrowband window energy intensity

[0044] The selected broadband detection frequency band is divided into 240 narrowbands, each with a window bandwidth of 1MHz. The average electromagnetic energy of the background signal in the 240 narrowbands over the past minute is calculated, and the average energy values ​​are sorted from high to low.

[0045] Step 3: Determine the narrowband trigger window

[0046] Excluding the 80 narrowband windows with high average energy, select 20 to 160 narrowband windows with low average energy as the trigger windows for lightning double pulse events; the selection of the number of trigger windows depends on how to balance the trigger rate of the lightning double pulse target signal and the false trigger rate of the background interference signal.

[0047] The lightning double-pulse signal identification algorithm module 13 operates in scientific detection mode. Through scientific detection, it identifies lightning double-pulse target signals online. The lightning double-pulse signal identification algorithm module 13 includes the following operation steps:

[0048] Step 1: Initially determine the narrowband trigger threshold multiplier.

[0049] The average energy of 20 to 160 narrowband trigger windows is statistically analyzed, and then multiplied by a constant term k as the trigger threshold increment for the linear term. The value of k ranges from 1 to 10, and the initial value is 2.

[0050] Step 2: Preliminary determination of narrowband trigger threshold plus coefficient

[0051] A constant term C is added to the trigger threshold in step one as the increment of the trigger threshold for the nonlinear term. The value of C ranges from 0.1 to 5, and the initial value is 1.5.

[0052] Step 3: Adjust the multiplication and addition coefficients in real time based on the detection results.

[0053] Conduct preliminary detection and adjust the settings of steps one and two based on the detection results; when the false trigger rate of the background interference signal is >10%, increase the multiplication factor and the addition factor; when the target signal trigger rate is <80%, decrease the multiplication factor and the addition factor.

[0054] Step 4: Adjust the number of narrowband trigger windows based on the detection results.

[0055] If, after adjusting the coefficients in step three, the false trigger rate of the background interference signal cannot be ≤10% even with increasing the multiplication and addition coefficients, then reduce the number of narrowband trigger windows; if, after decreasing the multiplication and addition coefficients, the target signal trigger rate cannot be ≥80%, then increase the number of narrowband trigger windows.

[0056] Step 5: Optimize the recognition algorithm

[0057] Repeat steps three and four until the false triggering rate of the background interference signal is ≤10% and the triggering rate of the target signal is ≥80%.

[0058] The beneficial effects of this invention are as follows: The near-space lightning double pulse detection device can be mounted on a near-space balloon platform to detect lightning double pulse signals. It features a sampling rate of up to 3GSps or higher, a weight of 3.5kg or less, power consumption of 30W or less, low computational load, and online identification capability. The interference signal elimination algorithm has strong environmental adaptability and can effectively reduce the false triggering rate of interference signals in complex environments to 10% or less, effectively improving the reliability of the system. The lightning double pulse signal identification algorithm module 13 can realize online monitoring of target signals and improve the target signal triggering rate to 80% or more. Attached Figure Description

[0059] Figure 1 Schematic diagram of near-space lightning double pulse detection device

[0060] In the diagram: 1. Broadband antenna; 2. GPS / BeiDou receiver module; 3. Analog front-end conditioning circuit; 4. Analog-to-digital converter (ADC); 5. Preprocessor FPGA; 6. Data buffer; 7. Postprocessor FPGA; 8. Data storage card; 9. Network module; 10. Communication module; 11. Embedded software module; 12. Interference cancellation algorithm module; 13. Lightning double pulse signal recognition algorithm module; 14. Aerial platform payload compartment management computer; 15. Host computer; 71. Inter-chip LVDS communication interface; 72. SD storage interface; 73. RGMII network interface; 74. SPI communication interface; 75. UART & IPPS interface; 76. LVCMOS control interface; 77. JESD204B & SPI interface.

