Unmanned aerial vehicle particle field uniformity detection device and method

Abstract

The application discloses a kind of unmanned aerial vehicle granular field uniformity detection device and detection method, belong to agricultural engineering and farmland detection technical field.The device includes detection device body, optical fiber sensing module, embedded PCB circuit board.The detection method is laid out and horizontally calibrated device, optical fiber sensing module collects the light signal variation of granular interruption detection beam, signal processing, effective counting obtains the number of particles per unit time, combined with single particle average mass conversion real-time spreading amount;Host computer collects multiple sampling point data to calculate the variation coefficient of spreading amount, compares preset threshold to determine uniformity grade, and accordingly guides unmanned aerial vehicle operation parameter adjustment.The application realizes real-time accurate detection of spreading uniformity, has strong environmental adaptability, is simple and easy to operate, effectively improves the intelligentization and high-efficiency level of field spreading monitoring, and is suitable for various agricultural unmanned aerial vehicle granular spreading operation scene.

Description

Technical Field

[0001] This invention belongs to the field of agricultural engineering and farmland testing technology, specifically relating to a device and method for detecting the uniformity of unmanned aerial vehicle (UAV) granule application in the field. Background Technology

[0002] In most field pellet spreading operations, the uniformity of pellet distribution is a key indicator for evaluating the scientific validity of drone operation parameters. During drone spreading, the pellets are easily unevenly distributed due to factors such as airflow disturbance, projection speed, and rotor airflow, resulting in strip accumulation or blank areas. Consequently, drone spreading operations may fail to meet the requirements of agricultural production.

[0003] Current methods for detecting the uniformity of particle distribution in the field mainly rely on manual weighing in the field or laboratory simulation analysis. Manual weighing in the field typically involves setting up multiple static collectors, sampling and weighing each collector after particle application, and then calculating the coefficient of variation (CV) of particle quantity at each measuring point to assess distribution uniformity. This method is cumbersome, labor-intensive, has a long data acquisition cycle, and low accuracy. While laboratory simulation analysis offers higher accuracy, the testing environment is idealized and fails to reflect real field conditions. Furthermore, the equipment is expensive and lacks mobility, making it unsuitable for rapid testing. For example, patent CN119935604A discloses an agricultural drone material application performance testing system and its operating method. This system obtains application distribution data by setting up collection trays and weighing the particles. This solution enables quantitative analysis of the drone application effect, but its device is large and complex to set up. The testing process must be completed offline after the operation, making real-time monitoring impossible. It is only suitable for indoor or windless conditions and cannot achieve field measurement. Summary of the Invention

[0004] To address the technical problems existing in the background art, this invention proposes a field uniformity detection device and method for unmanned aerial vehicle (UAV) granule application.

[0005] This invention adopts the following technical solution: A field uniformity detection device for unmanned aerial vehicle (UAV) granule application includes: The detection device body has several adjustable feet at its bottom; the upper surface of the detection device body is recessed from top to bottom to form a particle receiving cavity of a preset volume. An optical fiber sensing module is installed in the detection device body as required; the optical fiber sensing module includes: an optical fiber transmitter and an optical fiber receiver arranged opposite to each other, and a detection beam that transversely penetrates the entrance of the particle receiving cavity is formed between the optical fiber transmitter and the optical fiber receiver. The PCB circuit board is embedded in the detection device body; the PCB circuit board is also electrically connected to the fiber optic sensing module.

[0006] In a further embodiment, the adjustable leg includes: a telescopic bracket and a horizontal adjustment rod; The bottom of the telescopic support is used to abut against the field ground, the top of the telescopic support is connected to the bottom of the detection device body, and the horizontal adjustment rod passes through the telescopic support and is threaded into the telescopic support.

[0007] In a further embodiment, it also includes: a shock-absorbing material, laid at the bottom of the particle receiving cavity; the shock-absorbing material is used to absorb the impact energy of the falling particles when the drone is scattering them and to suppress particle rebound.

[0008] In a further embodiment, it further includes: a level, disposed on the upper surface of the detection device body; the level is used to monitor the horizontal state of the detection device body; A visual display screen is located on the outer side of the detection device body.

