FPGA-based low-frequency magnetic source detection and positioning system and application

Abstract

This invention belongs to the field of magnetic source detection and positioning technology, and discloses a low-frequency magnetic source detection and positioning system and its application based on FPGA. The system structure includes: a low-frequency magnetic source generation module for generating target low-frequency magnetic source signals; a low-frequency magnetic source detection module for detecting the target low-frequency magnetic source signals at the observation point; and a signal sampling and receiving module for receiving the detected three-dimensional magnetic signals and sampling them. The sampled data is processed by the FPGA processing module to output the spatial coordinates of the detected magnetic source target, thereby achieving accurate positioning. This invention provides a low-frequency magnetic source detection and positioning system based on lock-in amplification by combining low-frequency magnetic source detection technology with lock-in amplification signal processing method, and introduces FPGA hardware acceleration to construct a magnetic source detection and positioning system suitable for low-altitude detection and positioning applications of UAVs. It effectively solves the problems of insufficient real-time performance, low system integration, and high power consumption of existing magnetic source detection and positioning technologies when applied to UAV platforms.

Description

Technical Field

[0001] This invention belongs to the field of magnetic source detection and positioning technology, and more specifically, relates to a low-frequency magnetic source detection and positioning system and its application based on FPGA. Background Technology

[0002] With the rapid rise of the low-altitude economy, drones are increasingly being used in various fields such as logistics delivery, environmental monitoring, public safety, and power line inspection. For example, in logistics delivery, drones, with their core characteristics of high efficiency, flexibility, and freedom from ground traffic restrictions, effectively solve the pain point of the "last mile" of urban delivery, attracting widespread attention from logistics companies and the market. However, in practical applications, drones face many positioning challenges when flying at low altitudes in complex urban environments: the towering buildings in cities can easily block and reflect satellite navigation signals, and complex electromagnetic environments (such as urban power grids, communication base stations, and electronic devices) can easily interfere with navigation signals. This leads to frequent signal interruptions, decreased positioning accuracy, and even failures in traditional satellite navigation-based positioning methods. This not only affects the flight stability and mission execution efficiency of drones but also poses a serious threat to flight safety, easily causing drone collisions, loss of connection, and other safety accidents. Therefore, developing high-precision positioning technology adapted to complex low-altitude environments has become a key prerequisite for promoting the large-scale application of drones.

[0003] Among existing low-altitude target detection and positioning technologies, radar technology is the primary means of low-altitude detection, possessing all-weather operation capabilities and enabling the detection and monitoring of low-altitude targets. It is one of the core technologies in the low-altitude air defense system. However, in low-altitude scenarios, it is susceptible to interference from ground clutter and complex electromagnetic environments, limiting its ability to identify small, low-detectability targets. Furthermore, some radar systems have blind spots at close range, and the manufacturing and maintenance costs are relatively high. Optoelectronic detection technology, based on optical sensing, can clearly identify the appearance features of targets, allowing for intuitive identification and observation of low-altitude targets. It has application value in scenarios requiring precise target identification. However, it is significantly affected by lighting conditions, weather conditions, and obstructions, resulting in limited detection range and difficulty in achieving stable and continuous low-altitude monitoring. Acoustic detection technology is a passive detection method, detecting targets by capturing their acoustic signals without actively emitting signals. It is suitable for sensing low-altitude targets at close range. However, it is significantly affected by environmental noise and weather conditions, severely limiting its detection range and positioning accuracy, making it unsuitable for medium- and long-range low-altitude detection scenarios. Radio detection technology relies on receiving radio signals emitted by targets to conduct detection work. It can identify and locate targets radiating specific radio signals. Detection tasks require combining relevant signal characteristics, but it is overly dependent on radio signals actively emitted by targets, resulting in a narrow range of applications. Furthermore, it is susceptible to interference in complex electromagnetic environments, and its positioning accuracy typically requires multi-station coordination, leading to high system deployment complexity and maintenance difficulties. Multi-source information fusion solutions are comprehensive solutions that integrate multiple detection technologies. By aggregating data collected from different sensors, it achieves multi-dimensional information complementarity, improving the comprehensiveness and reliability of low-altitude detection and adapting to complex low-altitude detection scenarios. However, its system structure is complex, data processing volume is large, and it places high demands on computing resources, power consumption, and system integration. This makes it difficult to adapt to platforms with limited payload and energy consumption, such as UAVs, significantly restricting its application scenarios.

[0004] Compared to other low-altitude detection and positioning methods such as radar, photoelectric, and acoustic methods, magnetic source detection and positioning technology does not rely on lighting or weather conditions, is unaffected by obstructions such as tall buildings and trees, has strong resistance to electromagnetic interference, and boasts advantages such as stable operation, high positioning accuracy, and relatively compact equipment size. It is well-suited for applications in densely built-up urban environments with easily blocked signals and complex electromagnetic environments. Among existing magnetic source positioning technologies, detection methods based on static magnetic fields are widely used, but most can only determine the presence or absence of magnetic targets, making it difficult to accurately locate the spatial coordinates of the magnetic source. The core reason for this is that static magnetic field gradients decay rapidly and are susceptible to interference from the Earth's magnetic field and environmental ferromagnetic interference. Furthermore, existing technologies lack effective three-dimensional calculation mechanisms, making it impossible to establish a stable mapping between the magnetic field and spatial position, and thus difficult to deduce the specific spatial coordinates of the magnetic source.

[0005] Compared with static magnetic field detection, using dynamic magnetic fields (low-frequency magnetic fields) as magnetic sources for detection and positioning has the core advantage of making low-frequency magnetic signals easy to detect and capture. It can accurately obtain the signal amplitude and phase information of magnetic targets for positioning calculations. At the same time, it also has the inherent advantages of magnetic detection itself, such as resistance to obstruction and electromagnetic interference. Therefore, it can achieve high-precision positioning of the spatial coordinates of the magnetic source, adapt to complex low-altitude scenarios such as dense urban buildings, and effectively make up for the shortcomings of static magnetic field detection in positioning. However, the current mainstream positioning solution for this type of system is "magnetic gate detection + PC-based software calculation," which has significant limitations. Specifically, the large size, weight, and power consumption of magnetic gate sensors result in bulky detection devices that cannot meet miniaturization requirements. Furthermore, the limited bandwidth of magnetic gate sensors leads to poor performance in measuring dynamic magnetic fields. The PC-based calculation method suffers from data transmission and computation delays, resulting in insufficient real-time performance. The existing magnetic source detection positioning algorithms themselves are computationally complex, involving multi-channel signal processing, model fitting, and inversion calculations, further increasing the operational threshold. In addition, existing magnetic field detection positioning systems have a fragmented architecture and low integration. The magnetic field signals need to be transmitted to external computing units via communication links, which easily introduces additional delays, increases system power consumption and communication complexity, and affects overall reliability. Combined with the complexity of the overall algorithm and system architecture, this ultimately makes it difficult to achieve lightweight implementation in UAV positioning scenarios. Therefore, existing magnetic source positioning technology still needs further improvement. Summary of the Invention

[0006] In view of the above-mentioned defects or improvement needs of the existing technology, the present invention provides a low-frequency magnetic source detection and positioning system and application based on FPGA. Its purpose is to improve the real-time performance and accuracy of low-frequency magnetic source positioning, improve the integration, and realize the lightweight implementation of the technology.