[0061] Figure 2 Embedded software module operation flowchart

[0062] Figure 3 Interference elimination algorithm module operation flowchart

[0063] Figure 4 Operation flowchart of lightning double pulse signal recognition algorithm module Detailed Implementation

[0064] The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0065] Example 1

[0066] like Figure 1 As shown, the near-space lightning double pulse detection device includes a broadband antenna 1, a GPS / BeiDou receiver module 2, an analog front-end conditioning circuit 3, an analog-to-digital converter (ADC) module 4, a preprocessor FPGA 5, a data buffer 6, a postprocessor FPGA 7, a data storage card 8, a network module 9, and a communication module 10; the postprocessor FPGA 7 includes an embedded software module 11; the embedded software module 11 includes an interference elimination algorithm module 12 and a lightning double pulse signal recognition algorithm module 13; auxiliary equipment includes a floating platform payload compartment management computer 14 and a host computer 15.

[0067] The broadband antenna 1 is a small biconical commercial standard antenna with an effective frequency range of 25MHz to 1GHz. Its model number is ZN30503-1. The output terminal is connected to the signal input port of the analog front-end conditioning circuit 3. The connection line is a coaxial cable and the interface type is N.

[0068] The GPS / BeiDou receiver module 2 adopts a high-precision, low-power commercial standard module, model NB5442, with a timing accuracy of 30ns and a positioning accuracy of 3m. The output is connected to the UART & IPPS interface 75 of the post-processor FPGA7, and the output line is a coaxial cable with an N-type interface.

[0069] The analog front-end conditioning circuit 3 receives the RF signal detected by the broadband antenna 1 and processes it through a three-stage series circuit to meet the requirements of subsequent circuits for signal level, bandwidth, etc. The three-stage series circuit consists of: a first-stage impedance transformation circuit using broadband balun coupling to convert the single-ended input into a differential input. The balun uses Marki's BAL-0003 chip, with an input frequency range of 500kHz to 3GHz. The second-stage signal gain transformation circuit uses Analog Devices' ADA4961 RF programmable gain amplifier chip, with a -3dB bandwidth of 3.2GHz, a slew rate of 12000V / us, and an adjustable digital gain range of -6dB to +15dB. The third-stage low-pass filter circuit, which also serves as the interface circuit between the second-stage circuit and the analog-to-digital converter module ADC4, includes a 560nH RF inductor as the output load, a second-order Butterworth low-pass filter with a cutoff frequency of 1.5GHz, and is AC-coupled to the input interface of the ADC4 module.

[0070] The analog-to-digital converter (ADC4) uses a high-speed, dual-channel standard chip, model AD9208, with a precision of 14 bits, a sampling rate of 3 GSps, and a single-channel power consumption of 1.65W. The input terminal of the ADC4 is connected to the analog front-end conditioning circuit 3, and the output terminal is connected to the preprocessor FPGA5. The communication protocol uses the 8-lane JESD204B, and the signal format is a high-speed serial signal.

[0071] The preprocessor FPGA5 uses a commercial standard chip, model XCKU115-2FLVA1517I, which is specifically used to receive input data from the analog-to-digital converter module ADC4 and buffer the data to the data buffer 6. The output of the preprocessor FPGA5 is connected to the inter-chip LVDS communication interface 71 of the postprocessor FPGA7 through the LVDS high-speed data bus.

[0072] Data buffer 6 uses commercial standard components, has a precision of 64 bits, and a capacity of 9GB, and is used to temporarily store sampled data.

[0073] The post-processor FPGA7 uses a commercial standard chip, model Zynq-7020, and is the control center of this device. It is responsible for executing the embedded software module 11, realizing functions such as device initialization, local storage management, communication between the device and the payload cabin management computer 14 of the floating platform, control data export, and receiving time and location information from the GPS / BeiDou receiver module 2.

[0074] The data storage card 8 uses an industrial-grade SD card with a capacity of 32GB to store the probe data. It is connected to the SD storage interface 72 of the post-processor FPGA7. The processor FPGA can perform operations such as formatting, reading, writing, and querying on the data storage card 8.