[0009] The method for detecting the uniformity of field application of applied granules using the drone-based granule application uniformity detection device described above includes the following steps: Based on field parameter information and preset application width, determine the number and location coordinates of sampling points, and deploy uniformity detection devices accordingly; perform the following procedure for each set of uniformity detection devices: Based on the terrain undulation corresponding to the sampling point, the uniformity detection device is calibrated to ensure that the particle containment chamber is set horizontally. The drone operates according to the current parameters. When conducting field pellet spreading operations, the pellets enter the pellet receiving cavity under the action of gravity and pass through the detection beam of the fiber optic sensing module during the falling process. The detection beam formed by the fiber optic transmitter and receiver is momentarily blocked by a particle, causing continuous fluctuations in the light intensity received by the fiber optic receiver. This results in a continuous change in the optical signal that matches the frequency and velocity of the particle's passage. ; For the change in the continuous optical signal Perform signal shaping, filtering and noise reduction, and baseline correction preprocessing to extract effective blockage signal features; Based on the characteristics of effective blocking signals, blocking events are identified, and the effective counting of particle passage events is completed to obtain the unit time. The number of particles detected by the beam inside And calculate the real-time particle application rate at the corresponding sampling point. .

[0010] In a further embodiment, the following steps are also included: Summarize all sampling points at the same detection time Real-time particle application rate The coefficient of variation of the application rate was obtained by solving the problem. ; The coefficient of variation of the application rate With setting a uniformity evaluation threshold The comparison is performed to obtain the comparison results, and the current operating parameters of the drone are determined based on the comparison results. The following are the field granule application uniformity grades: like If the uniformity of granule application in the field is judged to be excellent; if If the uniformity of granule application in the field is judged to be good; if If the uniformity of granule application in the field is deemed poor, the current operating parameters should be adjusted. Get updated job parameters And detect; This is the evaluation coefficient.

[0011] In a further embodiment, the continuous optical signal variation amount The generation process is as follows: When a single particle passes through the detection beam, the detection beam emitted by the fiber optic transmitter is momentarily blocked, and the light intensity received by the fiber optic receiver decreases from the initial light intensity. This produces a single step attenuation, resulting in a single pulse-like change in the optical signal. : In the formula, The light intensity after a single particle blocks the light; When multiple particles pass through the detection beam, the detection beam emitted by the fiber optic transmitter is momentarily blocked multiple times in succession, and the light intensity received by the fiber optic receiver decreases from the initial light intensity. This generates continuous multi-step attenuation, with the superposition of adjacent single-pulse optical signal changes forming a continuous optical signal change. : In the formula, For particles passing through ordinal numbers, The number of particles passing through continuously. For the first Each particle blocks the light intensity in front. For the first Light intensity after being blocked by a single particle.

[0012] In a further embodiment, the method for extracting the effective blocking signal features is as follows: Changes in continuous optical signal Convert to digital sampling sequence For the digital sampling sequence Perform adaptive multi-level noise reduction processing to obtain the noise-reduced digital sampling sequence. ; Within a preset sliding time window, the denoised digital sampling sequence is... The least squares method is used to estimate the signal baseline sequence that changes synchronously with the sampling time in real time. and noise amplitude sequence And generate a digital sampling sequence after noise reduction. Same-dimensional dynamic adaptive double threshold sequence: , ; In the formula, The lower threshold is... The upper limit threshold, This is the threshold coefficient; Denoising digital sampling sequence By performing point-by-point real-time comparison with the dynamic adaptive dual threshold sequence, effective occlusion signal features are filtered and extracted. like If the signal is detected, it is determined to be the start of an effective blocking signal caused by particle occlusion, and the signal features of the corresponding sampling time of the sampling point are extracted. like If the signal is interrupted, it is determined that the single interruption signal has ended, and the feature extraction of the single interruption signal for the entire time period is completed. like If the signal is abnormal, it is identified as a sudden change in ambient light or electromagnetic interference, and the signal within the corresponding sampling time of that sampling point is discarded.

[0013] In a further embodiment, unit time The number of particles detected by the beam inside The determination process is as follows: Based on the characteristics of the effective blocking signal, a blocking event decision is made to obtain a blocking event sequence. ,in, This is indicated as a valid blocking event. This is indicated as an invalid blocking event; The time-segmented integration algorithm is used to divide the occlusion event sequence. According to unit time Divide the time into segments and determine each unit of time. Total number of effective blocking events within The unit time is obtained using the following formula. The number of particles detected by the beam inside : , ; Correspondingly, real-time particle application rate The conversion formula is as follows: , This represents the average mass parameter of a single particle.

[0014] In a further embodiment, the process for calculating the coefficient of variation of the application rate is as follows: Each sampling point based on the same detection time Real-time granular application rate The mean amount of granules applied at all sampling points was calculated using the following formula. : ; Based on mean The standard deviation of the particle application rate at all sampling points was calculated using the following formula. : ; The formula for calculating the coefficient of variation of the application rate is as follows: .