[0007] To achieve the above objectives, the present invention provides a low-frequency magnetic source detection and positioning system based on FPGA, comprising: Low-frequency magnetic source generation module, used to generate target low-frequency magnetic source signal; A low-frequency magnetic source detection module is set at at least two observation points on the ground to detect the target low-frequency magnetic source signal and obtain the three-dimensional magnetic detection signal detected at each observation point. The signal sampling and receiving module is used to sample the three-dimensional magnetic detection signal by channel to obtain magnetic detection data for each channel; The FPGA processing module is used to receive and store the sampled magnetic detection data, and extract the magnetic detection data from at least one channel for spectrum analysis to obtain the frequency of the target low-frequency magnetic source signal. The magnetic detection data from each channel undergoes a first phase-locked amplification and quadrature detection to obtain the phase of the target low-frequency magnetic source signal. The frequency of the reference signal used in the first phase-locked amplification operation and quadrature detection is... ; the frequency and the phase The frequency and phase of the reference signal used for the second phase-locked amplification operation and quadrature detection are used to perform the second phase-locked amplification operation and quadrature detection on the magnetic detection data of each channel to obtain the amplitude of the target low-frequency magnetic source signal; based on the amplitude, a three-dimensional magnetic source positioning algorithm is used to obtain the spatial coordinates of the target low-frequency magnetic source signal, thereby realizing the positioning of the low-frequency magnetic source.

[0008] Furthermore, the phase of the target low-frequency magnetic source signal for: α [- , ) in, , They respectively represent the use of , As a reference signal pair (t) The result of the first orthogonal detection. The amplitude of the reference signal, For magnetic detection data of one channel, characterization The x-axis component of the low-frequency magnetic source signal detected from the target at any given time; , They respectively represent the use of , As a reference signal pair The results of orthogonal detection For magnetic detection data of one channel, characterization The component of the low-frequency magnetic source signal detected at any time on the y-axis.

[0009] Furthermore, the amplitude of the target low-frequency magnetic source signal for:

[0010] in, ; ; ; ; ; ; In the formula, , , They respectively represent the use of As a reference signal pair , , The results of the second orthogonal detection; , , They respectively represent the use of As a reference signal pair , , The results of orthogonal detection; For magnetic detection data of one channel, characterization The z-axis component of the low-frequency magnetic source signal detected at any time.

[0011] Furthermore, in the FPGA processing module, an adjustable phase reference signal generator is used to generate the reference signal; wherein, the adjustable phase reference signal generator internally stores a single 1 / 4-cycle sin and cos signal data; the adjustable phase reference signal generator includes a phase accumulator, and during the first phase-locked amplification operation and quadrature detection, the phase control word of the phase accumulator is 0, and it is adjusted according to frequency. The cycle is repeated to generate a frequency. Reference signal with phase 0 and During the second phase-locked loop amplification operation and quadrature detection, the phase accumulator... For phase control words, by frequency The cycle is repeated to generate a frequency. Phase is Reference signal and .

[0012] Furthermore, in the FPGA processing module, a digital integrator is used to perform the integration of the magnetic detection data and the reference signal of each channel in the first and second phase-locked amplification operations. The digital integrator employs a three-stage pipeline of value acquisition, integration, and accumulation to achieve parallel execution of the integration operation. The value acquisition involves acquiring the magnetic detection data of the corresponding channel and its corresponding reference signal; the integration involves multiplying the magnetic detection data with its corresponding reference signal; and the accumulation involves summing the results of the integration operation to achieve the integration operation.

[0013] Furthermore, in the FPGA processing module, an arctan calculation module is used to realize the phase of the target low-frequency magnetic source signal. The solution is obtained by using the CORDIC algorithm in the arctan solution module. The solution.

[0014] Furthermore, magnetic detection data from at least one channel is extracted for spectral analysis to obtain the frequency of the target low-frequency magnetic source signal. ,include: At least one channel of magnetic detection data is extracted and subjected to spectral analysis using Fast Fourier Transform to initially obtain the frequency of the target low-frequency magnetic source signal. The initially obtained frequency is then fitted and optimized to obtain the final frequency of the target low-frequency magnetic source signal. .

[0015] Furthermore, the low-frequency magnetic source generating module is one or more spliced ​​neodymium iron boron permanent magnets rotating in a certain plane, or it is an energized coil; The low-frequency magnetic source detection module includes a sensor and a signal acquisition board; the sensor is used to acquire the components of the target low-frequency magnetic source signal in the x, y, and z axes; the signal acquisition board is used to process the target low-frequency magnetic source signal components in the three axes to obtain the three-dimensional magnetic detection signal; wherein, the processing includes one or more of AC coupling, signal amplification, and bandpass filtering; The signal sampling and receiving module is an ADC chip; the sampling parameters and sampling logic of the ADC chip are controlled by the FPGA processing module.

[0016] Furthermore, the sensor is a TMR sensor; the TMR sensor is composed of three single-axis TMR sensors arranged in a three-dimensional orthogonal manner, or the TMR sensor is an integrated three-dimensional TMR sensor chip; The present invention also provides a low-altitude magnetic source detection and positioning system for unmanned aerial vehicles (UAVs), wherein the magnetic source detection and positioning system is the FPGA-based low-frequency magnetic source detection and positioning system described in any of the above-mentioned embodiments; wherein the low-frequency magnetic source generating module is mounted on the UAV and uses the spatial coordinates of the obtained target low-frequency magnetic source signal as the position information of the UAV to achieve low-altitude positioning of the UAV.

[0017] In summary, the above-described technical solutions conceived in this invention can achieve the following beneficial effects: (1) To address the technical challenges of high computational complexity, limited positioning accuracy, and insufficient real-time performance caused by reliance on host computer software in existing magnetic source positioning algorithms, this invention combines low-frequency magnetic source detection technology with lock-in amplification signal processing to design a low-frequency magnetic source coordinate calculation method based on lock-in amplification. Specifically, considering that the low-frequency magnetic source signal is the superposition of two signals with the same frequency and phase and different amplitudes, this invention designs a double orthogonal detection lock-in amplification algorithm adapted to three-dimensional magnetic signals based on the characteristics of the low-frequency magnetic source signal. During the first lock-in amplification, the reference signal is at the same frequency as the magnetic source signal, and the phase of the low-frequency magnetic source signal is calculated through the orthogonal detection operation of the first lock-in amplification. And shift the phase of the reference signal to A second phase-locked amplification orthogonal detection operation is performed. During the second phase-locked amplification, the reference signal and the magnetic source signal are in phase and frequency. The phase-locked amplification result is only related to the amplitude of the magnetic source signal, thus obtaining the amplitude of the magnetic source signal and realizing the spatial positioning of the low-frequency magnetic source. This invention designs a phase-locked amplification algorithm based on the signal characteristics of the magnetic source signal, and separates the amplitude information of the magnetic source signal through phase locking. Phase-locked amplification processing can achieve narrowband filtering, effectively extracting weak magnetic source signals and suppressing background noise and environmental electromagnetic interference, improving the accuracy and stability of detection and positioning, and avoiding complex mathematical calculations. At the same time, the algorithm is hardware accelerated through FPGA, which greatly improves the computational parallelism of the algorithm, effectively improving the real-time performance of magnetic source positioning, and also improving the integration, realizing the lightweight implementation of the technology. The detection target of this invention is a low-frequency magnetic source signal, which effectively improves the detection range.

[0018] (2) Furthermore, in response to the common problems of large size, heavy weight and high power consumption of existing magnetic source detection devices in the magnetic source detection and positioning scenario, an efficient magnetic source signal acquisition scheme is proposed. By integrating a low-frequency magnetic source detection module including a TMR sensor and a signal acquisition board, the detection device is made lightweight and low-power. At the same time, the TMR sensor has the advantages of high sensitivity, low noise and wide dynamic range. The signal acquisition board realizes the AC coupling, signal amplification and bandpass filtering of the initial low-frequency magnetic source signal, which further improves the detection accuracy.

[0019] (3) In view of the problems that existing magnetic source detection and positioning systems are complex in structure and have low integration, which cannot meet the lightweight carrying requirements of UAVs and are difficult to implement, this invention constructs a magnetic source detection and positioning system with high integration and lightweight that can be directly used for UAVs based on FPGA. The system has a simple structure and is easy to implement in engineering and practical applications.