[0075] Network module 9 uses a commercial standard PHY chip, with one end connected to the RGMII network interface 73 of the post-processor FPGA7 and the other end connected to the TCP / IP network interface of the host computer 15 to realize network communication with the host computer 15. The host computer 15 is mainly used for device development and debugging, scientific data export and subsequent processing.

[0076] One end of the communication module 10 is connected to the SPI communication interface 74 of the post-processor FPGA7, and the other end is connected to the airborne platform payload compartment management computer 14 via an RS422 bus.

[0077] The post-processor FPGA7 is connected to the analog front-end conditioning circuit 3 through the LVCMOS control interface 76, which can control the gain of the analog front-end conditioning circuit 3.

[0078] The post-processor FPGA7 is connected to the analog-to-digital converter module ADC4 via the JESD204B & SPI interface 77, which can control the sampling rate of the analog-to-digital converter module ADC4.

[0079] Embedded software module 11 includes interference cancellation algorithm module 12 and lightning double pulse signal recognition algorithm module 13, which will be specifically combined Figure 2 , Figure 3 Please provide an explanation.

[0080] The airship platform payload compartment management computer 14 is responsible for receiving commands from the ground command system, controlling the operation of the equipment, and providing feedback on the execution results of the commands. The airship platform payload compartment management computer 14 is connected to the device's communication module 10.

[0081] like Figure 2 As shown, after the device is powered on, the embedded software module 11 executes the setting program in sequence, including initializing the analog front-end conditioning circuit 3, initializing the data buffer 6, loading the firmware into the configuration space of the post-processor FPGA7, initializing the peripherals, configuring the analog-to-digital converter module ADC4, setting the parameters of the pre-processor FPGA5, receiving information from the GPS / BeiDou receiver module 2, and entering the background survey mode or scientific exploration mode.

[0082] like Figure 3 As shown, the interference elimination algorithm module 12 operates in the background survey mode of the embedded software module 11. Through background survey, it removes various narrowband and broadband interferences, including those from broadcasting, television, radio, radar, electrostatic discharge, and mounted platform instruments and equipment. The interference elimination algorithm module 12 includes the following operation steps:

[0083] Step 1: Select the broadband detection frequency band

[0084] The selection of electromagnetic frequency bands needs to cover the characteristic frequency band of lightning double pulses and avoid common interfering electromagnetic frequency bands. Common interfering electromagnetic frequency bands may vary depending on the time and location, and need to be determined through actual measurements. In this embodiment, the experimental location was the Da Qaidam test site in Haixi Mongolian and Tibetan Autonomous Prefecture, Qinghai Province, during autumn daytime, with the floating platform flying at an altitude of 20km. After background signal measurements, frequency bands with strong interference were eliminated, and the selected detection frequency bands were 25MHz–75MHz and 110MHz–300MHz.

[0085] Step 2: Statistical analysis of narrowband window energy intensity

[0086] The selected electromagnetic frequency band is divided into multiple narrowbands. In this embodiment, the narrowband window width is set to 1MHz. 50 narrowband windows are selected in the range of 25MHz to 75MHz, and 190 narrowband windows are selected in the range of 110MHz to 300MHz, for a total of 240 narrowband windows. The average energy of the 240 narrowband windows over the past minute is calculated. Each narrowband window is sampled 600 times over the past minute (10 samples per second). The average energy of the 600 samples from the 240 narrowband windows is then sorted from highest to lowest.

[0087] Step 3: Determine the narrowband trigger window

[0088] Excluding the 80 narrowband windows with high average energy, the 65 narrowband windows with low average energy were selected as the narrowband trigger windows for lightning double pulse events.

[0089] like Figure 4 As shown, the lightning double pulse signal recognition algorithm module 13 operates in the scientific detection mode of the embedded software module 11. Through scientific detection, it accurately extracts the lightning double pulse target signal, improving the system trigger rate and reducing the false trigger rate of background interference signals. The lightning double pulse signal recognition algorithm module 13 includes the following operation steps:

[0090] Step 1: Initially determine the narrowband trigger threshold multiplier.