[0015] The beneficial effects of this invention are: The detection device of this invention is equipped with adjustable legs consisting of a telescopic bracket and a horizontal adjustment rod. Combined with a level, it can perform precise horizontal calibration according to the terrain undulation of different sampling points in the field, ensuring that the particles fall and are detected in a vertical direction, reducing detection errors caused by uneven terrain. The anti-shake adsorption material laid at the bottom of the particle receiving cavity can effectively absorb the impact energy of the falling particles and suppress particle rebound, avoiding false counting caused by particles repeatedly passing through the detection beam. At the same time, the optical fiber transmitter and receiver are installed with adjustable angles, which can ensure that the detection beam completely covers the entrance area of ​​the particle receiving cavity. The device as a whole can be adapted to different terrains such as plains and hills, as well as complex field environments with different wind speeds and light, solving the problem that existing detection devices have weak environmental adaptability and cannot meet the actual field operation needs.

[0016] This invention uses an optical fiber sensing module to monitor the process of particles passing through a detection beam in real time. Combined with a microcontroller control system on a PCB board, it completes on-site signal processing, counting, and application rate conversion. The detection data can be synchronously transmitted to a visual display screen for on-site display, and data from multiple sampling points can be wirelessly aggregated to a host computer for real-time statistical analysis. This eliminates the need for manual sampling and weighing after the operation, solving the problems of long data acquisition cycles and lag behind the operation process in traditional detection methods. It enables the simultaneous execution of drone application and uniformity detection, providing data support for real-time adjustment of operation parameters on-site.

[0017] This invention employs adaptive multi-level noise reduction processing and dynamic adaptive dual-threshold sequence point-by-point discrimination technology to address changes in optical signals. It uses the least squares method to estimate the signal baseline and noise amplitude in real time, and combines this with appropriate threshold coefficient values ​​to accurately screen for effective blocking signals. This effectively suppresses signal distortion caused by ambient light fluctuations, mechanical vibration glitches, and electromagnetic interference, while eliminating false blocking signals and abnormal interference signals. Combined with a time-segmented integral algorithm for counting, it eliminates counting overlap errors caused by instantaneous high-density application, significantly improving the accuracy of particle counting and application rate conversion. This solves the problems of existing detection methods being susceptible to interference from complex field environments and having low detection accuracy.

[0018] This invention quantifies the uniformity of field application by calculating the coefficient of variation of the application rate. It replaces traditional qualitative judgments with quantitative indicators such as mean, standard deviation, and coefficient of variation. By comparing the coefficient of variation with a preset threshold, the uniformity level of application can be directly determined, clarifying whether adjustments to drone operation parameters are needed. The test results are objective, repeatable, and comparable. Simultaneously, the test data can be stored locally and analyzed on a host computer, and a spatial distribution curve of particle application can be generated. This provides scientific and comprehensive quantitative data support for the precise optimization of drone operation parameters and the continuous improvement of application quality, contributing to the improvement of the operational quality and precision of agricultural drone particle application. Attached Figure Description

[0019] Figure 1 This is a structural diagram of the field uniformity detection device for unmanned aerial vehicle (UAV) granule application in Example 1.

[0020] Figure 2 This is a flowchart of the field uniformity testing method for applied granules in Example 2.

[0021] Figure 3 This is a layout diagram of the uniformity detection device during the field uniformity detection of applied granules in Example 2.

[0022] Figures 1 to 3 The components are labeled as follows: uniformity detection device 1, detection device body 101, particle receiving cavity 102, optical fiber transmitter 103, telescopic bracket 104, horizontal adjustment rod 105, anti-shake adsorption material 106, level 107, visual display screen 108, and optical fiber receiver 109. Detailed Implementation

[0023] The present invention will now be further described with reference to the accompanying drawings and embodiments.

[0024] Example 1 like Figure 1The present embodiment discloses a field uniformity detection device 1 for unmanned aerial vehicle (UAV) application of particles, including: a detection device body 101 with several adjustable feet at the bottom. The detection device body 101 in this embodiment is a lightweight aluminum alloy integral molding structure. Its upper surface is recessed from top to bottom to form a particle receiving cavity 102 with a preset volume. The particle receiving cavity 102 is an open cavity structure with its entrance area serving as a particle detection channel, which is adapted to the vertical falling trajectory of particles under gravity. The inner wall of the cavity is smooth and non-sticky to avoid particle adhesion and residue affecting the detection accuracy.