[0020] In summary, this invention provides a low-frequency magnetic source detection and positioning system based on lock-in amplification by combining low-frequency magnetic source detection technology with lock-in amplification signal processing method. It also introduces FPGA hardware acceleration to build a magnetic source detection and positioning system suitable for low-altitude detection and positioning applications of UAVs. This effectively solves the problems of insufficient real-time performance, low system integration, and high power consumption of existing magnetic source detection and positioning technologies when applied to UAV platforms. Attached Figure Description

[0021] Figure 1 This is a block diagram of the FPGA-based low-frequency magnetic source detection and positioning system in an embodiment of the present invention.

[0022] Figure 2 This is a circuit structure diagram of the low-frequency magnetic source detection module in an embodiment of the present invention.

[0023] Figure 3 This is a schematic diagram of the application structure of the UAV detection and positioning system in an embodiment of the present invention.

[0024] Figure 4 This is a diagram of the internal computing architecture of the FPGA processing module in an embodiment of the present invention.

[0025] Figure 5 This is a schematic diagram of the low-frequency magnetic source field distribution modeling in an embodiment of the present invention.

[0026] Figure 6 This is a diagram showing the distribution characteristics of the amplitude / signal-to-noise ratio of the sensors in the x, y, and z directions of the low-frequency magnetic source detection module of this invention as a function of distance.

[0027] Figure 7 This is a diagram showing the distribution characteristics of the anisotropic detection signal amplitudes of the sensors in the x, y, and z directions of the low-frequency magnetic source detection module of this invention as a function of the detection angle at equal distances.

[0028] Figure 8 This is a graph showing the signal-to-noise ratio enhancement performance of the lock-in amplification algorithm in the FPGA processing module of this invention.

[0029] Figure 9 This is a diagram showing the phase detection error characteristics of the lock-in amplification algorithm in the FPGA processing module of this invention.

[0030] Figure 10 This is a characteristic diagram of amplitude detection error in the phase-locked amplification algorithm in the FPGA processing module of this invention.

[0031] Figure 11 A quantitative comparison chart of the key performance of four low-altitude positioning technologies.

[0032] Figure 12 This is the result of testing 15 sets of positioning data on a horizontal plane with Z=1m in this embodiment of the invention.

[0033] Figure 13 This is the result of testing 15 sets of positioning data on the Z=3m horizontal plane in an embodiment of the present invention.

[0034] Figure 14 This is the result of testing 15 sets of positioning data on a horizontal plane with Z=5m in this embodiment of the invention.

[0035] Figure 15 This is the result of testing 15 sets of positioning data on the Z=7m horizontal plane in an embodiment of the present invention.

[0036] Figure 16 This is the result of testing 15 sets of positioning data on the Z=9m horizontal plane in an embodiment of the present invention.

[0037] Figure 17 This is the result of testing 15 sets of positioning data on a horizontal plane with Z=10m in this embodiment of the invention.

[0038] Figure 18 This refers to the relative error of three-dimensional positioning in the embodiments of the present invention under the conditions of Z=1 m and 3 m.

[0039] Figure 19 This shows the relative error distribution of each test point at different heights Z in this embodiment of the invention. Detailed Implementation

[0040] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. Furthermore, the technical features involved in the various embodiments of this invention described below can be combined with each other as long as they do not conflict with each other.

[0041] This invention provides a low-frequency magnetic source detection and positioning system based on FPGA, and applies it to low-altitude positioning scenarios of UAVs. The structure of the FPGA-based low-frequency magnetic source detection and positioning system is as follows: Figure 1 As shown, it includes a low-frequency magnetic source generating module, a low-frequency magnetic source detection module, a signal sampling and receiving module, and an FPGA processing module. The low-frequency magnetic source detection module is respectively located at at least two observation points near the signal sampling and receiving module and the FPGA processing module, as shown below. Figure 3 As shown, the detection and positioning system in this embodiment of the invention uses the main observation point O as the coordinate origin and an auxiliary observation point Q with known coordinates to detect low-frequency magnetic sources. Two-point detection is performed. The three-dimensional magnetic signals at each observation point are received by the low-frequency magnetic source detection module and then by the signal sampling and receiving module. The FPGA processing module then calculates and outputs the spatial coordinates of the detection magnetic source target, achieving precise positioning. The location of the low-frequency magnetic source is then determined. This refers to the position relative to the main observation point. The functions and implementation methods of each structural module are as follows: (1) Low-frequency magnetic source generation module The low-frequency magnetic source generation module generates low-frequency magnetic source signals, which are then used as detection targets for the system's spatial exploration. Simultaneously, it works in conjunction with low-frequency magnetic source detection modules at two observation points to achieve three-dimensional magnetic signal detection at at least two observation points. This module is installed on the UAV and works collaboratively with the low-frequency magnetic source detection modules at the two observation points. The low-frequency magnetic source detection modules at the observation points capture the low-frequency magnetic signals generated by the low-frequency magnetic source generation module and transmit them to the signal sampling and receiving module and the FPGA processing module for calculation, thereby realizing the dual functions of magnetic source spatial exploration and UAV coordinate calculation.

[0042] In this embodiment of the invention, the magnetic source signal frequency of the low-frequency magnetic source generation module is 140Hz, and the bandpass selection of the signal acquisition board is 140Hz±20%. The magnetic source signal frequency can be shifted to other frequency bands according to actual application requirements. The core hardware modules of the system (low-frequency magnetic source detection module, signal sampling and receiving module, and FPGA processing module) can be highly integrated. After integration, the device size can be controlled within approximately 10cm×10cm×2cm, and the weight does not exceed approximately 300g. It can be flexibly deployed at ground observation points, UAV platforms, or other mobile carriers, meeting the engineering requirements of UAV low-altitude applications for miniaturization, low power consumption, and low cost. Compared with electromagnetic wave signals, radar signals, and optical signals, the low-frequency magnetic field signal used by the system has the characteristics of strong diffraction ability and good penetration. It has strong anti-blocking and anti-interference capabilities in low-altitude application scenarios such as building obstruction and complex electromagnetic environments. Combined with the parallel processing hardware acceleration characteristics of FPGA, real-time hardware-level processing of magnetic field signals can be realized, taking into account positioning accuracy, real-time performance, and operational stability.

[0043] The low-frequency magnetic source generation module can be implemented using two schemes. Scheme 1 uses a neodymium iron boron permanent magnet rotating on a certain plane (such as the xoy plane, xoz plane, yoz plane) to generate a low-frequency magnetic field (low-frequency magnetic source signal) to be detected and located. In this embodiment, a model N52 neodymium iron boron permanent magnet with dimensions of 20mm×10mm×5mm is used, and multiple magnets can also be spliced ​​together. Scheme 2 uses an energized coil to generate the low-frequency magnetic field to be detected and located. In this embodiment, the input voltage of the energized coil is 12V, the output magnetic field strength is 0.5~2mT, and the operating frequency is adjustable at 150Hz. The installation position is at the center of the UAV fuselage or wing component, and the installation surface is parallel to the horizontal plane of the UAV to ensure uniform radiation of the magnetic signal. The function of the low-frequency magnetic source generation module is to generate a stable low-frequency magnetic signal as a signal source for detecting and locating the target. Its magnetic field distribution conforms to the sinusoidal distribution characteristic of attenuation with distance. This signal distribution law is the theoretical basis for the phase-locked amplification technology and the two-point coordinate algorithm in the detection and positioning system of this embodiment.