[0091] In this embodiment, the average energy of the 65 initially selected narrowband trigger windows is statistically analyzed and multiplied by a coefficient of 2 to serve as the trigger threshold increment for the linear term. The purpose is to dynamically reduce the false trigger rate of background interference signals within a second.

[0092] Step 2: Preliminary determination of narrowband trigger threshold plus coefficient

[0093] Add a coefficient of 1.5 to the trigger threshold determined in step one as the trigger threshold increment for the nonlinear term. The purpose is to statically reduce the false trigger rate of background interference signals within a minute time.

[0094] Step 3: Adjust the multiplication and addition coefficients in real time based on the detection results.

[0095] A preliminary detection is conducted, and the settings for steps one and two are adjusted based on the detection results. In this embodiment, the false trigger rate of the background interference signal is 1.5%, which meets the requirement of ≤10%, but the trigger rate of the target signal is low, only 70%. Therefore, the multiplication factor is reduced to 1.8, and the addition factor is reduced to 1.4.

[0096] Step 4: Adjust the number of narrowband trigger windows based on the detection results.

[0097] Based on the preliminary detection results of step three, the false triggering rate of the background interference signal is 1.9%, which meets the requirement of ≤10%, but the triggering rate of the target signal is still low, at only 78%. Therefore, the number of narrowband trigger windows is adjusted to 55.

[0098] Step 5: Optimize the recognition algorithm

[0099] Repeat steps three and four, finally setting the number of trigger windows to 58, multiplying by a factor of 1.7, and adding by a factor of 1.35. The target signal trigger rate is 90.2%, and the false trigger rate of background interference signals is 2.5%, meeting the usage requirements.

[0100] The near-space lightning double-pulse detection device of this invention can communicate with a near-space floating platform, receive platform commands, and transmit status data. It has timing and positioning functions, can collect and store measurement data, accurately identify lightning double-pulse target signals, and effectively suppress interference from complex environments and artificial communication. This device features a sampling rate of up to 3GSps, a weight as light as 3.5kg, power consumption as low as 30W, a false trigger rate as low as 2.5%, a target signal trigger rate as high as 90.2%, low computational load, online identification capability, and strong environmental adaptability.