[0025] An optical fiber sensing module is fixedly installed at preset positions on both sides of the inlet of the particle receiving cavity 102 within the main body 101 of the detection device, as required. In this embodiment, the optical fiber sensing module includes an optical fiber transmitter 103 and an optical fiber receiver 109 arranged opposite to each other. Both the optical fiber transmitter and receiver employ an adjustable-angle installation structure, allowing the installation angle to be adjusted on-site according to the inlet size of the particle receiving cavity 102 and the particle application range. This ensures that the infrared detection beam formed between the optical fiber transmitter and receiver completely penetrates the entire detection channel at the inlet of the particle receiving cavity 102 laterally, without any blind spots. Furthermore, the optical fiber sensing module has a sensitivity adaptive function, dynamically adjusting the beam emission intensity according to field ambient light and particle size to ensure the stability of the interruption detection.

[0026] To enable real-time processing of optical signals, particle counting, application rate conversion, and data storage and transmission, a PCB circuit board is embedded within the main body 101 of the detection device. The PCB circuit board is installed in a sealed manner, and the electronic cavity in which it is located is treated to be waterproof, moisture-proof, and electromagnetic interference-proof. The connection points are equipped with silicone sealing rings and waterproof connectors. The surface of the circuit board is coated with conformal coating to adapt to complex working environments such as field dew, mud splashes, and dust. Furthermore, the PCB circuit board is electrically connected to the fiber optic sensing module through wires, enabling rapid reception and transmission of optical signals.

[0027] It should be noted that the PCB circuit board described in this embodiment is an integrated control circuit board, which integrates a microcontroller control system, a data storage module, a wireless communication module (such as a LoRa mode combined with a 4G dual-mode wireless communication module) and an A / D conversion module.

[0028] To adapt to the complex terrain undulations in the field and ensure that the detection device body 101 is set horizontally to eliminate detection errors caused by particle trajectory deviation, the adjustable support legs described in this embodiment include a telescopic bracket 104 and a horizontal adjustment rod 105. The adjustable support legs are symmetrically arranged at the four corners of the bottom of the detection device body 101, forming a four-point stable support structure, which significantly improves the stability of the device on different ground surfaces such as soft soil and paddy field substrates.

[0029] The telescopic support 104 adopts a multi-section nested telescopic structure, and its telescopic stroke is adapted to the terrain undulation height of 0~20cm in the field. The bottom is equipped with a non-slip and wear-resistant rubber support pad to increase the contact area with the field ground and prevent the support from sinking. The top of the telescopic support 104 is detachably connected to the bottom of the detection device body 101 by a rotating buckle, which facilitates the disassembly, assembly and storage of the device. The horizontal adjustment rod 105 is a threaded adjustment rod, which passes through the side adjustment hole of the telescopic support 104 and is threadedly engaged with the telescopic support 104. The end of the adjustment rod is equipped with a non-slip rotary knob connected to the detection device body 101.

[0030] Operators can make minute height adjustments to the telescopic bracket 104 by turning the knob, with an adjustment accuracy of up to 1mm. Combined with the level 107 on the upper surface of the detection device body 101, the telescopic bracket 104 is used to complete the coarse level calibration of the device, and then the level adjustment rod 105 is used to complete the precise level calibration. This ensures that the particle receiving cavity 102 always remains horizontal, so that the particles fall vertically under gravity and stably pass through the detection beam. This effectively reduces particle offset and counting errors caused by device tilt, and improves the comparability and consistency of detection results at different sampling points.

[0031] To avoid the rebound caused by particles falling and hitting the bottom of the cavity, and to avoid miscounting caused by repeated crossing of the detection beam, and to improve the accuracy and repeatability of particle detection, a shock-absorbing material 106, such as a high-damping polyurethane microporous foam composite material, is laid on the bottom of the particle receiving cavity 102.

[0032] The anti-shake adsorption material 106 is attached and fixed to the bottom of the particle receiving cavity 102. Its coverage area covers the entire effective area where the particles fall, which can fully absorb the impact energy when the particles fall from the drone, quickly attenuate the kinetic energy of the particles, effectively suppress the rebound, bouncing or secondary bounce of the particles, and keep the particles in a stable static state after passing through the detection beam. This fundamentally eliminates the problem of multiple interruptions of the detection beam caused by particle rebound, ensures that the fiber optic sensing module responds accurately to the particle passing event in one go, and greatly reduces the counting error.

[0033] To better enable rapid leveling of the testing device and on-site visualization of testing data, this embodiment also includes a level 107 and a visualization display screen 108. The level 107, a high-precision bubble level, is embedded in the center of the upper surface of the testing device body 101. Its scale markings are clear, allowing for direct monitoring of the levelness of the testing device body 101. Combined with adjustable feet, it performs both coarse and fine leveling operations, ensuring that the particle receiving cavity 102 is always horizontal.