[0044] (2) Low-frequency magnetic source detection module The low-frequency magnetic source detection modules are respectively set at two observation points on the ground. Each low-frequency magnetic source detection module includes a sensor and a signal acquisition board. The sensor is used to acquire the components of the low-frequency magnetic source signal generated by the low-frequency magnetic source generation module in the x, y, and z axes. The signal acquisition board is used to process the low-frequency magnetic source signal components in the x, y, and z axes, including AC coupling, signal amplification, and bandpass filtering, to obtain a three-dimensional magnetic detection signal. Preferably, the sensor is a high-sensitivity, low-noise, and wide dynamic range tunnel magnetoresistive (TMR) sensor (e.g., TMR2905, TMR8533, etc.).

[0045] In this embodiment of the invention, the TMR sensor used is a single-axis TMR sensor (model TMR2905, sensitivity ≥45mV / V / Gs, noise density 2nT / ). Therefore, a three-dimensional magnetic field network is constructed using three single-axis TMR sensors arranged in a three-dimensional orthogonal configuration (@1Hz). The spacing between adjacent sensors is less than 0.5cm, and all connecting components between the sensors are made of non-magnetic materials to avoid introducing additional magnetic field interference. During operation, each single-axis TMR sensor outputs a differential signal (used to characterize the low-frequency magnetic source signal in the corresponding direction) and is connected one-to-one with the differential input channel of the signal acquisition board. The sensor power supply terminals (VDD, GND) are connected to the dedicated power supply module of the signal acquisition board, with an operating voltage of 5V. The core function of this three-dimensional detection network is to realize the three-dimensional magnetic signal detection at the observation point. The three sensors respectively capture the low-frequency magnetic source signal components in the x, y, and z axes, thereby providing complete and high-precision magnetic field data support for subsequent three-dimensional magnetic source detection and positioning algorithms. In other embodiments, an integrated three-dimensional TMR sensor chip can also be used, which can directly replace this discrete sensor network, simplifying the hardware architecture.

[0046] Correspondingly, in this embodiment of the invention, the signal acquisition board processes the differential signals output by each TMR sensor through differential signal to single-ended signal conversion, AC coupling, signal amplification, bandpass filtering, etc., to obtain a three-dimensional magnetic detection signal, thereby improving the signal-to-noise ratio of the original signal. In this embodiment of the invention, differential AC coupling filters out the DC drift component in the differential signals output by each TMR sensor through an RC network; the differential-to-single-ended circuit uses an instrumentation amplifier (instrumentation amplifier chip AD8421) to convert the differential signal into a single-ended signal to adapt to subsequent amplification and filtering circuits; the instrumentation amplifier and the composite amplifier circuit work together to provide controllable gain for weak signals, amplifying the signal amplitude to a processable range. In this embodiment of the invention, a series composite amplification architecture is used to achieve a signal gain of 86dB together with the instrumentation amplifier; the fourth-order bandpass filter and the pre-stage AC coupling circuit together constitute the bandpass filtering characteristics, realizing the screening of target frequency band signals, clutter suppression, and noise passband compression, ultimately providing a high signal-to-noise ratio and high stability input signal for the signal sampling and receiving module. In this embodiment of the invention, the fourth-order bandpass filter is implemented using a Sallen-Key configuration design, with a passband bandwidth of 140Hz±20%. The signal acquisition board has a built-in DC-DC power supply module that can stably provide ±5V operating voltage to power the onboard functional circuits and magnetic sensor array. At the signal connection level, the differential signals output by each TMR sensor chip in the onboard 3D magnetic detection array are converted into single-ended signals by the signal acquisition board, and then amplified and bandpass filtered. The 3D magnetic detection signals processed by the signal acquisition board are synchronously connected to different channels (AIN0~AIN5) of the ADC chip (signal sampling and receiving module). Since two observation points are set in this embodiment, the signal acquisition board of the two observation points outputs two 3D magnetic detection signals, for a total of six channels of signals. The structure of each circuit in the signal acquisition board is as follows: Figure 2 As shown, the low-frequency magnetic source detection module at each observation point is directly connected to the signal sampling and receiving module, outputting the processed signal to the signal sampling and receiving module to provide a high-quality signal source for subsequent signal sampling.

[0047] (3) Signal sampling and receiving module The core function of the signal sampling and receiving module is to sample and temporarily store the processed signals output by the low-frequency magnetic source detection modules at each observation point. The ADC chip performs signal sampling according to preset parameters, and the sampled signals are then transmitted to the internal storage resources of the FPGA processing module. One end of this module connects to the low-frequency magnetic source detection modules at each observation point to receive the processed magnetic source signals; the other end connects to the FPGA processing module, which controls the signals under unified FPGA control (including hardware configuration and logic control such as sampling frequency, sampling time, and oversampling rate), achieving accurate transmission and storage of the sampled signals.

[0048] In this embodiment of the invention, the signal sampling and receiving module uses an 8-channel 16-bit parallel synchronous sampling ADC chip AD7606, which has a maximum throughput of 200kSPS, an analog input voltage range covering ±5V, and a built-in analog front-end buffer, effectively improving signal driving capability and sampling stability. At the hardware connection level, the ADC chip's six analog signal input pins (AIN0~AIN5) are connected one-to-one with the x, y, and z axis signal output terminals of the two low-frequency magnetic source detection modules at the two observation points to receive the processed analog signals (three-dimensional magnetic detection signals output by the signal acquisition board) from the corresponding low-frequency magnetic source detection modules. The chip's control pins (CS, RD, CONVST, etc.) are logic-controlled through the FPGA's GPIO interface, and the data output pins (DB0~DB15) are directly connected to the FPGA's receiving end. The chip's power supply is connected to a 5V DC regulated power supply. The core function of this ADC chip is high-precision conversion from analog to digital signals: based on the control instructions issued by the FPGA, it performs sampling according to the configuration parameters (4096Hz sampling frequency, 0.5s / frame continuous sampling duration, 4 times oversampling rate, etc.), quantizes the continuous analog signal output by the acquisition board into a digital signal, and transmits it to the FPGA on-chip storage in real time, providing high-fidelity data input for subsequent digital signal processing such as lock-in amplification and coordinate calculation.

[0049] (4) FPGA processing module The FPGA processing module, as the core control and computing unit of the system, performs core functions of processing and calculating the coordinates of the sampled signals. Specifically, it performs frequency detection and phase-locked amplification (extracting the phase and amplitude of the magnetic source signal and further improving the signal-to-noise ratio through phase-locked narrowband filtering) on ​​the sampled signals output from the signal sampling and receiving module. It also performs three-dimensional magnetic source signal amplitude calculation based on two-point observations and relative coordinate calculation of the low-frequency magnetic source target (using the observation point as a reference point). This module is bidirectionally connected to the signal sampling and receiving module, controlling the sampling parameters and logic of the ADC chip (signal sampling and receiving module) on one hand, and receiving and storing the sampled signals on the other. Simultaneously, it provides core computational support for the entire system and is a key module connecting signal processing and coordinate calculation.

[0050] In this embodiment of the invention, the Xilinx Artix-7 series XC7A100TFGG484-2 FPGA chip is used as the core control and computing module. By integrating and deploying an ADC control unit, FFT unit, high-precision frequency fitting unit, phase-locked loop preprocessing unit, phase-locked loop amplification unit (including phase-locked loop amplification and orthogonal detection), phase calculation unit, and two-point coordinate calculation unit, the core positioning function of the magnetic source target is realized. In this embodiment of the invention, a mathematical model of the low-frequency rotating magnetic source is established to model the magnetic source distribution (characterizing the amplitude information of the magnetic source at different positions in space) generated by the low-frequency magnetic source generation module. A specially designed double orthogonal detection phase-locked loop amplification algorithm adapted to three-dimensional magnetic signals is implemented, and the entire algorithm is mapped to FPGA hardware logic implementation, ensuring the real-time performance and high accuracy of signal calculation from the architectural level. The FPGA chip operates at a frequency of 50MHz. It configures and controls the timing of the ADC chip through the GPIO interface, and can flexibly set operating parameters such as sampling frequency and number of sampling points. After positioning is completed, the target coordinate results (magnetic source spatial position coordinates, i.e., the current position of the UAV) can be transmitted to the back-end device through a dedicated output port, or directly drive the digital tube or display screen for real-time display.