Claims

1. A near-space lightning double-pulse detection device, characterized in that: The device includes a broadband antenna (1), a GPS / BeiDou receiver module (2), an analog front-end conditioning circuit (3), an analog-to-digital converter (ADC) module (4), a preprocessor FPGA (5), a data buffer (6), a post-processor FPGA (7), a data storage card (8), a network module (9), and a communication module (10). The analog signal detected by the broadband antenna (1) is input to the analog front-end conditioning circuit (3), and after impedance transformation, signal gain transformation, and low-pass filtering by the analog front-end conditioning circuit (3), it is output to the analog-to-digital converter (ADC) module (4). The analog-to-digital converter (ADC) module (4) converts the analog signal into a digital signal and outputs it to the preprocessor FPGA (5). (5) Store data in a data buffer (6) and call it at any time; the output of the preprocessor FPGA (5) is connected to the inter-chip LVDS communication interface (71) of the postprocessor FPGA (7); the data storage card (8) is connected to the SD storage interface (72) of the postprocessor FPGA (7); one end of the network module (9) is connected to the RGMII network interface (73) of the postprocessor FPGA (7), and the other end of the network module (9) is connected to the host computer (15); one end of the communication module (10) is connected to the SPI communication interface (74) of the postprocessor FPGA (7), and the other end of the communication module (10) is connected to the airborne platform payload cabin management computer (14); GPS / BeiDou receiver The output of module (2) is connected to the UART & IPPS interface (75) of the post-processor FPGA (7); the LVCMOS control interface (76) of the post-processor FPGA (7) is connected to the analog front-end conditioning circuit (3) and controls the gain of the analog front-end conditioning circuit (3); the JESD204B & SPI interface (77) of the post-processor FPGA (7) is connected to the analog-to-digital converter module ADC (4) and controls the sampling rate of the analog-to-digital converter module ADC (4); the post-processor FPGA (7) includes an embedded software module (11); the embedded software module (11) includes an interference cancellation algorithm module (12) and a lightning double pulse signal recognition algorithm module (13). Interference elimination algorithm module (12) and lightning double pulse signal recognition algorithm module (13) are connected in sequence; the operation steps of embedded software module (11) are as follows: Start → Initialize analog front-end conditioning circuit (3) → Initialize data buffer (6) → Load firmware to post-processor FPGA (7) configuration space → Initialize peripherals → Configure analog-to-digital conversion module ADC (4) → Set pre-processor FPGA (5) parameters → Receive GPS / BeiDou receiver module (2) information → Enter background survey mode or scientific exploration mode → End. Among them, the background survey mode requires running the interference elimination algorithm module (12), and the scientific exploration mode requires running the lightning double pulse signal recognition algorithm module (13).The operation steps of the interference elimination algorithm module (12) are as follows: Step 1: Select a broadband detection frequency band. The selection of the broadband detection frequency band needs to cover the characteristic frequency band of lightning double pulse and avoid common interference electromagnetic frequency bands. The broadband detection frequency band is 25MHz~75MHz and 110MHz~300MHz; Step 2: Statistically calculate the energy intensity of the narrowband window. Divide the selected broadband detection frequency band into 240 narrowbands. The bandwidth of each narrowband window is 1MHz. Statistically calculate the average electromagnetic energy of the background signal in the 240 narrowbands in the past 1 minute. Rank the average energy from high to low. Low sorting; Step 3: Determine the narrowband trigger window, exclude the 80 narrowband windows with high average energy, and select 20 to 160 narrowband windows with low average energy as the trigger windows for the lightning double pulse event. The selection of the number of trigger windows depends on how to balance the trigger rate of the lightning double pulse target signal and the false trigger rate of the background interference signal; The operation steps of the lightning double pulse signal recognition algorithm module (13) are as follows: Step 1: Preliminarily determine the narrowband trigger threshold multiplier, statistically calculate the average energy of the selected 20 to 160 narrowband trigger windows, and multiply it by a constant term k. Step 1: As the trigger threshold increment for the linear term, k ranges from 1 to 10, with an initial value of 2; Step 2: Initially determine the narrowband trigger threshold plus coefficient, adding a constant term C to the trigger threshold in Step 1 as the trigger threshold increment for the nonlinear term, with C ranging from 0.1 to 5, and an initial value of 1.5; Step 3: Adjust the multiplication and addition coefficients in real time based on the detection results, conduct preliminary detection, and adjust the settings of Step 1 and Step 2 based on the detection results; When the false trigger rate of the background interference signal is >10%, increase the multiplication and addition coefficients, and when the target signal trigger rate... When the false trigger rate is less than 80%, reduce the multiplication and addition coefficients; Step 4: Adjust the number of narrowband trigger windows based on the detection results. Check the detection results after adjusting the coefficients in Step 3. If increasing the multiplication and addition coefficients still cannot make the false trigger rate of the background interference signal ≤10%, then reduce the number of narrowband trigger windows. If decreasing the multiplication and addition coefficients still cannot make the target signal trigger rate ≥80%, then increase the number of narrowband trigger windows; Step 5: Optimize the recognition algorithm. Repeat Steps 3 and 4 until the false trigger rate of the background interference signal is ≤10% and the target signal trigger rate is ≥80%.

2. The near-space lightning double-pulse detection device according to claim 1, characterized in that: The ADC(4) module has an accuracy of 12-bit to 14-bit and a sampling rate of 2GSps to 3GSps.

3. The near-space lightning double-pulse detection device according to claim 1, characterized in that: The overall bandwidth of the device is 25MHz to 1GHz.

4. A near-space lightning double-pulse detection device according to claim 1, characterized in that: The data cache (6) has a precision of 32-bit to 64-bit and a capacity of 4-GB to 40-GB.

Images