[0034] The visualization display screen 108 adopts a low-power high-definition segment code screen, which is embedded in a prominent position on the outside of the detection device body 101 and electrically connected to the PCB circuit board. It displays key information such as detection time, particle count per unit time, real-time particle application rate, and device working status in real time. It can realize on-site data viewing without external equipment, making it convenient for operators to keep track of the detection situation in real time. At the same time, a waterproof and scratch-resistant protective cover is added to the surface of the display screen to adapt to the complex field working environment and effectively extend its service life.

[0035] Finally, the detection device body 101 described in this embodiment is also provided with a material guide groove. The material guide groove is inclinedly opened on one side of the bottom of the particle receiving cavity 102. Its inlet end is smoothly connected to the bottom of the particle receiving cavity 102, and its outlet end penetrates through the side wall of the detection device body 101 and extends outward. The inner wall of the groove is polished to prevent sticking and reduce the resistance of particle slippage.

[0036] The feed chute is used to discharge the particles after the detection is completed. It allows the particles, after being detected and counted by the fiber optic sensing module, to slide smoothly out of the detection device under the action of gravity. This prevents the particles from accumulating in the particle receiving cavity 102, thus preventing the accumulated particles from blocking the detection beam and interfering with subsequent signal acquisition and counting. It also avoids secondary detection errors caused by the particles bouncing repeatedly in the cavity. At the same time, it realizes automatic discharge of the particles after detection, eliminating the need for manual cleaning of the cavity and ensuring that the detection device can continuously and stably carry out field application particle detection operations.

[0037] Example 2 Based on the unmanned aerial vehicle (UAV) particle application uniformity detection device disclosed in Example 1, such as Figure 2 This embodiment also discloses a method for detecting the uniformity of granule application in the field, including the following steps: Based on field parameter information and preset application width, the number and location coordinates of sampling points are determined, and corresponding uniformity detection devices are deployed; for example... Figure 3 As shown, based on field topographic elevation data, in sloping and gully areas with topographic relief greater than 10°, the sampling point spacing is reduced to 0.3~0.5m, and the detection devices are deployed more densely. In flat areas, the sampling points are deployed at equal intervals of 0.5~2.0m, ensuring detection accuracy in complex terrain areas while avoiding resource waste caused by excessive deployment in flat areas. The number of sampling points can be 5 to 21.

[0038] The following procedure is performed for each group of uniformity detection devices: Based on the topographic relief corresponding to the sampling point, the uniformity detection device is calibrated to ensure that the particle containment cavity is set horizontally. As described in Example 1, the retractable bracket is first adjusted to complete the coarse horizontal calibration of the detection device, and then the horizontal adjustment rod is used for precise fine adjustment until the bubble in the level returns to the center, so that the particle containment cavity is kept horizontal.

[0039] The drone operates according to the current parameters. During field pellet application, the pellets enter the pellet container under gravity and pass through the detection beam of the fiber optic sensor module during their descent; the current operation parameters described in this embodiment... This includes flight altitude, flight speed, and spray rotation speed, among which... This represents the number of times the job parameters are updated.

[0040] The detection beam formed by the fiber optic transmitter and receiver is momentarily blocked by a particle, causing continuous fluctuations in the light intensity received by the fiber optic receiver. This results in a continuous change in the optical signal that matches the frequency and velocity of the particle's passage. And transmit it to the PCB circuit board described in Example 1.

[0041] The PCB circuit board's response to the change in the continuous optical signal Perform signal shaping, filtering and noise reduction, and baseline correction preprocessing to extract effective blockage signal features; Based on the characteristics of effective blocking signals, blocking events are identified, and the effective counting of particle passage events is completed to obtain the unit time. The number of particles detected by the beam inside And calculate the real-time particle application rate at the corresponding sampling point. .

[0042] Considering the need for quantitative evaluation of the uniformity of particle spreading in the field and real-time optimization of UAV operation parameters, this embodiment also includes the following to achieve accurate determination of spreading effect and closed-loop adjustment of operation parameters: Summarize all sampling points at the same detection time Real-time particle application rate The coefficient of variation of the application rate was obtained by solving the problem. ; The coefficient of variation of the application rate With setting a uniformity evaluation threshold The comparison is performed to obtain the comparison results, and the current operating parameters of the drone are determined based on the comparison results. The following are the field granule application uniformity grades: like If the uniformity of particle application in the field is good, it indicates that the application is highly uniform and meets the requirements of agricultural production. The current operating parameters are also satisfactory. No adjustments are needed; the drone will continue its application operation according to the existing parameters.