[0051] like Figure 4 As shown, the FPGA adopts a modular design to form an integrated in-memory computing architecture of "data reception - feature processing - phase-locked amplification - coordinate calculation". The functions of each unit and the data flow logic are as follows: First, the ADC control unit is responsible for initializing and configuring the ADC hardware (signal sampling and receiving module) and driving control. It also manages the reception, synchronous buffering and storage logic of the sampled data, and stores the six-channel magnetic detection sampled data from the two observation points into the on-chip block random access memory (BRAM_CH0~BRAM_CH5) in sequence according to the channel.

[0052] Second, the FFT unit extracts the magnetic detection data stored in one channel and performs spectrum analysis using Fast Fourier Transform to initially obtain the characteristic frequency (low-resolution characteristic frequency) of the target low-frequency magnetic signal. Preferably, considering that the planar signal is stronger, the FFT unit extracts the magnetic detection sampling data corresponding to the x-direction or y-direction for spectrum analysis. In this embodiment of the invention, the FFT unit extracts the stored data of the main channel CH0 (magnetic detection sampling data corresponding to the x-direction) of the main observation point for spectrum analysis to initially obtain the characteristic frequency of the target low-frequency magnetic signal.

[0053] Third, the high-precision frequency fitting unit optimizes the characteristic frequencies calculated by the FFT unit to obtain the characteristic frequencies of the high-resolution target magnetic source signal after fitting. This further improves the frequency detection accuracy and generates a high-precision reference frequency signal for the lock-in amplifier unit.

[0054] Fourth, the phase-locked amplification process includes a phase-locked preprocessing unit, a phase-locked amplification unit, and a phase calculation unit. The phase-locked preprocessing unit, based on the fitted target frequency, completes the phase-locked control logic configuration and outputs the input data and reference signal required by the phase-locked amplification unit. Specifically, the phase-locked preprocessing unit uses the fitted high-resolution target frequency as the frequency of the reference signal in the phase-locked amplification and determines the integration time in the phase-locked amplification based on the fitted high-resolution target frequency. Simultaneously, the frequency and integration time of the reference signal are provided to the lock-in amplifier unit. Based on the frequency of the reference signal and the modeled low-frequency magnetic source distribution, the lock-in amplifier unit performs the first lock-in amplification operation and quadrature detection on the magnetic detection data of each channel, demodulating the parameters characterizing the amplitude and phase of the magnetic source signal. These parameters are then provided to the phase calculation unit for phase calculation to obtain the phase of the low-frequency magnetic source signal. Phase of the magnetic source signal The phase-locked loop (PLL) preprocessing unit is provided with the phase of the magnetic source signal. The phase, frequency, and integration time of the reference signal are provided to the lock-in amplifier unit. Based on the frequency and phase of the reference signal and the modeled low-frequency magnetic source distribution, the lock-in amplifier unit performs a second lock-in amplification operation and quadrature detection on the magnetic detection data of each channel, demodulating the amplitude information of the low-frequency magnetic source signal. In the first phase-locked amplification, the signal of the magnetic source under test is only in the same frequency as the reference signal. In the second phase-locked amplification, the signal of the magnetic source under test and the reference signal are in the same frequency and phase.

[0055] Fifth, the two-point coordinate calculation unit calls the three-dimensional magnetic source positioning algorithm to complete the calculation and solve based on the amplitude information of the low-frequency magnetic source signal output by the second orthogonal detector, and obtains the accurate spatial coordinates of the low-frequency magnetic source.

[0056] To further understand the inventive point, the magnetic source coordinate calculation algorithm based on lock-in amplification and its FPGA hardware acceleration implementation provided by this invention will be further explained.

[0057] (1) Modeling of low-frequency magnetic source field distribution: like Figure 5 As shown, the location of the magnetic source is The observation point is The low-frequency magnetic source generating module can be approximated as a magnetic dipole, equivalent to a three-dimensional magnetic moment concentrated at a single point. The distribution of the magnetic field, , , Three-dimensional magnetic moments The components along the x, y, and z axes. According to the Biot-Savart law, the three-dimensional magnetic field distribution of a magnet. follow: , The permeability of free space, For observation point To the detection point The relative distance, Let be the relative distance vector from the observation point to the detection point. Therefore, when the magnet at the magnetic source rotates along the z-axis in the horizontal plane, its magnetic field distribution in space satisfies: ; in, , , These represent the components of the low-frequency magnetic source signal along the three axes. , , These represent the magnitudes of the magnetic moment in the x, y, and z directions, respectively.

[0058] again , Therefore, it can be , , Transformed into: = ; in, Magnetic moment component and The vector sum, The characteristic frequencies of the fitted, high-resolution target magnetic source signal output by the high-precision frequency fitting unit are given. For time, The phase of the low-frequency magnetic source signal (determined by the phase calculation unit mentioned above).

[0059] remember ; Then there is .

[0060] At the current moment, the relative positions of the observation point and the magnetic source can be considered fixed, that is... Constant, in its physical sense, refers to the magnetic moment of the magnetic source and the observation point. To the detection point The numerical representation of relative distance, which indicates the amplitude information of the low-frequency magnetic source, will be simplified by using... Recorded as: ; ; ; ; The detected magnetic source signal , , Therefore The amplitude is and the frequency is . Phase is The sinusoidal signal.

[0061] (2) Lock-in amplification principle and its improvement for magnetic source detection By acquiring and filtering out interference signals (such as DC, signal harmonics, high-frequency magnetic signals, etc.), constructing a 140Hz±20% bandpass, and amplifying the signal to a processable range, the signal actually acquired by the low-frequency magnetic source detection module is transformed into: ; ; ; in, For signal gain, , , For the corresponding noise term; , , Related to the DC component, after the DC component is filtered out, , , It is 0.

[0062] The frequency detection value of the magnetic source signal can be obtained through FFT and high-precision frequency fitting. The general principle of a lock-in amplifier design is to introduce a reference signal of the same frequency and integrate the original signal to achieve narrowband filtering, thereby demodulating the amplitude and phase of the original signal. However, unlike the general case, low-frequency magnetic source signals ( , , ( ) is the superposition of two signals of the same frequency and phase with different amplitudes. Therefore, this invention makes adjustments to the general lock-in amplification principle to be suitable for low-frequency magnetic source signals. The main idea is as follows: through the orthogonal detection operation of the first lock-in amplification, information about four variables ( The four equations (first time targeting) and (Perform orthogonal detection), the phase calculation unit uses mathematical techniques to first solve for the phase of the low-frequency magnetic source signal. Then, the phase shift of the reference signal is adjusted to... And perform a second phase-locked amplification orthogonal detection operation (the second operation simultaneously...) , and (Perform quadrature detection). At this time, the reference signal and the magnetic source signal are in phase, and the lock-in amplification result is only related to the amplitude of the magnetic source signal. Related, that is The demodulated component is positively correlated with the second orthogonal detector; at this point, the amplitude and phase of the magnetic source signal have been calculated. In this embodiment of the invention, the phase-locked amplification design has been improved for applicability to the signal characteristics of the magnetic source signal, and the calculation process uses phase-locking to convert the amplitudes of the two signals that synthesize the low-frequency magnetic source signal ( and , and , and The separation (in the prior art, it is the sum of squares of two parameters) avoids complex mathematical calculations.