[0043] like If the uniformity of particle spreading in the field is good, it indicates that the uniformity of spreading meets the basic requirements of agricultural production operations. There is no need to adjust the operating parameters, and spreading operations can continue to be carried out while the changes in spreading uniformity are continuously monitored in real time. Alternatively, the parameters can be further optimized if necessary.

[0044] like If the uniformity of particle spreading in the field is poor, it indicates that the particle spreading is uneven, with problems such as strip accumulation and blank areas. The current operational parameters, such as the drone's flight altitude, flight speed, spreading speed, and flight path spacing, need to be adjusted to obtain updated operational parameters. The drone then operated according to the updated parameters. When conducting field pellet spreading operations, this detection device simultaneously monitors the spreading amount and determines the uniformity level under the updated operating parameters in real time until the spreading uniformity level reaches the excellent or good standard, thus realizing closed-loop optimization and adjustment of operating parameters.

[0045] in, This is the evaluation coefficient, and its value ranges from 0.7 to 0.9.

[0046] To address the issue that fiber optic sensing modules lack clear response characteristics to the optical signals passing through particles, making it impossible to accurately distinguish the optical signal variation patterns between single particles and multiple particles passing continuously, the generation process of the continuous optical signal variation is as follows: When a single particle passes through the detection beam, the detection beam emitted by the fiber optic transmitter is momentarily blocked, and the light intensity received by the fiber optic receiver decreases from the initial light intensity. This produces a single step attenuation, resulting in a single pulse-like change in the optical signal. : In the formula, The light intensity after a single particle blocks the light; When multiple particles pass through the detection beam, the detection beam emitted by the fiber optic transmitter is momentarily blocked multiple times in succession, and the light intensity received by the fiber optic receiver decreases from the initial light intensity. This generates continuous multi-step attenuation, with the superposition of adjacent single-pulse optical signal changes forming a continuous optical signal change. : In the formula, For particles passing through ordinal numbers, The number of particles passing through continuously. For the first Each particle blocks the light intensity in front. For the first Light intensity after being blocked by a single particle.

[0047] The above technical solution characterizes the light intensity response pattern when a single particle and multiple particles pass through continuously, establishing a correspondence between the change in light signal and the frequency and speed of particle passage. This provides a clear and identifiable original signal basis for subsequent signal preprocessing and effective blocking signal feature extraction, effectively avoiding misjudgment of particle counting due to fuzzy light signal change patterns, and improving the recognition accuracy of the fiber optic sensing module for particle passage events.

[0048] Based on the above description, the method for extracting effective blocking signal features is as follows: The change in the continuous optical signal collected by the fiber optic receiver Convert the digital sampling sequence using an A / D conversion module. For the digital sampling sequence Perform adaptive multi-level noise reduction processing to obtain the noise-reduced digital sampling sequence. The adaptive multi-level noise reduction processing described in this embodiment is as follows: For the digital sampling sequence... The adaptive multi-stage noise reduction process sequentially performs dynamic window mid-range filtering, moving average filtering, and amplitude limiting to remove signal noise caused by ambient light fluctuations, mechanical vibrations, and electromagnetic interference.

[0049] Within a preset sliding time window, the denoised digital sampling sequence is... The least squares method is used to estimate the signal baseline sequence that changes synchronously with the sampling time in real time. and noise amplitude sequence And generate a digital sampling sequence after noise reduction. Same-dimensional dynamic adaptive double threshold sequence: , ; In the formula, The lower threshold is... The upper limit threshold, The threshold coefficient, with a value range of 1.5 to 3.0, is a core adjustment parameter for adapting to complex field detection environments. It needs to be flexibly selected according to the intensity of environmental interference and particle size in the actual operation scenario, so as to control the screening boundary of effective blocking signals and avoid missing effective signals due to excessively high threshold settings or misjudging interference signals due to excessively low threshold settings.

[0050] The specific threshold selection principle is as follows: In complex field scenarios with large ambient light fluctuations and significant electromagnetic interference, or when detecting small-diameter particles (≤5mm), a high threshold coefficient of 2.0 to 3.0 should be selected to effectively improve the stringency of signal screening and filter out false signals caused by environmental interference to the greatest extent. In conventional field scenarios with stable lighting and less interference, or when detecting large-diameter particles (>5mm), a low threshold coefficient of 1.5 to 2.0 should be selected to ensure the complete identification of effectively blocked signals and to avoid missed detection of particle passing events due to excessively high threshold values.