[0063] In this embodiment of the invention, the quadrature detection of the first phase-locked amplification:

[0064] in, Indicates adoption When used as a reference signal (t) Perform detection operation. The amplitude of the reference signal is a fixed value, and the narrowband noise n(t) is expressed at the center frequency. It can be decomposed into: ; in, and Noise and The component in the x-direction. After passing through the LPF (implemented by a digital integrator), only... , , , Partial retention; the narrower the equivalent bandwidth of LPF, the better. The stronger the suppression of even a tiny portion of noise, the greater the improvement in signal-to-noise ratio. Ignoring the noise term after passing through the LPF, we get: ; in, The integration time in lock-in amplification is determined based on the fitted high-resolution target frequency.

[0065] Similarly, we can conclude that: ; ; ; again , , ; in, Indicates adoption For reference signal pair Perform detection operation. , They respectively represent the use of , For reference signal pair Perform detection operation.

[0066] The second phase-locked amplification quadrature detection: With phase α, a second quadrature detection demodulation can be performed to extract the amplitude, similarly using... For example ( , , , , Similarly): ; Similarly: ; ; ; ; ; in, , , They respectively represent the use of For reference signal pair (t), , Perform detection operation; , , They respectively represent the use of For reference signal pair , , Perform detection operation.

[0067] Lock-in amplification (LPA) can achieve narrowband filtering, greatly improving the signal-to-noise ratio (SNR). The magnitude of the SNR improvement is related to the integration time. The integration time is inversely proportional to the LPF bandwidth; the longer the integration time (or the larger the number of integration points), the narrower the bandwidth and the stronger the noise suppression capability. Furthermore, LPA can directly solve for the phase and amplitude of the magnetic source signal, performing two-point coordinate calculations.

[0068] (3) FPGA implementation of lock-in amplification The core functional modules of the FPGA implementation of lock-in amplification include an adjustable phase reference signal generator (used to generate reference signals for the two lock-in amplification operations and quadrature detection), a digital integrator (used for the implementation of the above two lock-in amplification and quadrature detection, i.e., the implementation of the integration formula, outputting the amplitude A of the low-frequency magnetic source), and an arctan calculation module (i.e., the phase calculation unit, used to implement...). The calculations output the phase of the low-frequency magnetic source. The functions, quantization parameters, and technical implementation details of each module are as follows: Adjustable phase reference wave generator: Its core function is to generate four reference signals with the same frequency as the magnetic source signal, namely... , The adjustable phase reference wave generator internally stores 1 / 4 cycle of sin / cos wave data (preferably 1 / 4 cycle to save FPGA on-chip ROM resources). The sin / cos wave data has a precision of 14-bit signed numbers and is stored in the FPGA on-chip ROM resources. The ROM storage depth is 1024×14bit, which meets the reference wave precision requirements. A 32-bit phase accumulator is used. During the first phase-locked loop (PLL) control, the phase control word of the phase accumulator is 0, and the frequency detection value is used... The loop steps are calculated based on the FPGA system clock frequency (50MHz) and the target reference signal frequency (120~160Hz), and the resulting frequency is... A reference signal with a phase of 0 ( , (The phase is to be obtained by the arctan solution module) After that, with The phase control word for the phase accumulator generates the phase as... Two reference signals The reference signal amplitude Vr is a 12-bit signed number with an amplitude range of 0~4095 LSB, ensuring that the reference signal amplitude is stable and meets the accuracy requirements.

[0069] Digital integrator: Its core function is to perform integration operations between the magnetic source signal and the reference signal. A typical calculation formula is as follows: = The magnetic source signal needs to be completed. With reference signal , Summation of integrals, etc. This module is categorized by "value selection" (e.g., selecting...). and reference signal - Quadrature (calculation) and (product) - summation ( The three-stage pipeline design (integration) has the pipeline clock frequency consistent with the FPGA system clock (50MHz), the maximum number of integration points is 2048, and the integration time is the number of integration points × clock cycle, that is, 2048 × 20ns = 40.96μs, which is significantly shortened compared with traditional non-pipelined integrators, effectively improving the real-time performance of signal processing.

[0070] Arctan calculation module: Its core function is to perform arctan calculations using the formula... Phase calculation This module implements the arctan algorithm using the CORDIC algorithm. The algorithm iterates 16 times with an iteration clock frequency of 50MHz, and the single solution time is ≤320ns, ensuring a fast solution speed. The phase results are then processed. Scaling to the [-1024, 1023) interval, this interval is precisely mapped to the [-π, π) phase interval, with a mapping ratio of 1024 / π, that is, 1 LSB corresponds to π / 1024 rad (approximately 0.176°), and the phase calculation accuracy is ≤0.2°, which meets the phase calculation requirements for magnetic source positioning.

[0071] In this embodiment of the invention, the core value of the FPGA processing module lies in two aspects: First, by strictly adhering to the magnetic field distribution model of the magnetic dipole and the sinusoidal superposition characteristics of the magnetic signal, the dedicated lock-in amplification algorithm is implemented in hardware, ensuring a high degree of matching between the algorithm and the physical characteristics of the magnetic source, thus significantly improving the accuracy of amplitude and phase calculations. Second, by ensuring the stability and synchronization of ADC sampling through hardware-based timing control, and relying on the FPGA parallel computing architecture, high-speed feature processing, lock-in amplification, and coordinate calculation of the sampled data are achieved, ultimately achieving microsecond-level response speed and low-power hardware acceleration, providing core support for the real-time and accurate positioning of the magnetic source target.

[0072] The performance of the detection and positioning system of the present invention will be further illustrated below with specific examples.

[0073] Example 1: (1) Test environment: The test site is an indoor enclosed space with dimensions of 20m×20m×5m (length×width×height). A moderately obstructed working environment is set up, with obstructions including wall corners, metal / wooden cabinets and experimental instruments, etc., with an obstruction rate of ≥60%, simulating the complex environment in actual applications.

[0074] (2) Hardware configuration: The magnetic source simulation uses a small motor equipped with neodymium iron boron permanent magnets to replace the actual UAV, with a working frequency range of 50~150Hz; the detection and positioning system includes a low frequency magnetic source detection module, a signal sampling and receiving module and an FPGA (FPGA chip model xc7a100tfgg484-2) processing module. The ADC sampling rate in the signal sampling and receiving module is set to 4096Hz, and 4 times oversampling technology is used to improve the signal sampling accuracy.

[0075] (3) Test procedure: In this test, the magnetic source was moved along the x, y, and z axes, and the detection signal data of the x-axis sensor at the main observation point was recorded. The signal amplitude distribution was analyzed to determine whether it met the requirements of the magnetic source signal modeling. At the same time, at a fixed distance of 2m, angle tests were conducted on the xoy, xoz, and yoz planes respectively. The signal amplitude changes caused by the angle changes under different planes were recorded to further verify the correctness of the magnetic source signal modeling and the high-quality sampling of the signal sampling and receiving module.

[0076] (4) Test results: The results are respectively from Figure 6 , Figure 7 As shown, Figure 6 The characteristics of amplitude and signal-to-noise ratio of the sensor detection signal at the main observation point x-axis in magnetic source detection are presented. The signal characteristics are cubic attenuation, which is consistent with the attenuation law of theoretical modeling. Moreover, the signal amplitude in the three directions meets the 2 times numerical relationship of theoretical modeling. This shows that the signal sampling and receiving module can effectively capture magnetic source signals within 10m, providing reliable data support for subsequent signal amplitude calculation and coordinate calculation. Figure 7 The characteristics of the amplitude of the signal detected by the x-axis sensor at the main observation point in magnetic source detection are presented. The amplitude of the signal varies with the angle at different planes at a fixed distance of 2m. The amplitude of the signal varies with the angle and exhibits a sinusoidal distribution, which satisfies the theoretical modeling. This shows that the signal sampling and receiving module can effectively capture magnetic source signals of all directions within the range.