[0051] Meanwhile, this threshold coefficient can be linked with the sensitivity adaptive function of the fiber optic sensing module, and optimized synchronously with the real-time adjustment of beam intensity parameters and response thresholds. This ensures that the dynamic adaptive dual threshold sequence always maintains the optimal signal discrimination accuracy under different detection conditions, providing reliable parameter support for the effective extraction of blockage signal features.

[0052] Denoising digital sampling sequence By performing point-by-point real-time comparison with the dynamic adaptive dual threshold sequence, effective occlusion signal features are filtered and extracted. like If the signal is detected, it is determined to be the start of an effective blocking signal caused by particle occlusion. The signal features corresponding to the sampling time of the sampling point are extracted, such as signal amplitude, attenuation slope and other signal features.

[0053] like If the signal is interrupted, it is determined that the single interruption signal has ended, and the features such as amplitude change and duration of the single interruption signal from start to recovery are extracted. like If the signal is abnormal, such as a sudden change in ambient light or electromagnetic interference, the signal within the corresponding sampling time of that sampling point will be discarded to avoid interfering with the extraction accuracy of the signal features.

[0054] Based on the extraction of the above-mentioned effective blocking signal features, in order to overcome the counting overlap error caused by multiple particles passing continuously and instantaneous high-density application, and to avoid the problem of inaccurate particle counting caused by interference from ineffective blocking signals, the unit time described in this embodiment... The number of particles detected by the beam inside The determination process is as follows: Based on the characteristics of the effective blocking signal, a blocking event decision is made to obtain a blocking event sequence. ,in, This is indicated as a valid blocking event. This is indicated as an invalid blocking event; The time-segmented integration algorithm is used to divide the occlusion event sequence. According to unit time Divide the time into segments and determine each unit of time. Total number of effective blocking events within The unit time is obtained using the following formula. The number of particles detected by the beam inside : , ; Correspondingly, real-time particle application rate The conversion formula is as follows: , This is the average mass parameter of a single particle, which is pre-calibrated by sampling from batches of particulate material.

[0055] This method enables effective counting of particle passing events while ensuring the accuracy of real-time particle application rate conversion. It provides precise and reliable basic data support for the quantitative analysis of application uniformity at each sampling point, effectively improving the counting accuracy and data reliability of the entire detection method.

[0056] Finally, considering the introduction of the coefficient of variation of the application rate in this embodiment, the solution process is as follows: Each sampling point based on the same detection time Real-time granular application rate The mean amount of granules applied at all sampling points was calculated using the following formula. : ; Based on mean The standard deviation of the particle application rate at all sampling points was calculated using the following formula. : ; The formula for calculating the coefficient of variation of the application rate is as follows: .

[0057] By using the above-mentioned quantitative calculation method, the real-time application amount at each sampling point is transformed into a coefficient of variation index that reflects the overall distribution uniformity. This eliminates the subjectivity of traditional qualitative evaluation and achieves accurate quantitative judgment of the uniformity of particle application in the field. At the same time, this calculation method is compatible with the statistical analysis system of the host computer and can quickly complete the summary calculation of data from multiple sampling points. This provides a scientific and unified quantitative basis for subsequent comparison with preset uniformity evaluation thresholds and determination of the application uniformity level, effectively improving the objectivity and credibility of the uniformity evaluation results.

Claims

1. A device for detecting the uniformity of unmanned aerial vehicle (UAV) particle application in the field, characterized in that, include: The detection device body has several adjustable feet at its bottom; The upper surface of the detection device body is recessed from top to bottom to form a particle-containing cavity of a preset volume; An optical fiber sensing module is installed in the detection device body as required; the optical fiber sensing module includes: an optical fiber transmitter and an optical fiber receiver arranged opposite to each other, and a detection beam that transversely penetrates the entrance of the particle receiving cavity is formed between the optical fiber transmitter and the optical fiber receiver. The PCB circuit board is embedded in the detection device body; the PCB circuit board is also electrically connected to the fiber optic sensing module.

2. The field uniformity detection device for unmanned aerial vehicle (UAV) particle application according to claim 1, characterized in that, The adjustable support leg includes: a telescopic bracket and a horizontal adjustment rod; The bottom of the telescopic support is used to abut against the field ground, the top of the telescopic support is connected to the bottom of the detection device body, and the horizontal adjustment rod passes through the telescopic support and is threaded into the telescopic support.

3. The field uniformity detection device for unmanned aerial vehicle (UAV) particle application according to claim 1, characterized in that, Also includes: A shock-absorbing material is laid at the bottom of the particle receiving cavity; the shock-absorbing material is used to absorb the impact energy of the falling particles when the drone is spreading them and to suppress particle rebound.