[0077] Example 2: (1) Test environment: In the laboratory test environment, the DDS signal generator is used to generate test signals to verify the phase extraction and amplitude extraction functions of the FPGA-based lock-in amplifier module in the embodiment of the present invention.

[0078] (2) Hardware configuration: DDS signal generator, which can simulate and generate actual detection magnetic source signals; FPGA (FPGA chip model xc7a100tfgg484-2) processing module, wherein the FPGA processing module deploys the positioning solution algorithm with lock-in amplification function designed in this embodiment of the invention.

[0079] (3) Test procedure: Use DDS to simulate the actual magnetic source signal, change the initial phase and amplitude of the signal respectively, and record the FPGA phase-locked output result; use DDS to simulate the actual magnetic source signal and add an adjustable noise signal, fix the magnetic source signal amplitude and change the noise power, and record the FPGA phase-locked amplification output signal signal-to-noise ratio under different types of noise and different input signal signal-to-noise ratios.

[0080] (4) Test Results: The tests verified the phase and amplitude extraction capabilities of the lock-in amplifier, as well as its ability to improve the signal-to-noise ratio (SNR) within a narrow bandwidth. The FPGA lock-in amplifier design can achieve high-quality extraction of the phase and amplitude of the target signal and further improve the SNR of the input signal. The data from the three experiments are as follows: Figures 8-10 As shown. It can be seen that, Figure 8This paper presents the signal-to-noise ratio (SNR) improvement characteristics of the FPGA-based lock-in amplifier design in this invention under uniformly distributed noise and Gaussian distributed noise. The results in the figure quantify the performance boundaries of the module under two typical noise environments: First, it verifies the SNR enhancement capability of the lock-in amplifier design for the input signal, clarifying that it can effectively improve signal quality in magnetic source detection scenarios (input SNR ≥ 0dB within the detection distance); Second, this lock-in amplifier design can enhance the effective signal characteristics masked by noise, providing reliable signal support for high-precision implementation of signal amplitude calculation and magnetic source coordinate calculation, and has direct engineering practical value.

[0081] Figure 9 This invention demonstrates the error characteristics of signal phase calculation in an FPGA-based lock-in amplifier design: with the input signal phase (0~2...) Using rad as a variable, the relative error (left axis, %) and absolute error (right axis, mrad) of the solution results are presented simultaneously, indicating that the overall phase solution error of the lock-in amplification is controlled within 0.4%, and the relative error of most phase intervals is less than 0.1% and the absolute error is less than 2%. The mrad data not only verified the stability of the phase calculation of the lock-in amplifier design, but also demonstrated that it can provide high-precision phase data support for the accurate calculation of signal amplitude in magnetic source detection scenarios.

[0082] Figure 10 This paper presents the error characteristics of the FPGA-based lock-in amplifier design in signal amplitude calculation in this embodiment of the invention, with the input signal amplitude tested ranging from 5mV to 2.5V. The results show that the absolute error of the amplitude calculation is related to the input amplitude, exhibiting a peak in the range of larger input amplitudes, but the overall absolute error is controlled within 2mV. The absolute error in the low input amplitude range is less than 0.5mV, while the relative error increases as the input amplitude decreases (relative error >5% when input amplitude <10mV), and the relative error is <2% in most ranges where the input amplitude >10mV. These results verify the stability of the amplitude calculation of this lock-in amplifier design, indicating that it can provide high-precision signal amplitude calculation results for magnetic source detection.

[0083] like Figure 11 As shown, in this embodiment of the invention, the magnetic source positioning method proposed in this embodiment of the invention is also compared with the key performance of commonly used low-altitude positioning technologies (radar positioning, photoelectric / visual positioning, acoustic positioning). It can be seen that the magnetic source positioning of the present invention has the advantages of anti-obstruction, low latency, high accuracy and high integration in low-altitude short-range detection and positioning compared with other positioning technologies.

[0084] Example 3: (1) Test environment: The FPGA-based low-frequency magnetic source detection and positioning system of the present invention was systematically positioned and tested in an outdoor environment.

[0085] (2) Hardware configuration: The FPGA-based low-frequency magnetic source detection and positioning system in this embodiment of the invention.

[0086] (3) Test process: The algorithm of this invention is used to locate the target magnetic source at different heights and positions. The actual coordinates and the FPGA calculated coordinates are recorded and the error is analyzed.

[0087] (4) Test results: Fifteen sets of positioning data were tested on each of the six horizontal planes at heights Z=1m, 3m, 5m, 7m, 9m, and 10m for the low-frequency magnetic source generating module. The x and y axes increased from 2m to 11m in increments of 3m. The positions of each set of positioning data were denoted as (X, Y). The actual coordinates of each set of positioning data at different heights Z and the coordinates calculated using the FPGA-based low-frequency magnetic source detection and positioning system in this embodiment of the invention are shown below. Figures 12-17 As shown, it can be seen that the coordinates calculated by the FPGA-based low-frequency magnetic source detection and positioning system in this embodiment of the invention are basically consistent with the actual coordinates.

[0088] like Figure 18 As shown, under the conditions of relatively close distances of Z=1m and 3m, the relative error of three-dimensional positioning is less than 1.5%; as the detection distance increases to Z=5m and above, the positioning error increases, but it is still controlled within 6% in the effective detection area.

[0089] At different heights Z, the positioning error of the test point sequence (X,Y) was tested in the order of (5,2) → (8,2) → (11,2) → (2,5) → (5,5) → (8,5) → (11,5) → (2,8) → (5,8) → (8,8) → (11,8) → (2,11) → (5,11) → (8,11) → (11,11). The relative error distribution of each point was obtained as follows: Figure 19 As shown, it can be seen that within the test area where X≥2 m and Y≥5 m, the relative error of the three-dimensional positioning of all points is <6%.

[0090] Based on the above experimental test data and the inherent technical advantages of magnetic detection positioning, it can be concluded that the low-altitude detection technology in this embodiment of the invention significantly improves its anti-obstruction and anti-interference capabilities. The low-frequency magnetic signal of 140Hz (±20%) possesses strong diffraction and penetration properties, capable of penetrating obstructions such as urban building clusters, trees, and glass curtain walls. It can still operate stably in complex environments with obstruction rates ≥60% (dense buildings, underground parking garages, indoor spaces), solving problems associated with other detection methods such as radar signal loss at low altitudes, environmental influences on photoelectric detection, and severe obstruction of radio signals. Phase-locked amplification processing enables narrowband filtering (bandwidth ≤1Hz), effectively extracting weak magnetic source signals and suppressing background noise and environmental electromagnetic interference, improving the data accuracy and stability of detection and positioning, with no drift or jumps in the positioning results. Furthermore, the detection and positioning accuracy and real-time performance of this embodiment of the invention reach a high level, with a relative error of <6% within the target range for three-dimensional positioning, superior to other low-altitude detection technologies, meeting the high-precision requirements of UAV low-altitude operations. The FPGA parallel processing architecture enables hardware acceleration, reducing the need for serial operation by 80... Under a pipelined architecture, the positioning response time can be further reduced to 40 ms. Superior to CPU-based software processing solutions (response time ≥ 1ms), this invention, as a standalone hardware device, can support independent and long-term operation of UAVs. Finally, the miniaturization, low power consumption, and wide frequency band characteristics of this invention are well-suited to the application scenarios of UAVs. The integrated core hardware (sensor + acquisition board, ADC, FPGA) has a volume ≤ 10cm × 10cm × 2cm and a weight ≤ 300g, allowing direct deployment at ground observation points or other carriers. Compared to LiDAR detection and positioning systems (typically ≥ 20cm × 15cm × 10cm, weight ≥ 1kg), its miniaturization advantage is significant. The core component, the FPGA, consumes ≤ 0.5W, far lower than other detection and positioning technologies. The operating frequency band is adjustable, allowing for configuration and switching of the operating frequency band according to the actual UAV model for accurate positioning. It also has the ability to migrate to other frequency bands, adapting to positioning needs under different hardware configurations, and exhibits superior compatibility compared to fixed-frequency detection and positioning technologies.

[0091] Furthermore, the hardware cost of this invention is low, with core components selected from low-cost industrial-grade products (such as TMR sensors, AD7606 ADC chips, and Xilinx Artix-7 series FPGA chips). This results in lower overall hardware costs, which can be further reduced during mass production. Compared to LiDAR detection and positioning systems and INS / vision / UWB fusion systems, it offers a price advantage for large-scale application. Secondly, deployment and maintenance costs are low, eliminating the need for complex base stations (such as UWB base stations or GNSS differential base stations). Only low-frequency magnetic source detection modules need to be deployed at ground observation points, resulting in a short deployment cycle. The core chip components have high reliability, with no mechanically worn parts, leading to low maintenance costs.

[0092] This invention provides a high-precision, low-cost technical solution for low-altitude UAV positioning. In future air traffic, this solution can provide high-precision short-range detection to solve takeoff and landing deviation problems, ensuring safe and efficient air traffic. It can also fill the positioning gaps in indoor and underground logistics hubs, achieve seamless connection between air vehicles and indoor docking points, provide accurate positioning guidance for UAVs in complex environments, or serve as a UAV identification method to prevent unauthorized "black flight" behavior by other non-compliant UAVs, becoming an important technological complement to the future three-dimensional, intelligent air traffic network. Furthermore, the unique technical advantages of this invention can complement other technologies, serving as a precise detection and positioning method for the last 10 meters in multi-source fusion detection technology.

[0093] Those skilled in the art will readily understand that the above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A low-frequency magnetic source detection and positioning system based on FPGA, characterized in that, include: Low-frequency magnetic source generation module, used to generate target low-frequency magnetic source signal; A low-frequency magnetic source detection module is set at at least two observation points on the ground to detect the target low-frequency magnetic source signal and obtain the three-dimensional magnetic detection signal detected at each observation point. The signal sampling and receiving module is used to sample the three-dimensional magnetic detection signal by channel to obtain magnetic detection data for each channel; The FPGA processing module is used to receive and store the sampled magnetic detection data, and extract the magnetic detection data from at least one channel for spectrum analysis to obtain the frequency of the target low-frequency magnetic source signal. The magnetic detection data from each channel undergoes a first phase-locked amplification and quadrature detection to demodulate and obtain the phase of the target low-frequency magnetic source signal. The frequency of the reference signal used in the first phase-locked amplification operation and quadrature detection is... ; the frequency and the phase The frequency and phase of the reference signal used for the second phase-locked amplification and quadrature detection are used to perform the second phase-locked amplification and quadrature detection on the magnetic detection data of each channel, and the amplitude of the target low-frequency magnetic source signal is obtained by demodulation. Based on the amplitude, a three-dimensional magnetic source localization algorithm is used to obtain the spatial coordinates of the target low-frequency magnetic source signal, thereby realizing the localization of the low-frequency magnetic source.

2. The FPGA-based low-frequency magnetic source detection and positioning system according to claim 1, characterized in that, The phase of the target low-frequency magnetic source signal for: ，α [- ， ) in, , They respectively represent the use of , As a reference signal pair (t) The result of the first orthogonal detection. The amplitude of the reference signal, For magnetic detection data of one channel, characterization The x-axis component of the low-frequency magnetic source signal detected from the target at any given time; , They respectively represent the use of , As a reference signal pair The results of orthogonal detection For magnetic detection data of one channel, characterization The component of the low-frequency magnetic source signal detected at any time on the y-axis.

3. The FPGA-based low-frequency magnetic source detection and positioning system according to claim 2, characterized in that, The amplitude of the target low-frequency magnetic source signal for: in ; ; ; ； ； ； In the formula, , , They respectively represent the use of As a reference signal pair , , The results of the second orthogonal detection; , , They respectively represent the use of As a reference signal pair , , The results of orthogonal detection; For magnetic detection data of one channel, characterization The z-axis component of the low-frequency magnetic source signal detected at any time.

4. The FPGA-based low-frequency magnetic source detection and positioning system according to claim 3, characterized in that, In the FPGA processing module, an adjustable phase reference signal generator is used to generate the reference signal; wherein, the adjustable phase reference signal generator internally stores a single 1 / 4-cycle sin and cos signal data; the adjustable phase reference signal generator includes a phase accumulator, and during the first phase-locked amplification operation and quadrature detection, the phase control word of the phase accumulator is 0, and it is adjusted according to frequency. The cycle is repeated to generate a frequency. Reference signal with phase 0 and During the second phase-locked loop amplification operation and quadrature detection, the phase accumulator... For phase control words, by frequency The cycle is repeated to generate a frequency. Phase is Reference signal and .

5. The FPGA-based low-frequency magnetic source detection and positioning system according to claim 3, characterized in that, In the FPGA processing module, a digital integrator is used to perform the integration of magnetic detection data and reference signal of each channel in the first and second phase-locked amplification operations. The digital integrator employs a three-stage pipeline of value acquisition, integration, and accumulation to achieve parallel execution of the integration operation. The value acquisition involves acquiring the magnetic detection data of the corresponding channel and its corresponding reference signal; the integration involves multiplying the magnetic detection data with its corresponding reference signal; and the accumulation involves summing the results of the integration operation to achieve the integration operation.

6. The FPGA-based low-frequency magnetic source detection and positioning system according to claim 2, characterized in that, In the FPGA processing module, an arctan calculation module is used to realize the phase of the target low-frequency magnetic source signal. The solution is obtained by using the CORDIC algorithm in the arctan solution module. The solution.

7. The FPGA-based low-frequency magnetic source detection and positioning system according to any one of claims 1-6, characterized in that, Magnetic detection data from at least one channel are extracted and subjected to spectral analysis to obtain the frequency of the target low-frequency magnetic source signal. ,include: At least one channel of magnetic detection data is extracted and subjected to spectral analysis using Fast Fourier Transform to initially obtain the frequency of the target low-frequency magnetic source signal. The initially obtained frequency is then fitted and optimized to obtain the final frequency of the target low-frequency magnetic source signal. .

8. The FPGA-based low-frequency magnetic source detection and positioning system according to any one of claims 1-6, characterized in that, The low-frequency magnetic source generating module is one or more spliced ​​neodymium iron boron permanent magnets rotating in a certain plane, or it is an energized coil; The low-frequency magnetic source detection module includes a sensor and a signal acquisition board; the sensor is used to acquire the components of the target low-frequency magnetic source signal in the x, y, and z axes; the signal acquisition board is used to process the target low-frequency magnetic source signal components in the three axes to obtain the three-dimensional magnetic detection signal; wherein, the processing includes one or more of AC coupling, signal amplification, and bandpass filtering; The main component of the signal sampling and receiving module is an ADC chip; the sampling parameters and sampling logic of the ADC chip are controlled by the FPGA processing module.

9. The FPGA-based low-frequency magnetic source detection and positioning system according to claim 8, characterized in that, The sensor is a TMR sensor; the TMR sensor is composed of three single-axis TMR sensors arranged in a three-dimensional orthogonal manner, or the TMR sensor is an integrated three-dimensional TMR sensor chip.

10. A low-altitude magnetic source detection and positioning system for unmanned aerial vehicles (UAVs), characterized in that, The magnetic source detection and positioning system is the FPGA-based low-frequency magnetic source detection and positioning system according to any one of claims 1-9; wherein, the low-frequency magnetic source generating module is mounted on the UAV and uses the spatial coordinates of the obtained target low-frequency magnetic source signal as the position information of the UAV to achieve low-altitude positioning of the UAV.

Images