4. The field uniformity detection device for unmanned aerial vehicle (UAV) particle application according to claim 1, characterized in that, Also includes: A level is installed on the upper surface of the detection device body; The level is used to monitor the horizontal status of the detection device body; A visual display screen is located on the outer side of the detection device body.

5. A method for detecting the uniformity of field application of applied particles using the unmanned aerial vehicle (UAV) particle application uniformity detection device as described in any one of claims 1 to 4, characterized in that, Includes the following steps: Based on field parameter information and preset application width, determine the number and location coordinates of sampling points, and deploy uniformity detection devices accordingly; perform the following procedure for each set of uniformity detection devices: Based on the terrain undulation corresponding to the sampling point, the uniformity detection device is calibrated to ensure that the particle containment chamber is set horizontally. The drone operates according to the current parameters. When performing field pellet spreading operations, the pellets enter the pellet holding cavity under the action of gravity and pass through the detection beam of the fiber optic sensing module during the falling process. The detection beam formed by the fiber optic transmitter and receiver is momentarily blocked by a particle, causing continuous fluctuations in the light intensity received by the fiber optic receiver. This results in a continuous change in the optical signal that matches the frequency and velocity of the particle's passage. ; For the change in the continuous optical signal Perform signal shaping, filtering and noise reduction, and baseline correction preprocessing to extract effective blockage signal features; Based on the characteristics of effective blocking signals, blocking events are identified, and the effective counting of particle passage events is completed to obtain the unit time. The number of particles is detected by the internal beam. And calculate the real-time particle application rate at the corresponding sampling point. .

6. The method for detecting the field uniformity of applied granules according to claim 5, characterized in that, It also includes the following steps: Summarize all sampling points at the same detection time Real-time particle application rate The coefficient of variation of the application rate was obtained by solving the problem. ; The coefficient of variation of application rate With setting a uniformity evaluation threshold The comparison is performed to obtain the comparison results, and the current operating parameters of the drone are determined based on the comparison results. The following are the field granule application uniformity grades: like If the uniformity of granule application in the field is judged to be excellent; if If the uniformity of granule application in the field is judged to be good; if If the uniformity of granule application in the field is deemed poor, the current operating parameters should be adjusted. Get updated job parameters And detect; This is the evaluation coefficient.

7. The method for detecting the field uniformity of applied granules according to claim 5, characterized in that, The change in the continuous optical signal The generation process is as follows: When a single particle passes through the detection beam, the detection beam emitted by the fiber optic transmitter is momentarily blocked, and the light intensity received by the fiber optic receiver decreases from the initial light intensity. This produces a single step attenuation, resulting in a single pulse-like change in the optical signal. : In the formula, The light intensity after a single particle blocks the light; When multiple particles pass through the detection beam, the detection beam emitted by the fiber optic transmitter is momentarily blocked multiple times in succession, and the light intensity received by the fiber optic receiver decreases from the initial light intensity. This generates continuous multi-step attenuation, with the superposition of adjacent single-pulse optical signal changes forming a continuous optical signal change. : In the formula, For the particle through ordinal number, The number of particles passing through continuously. For the first Each particle blocks the light intensity in front. For the first Light intensity after being blocked by a single particle.

8. The method for detecting the field uniformity of applied granules according to claim 5, characterized in that, The method for extracting the effective blocking signal features is as follows: Changes in continuous optical signal Convert to digital sampling sequence For the digital sampling sequence Perform adaptive multi-level noise reduction processing to obtain the noise-reduced digital sampling sequence. ; Within a preset sliding time window, the denoised digital sampling sequence is... The least squares method is used to estimate the signal baseline sequence that changes synchronously with the sampling time in real time. and noise amplitude sequence And generate a digital sampling sequence after noise reduction. Same-dimensional dynamic adaptive double threshold sequence: ， ； In the formula, The lower threshold is... The upper limit threshold, This is the threshold coefficient; Denoising digital sampling sequence By performing point-by-point real-time comparison with the dynamic adaptive dual threshold sequence, effective blocking signal features are filtered and extracted. like If the signal is detected, it is determined to be the start of an effective blocking signal caused by particle occlusion, and the signal features of the corresponding sampling time of the sampling point are extracted. like If the signal is interrupted, it is determined that the single interruption signal has ended, and the feature extraction of the single interruption signal for the entire time period is completed. like If the signal is abnormal, it is identified as a sudden change in ambient light or electromagnetic interference, and the signal within the corresponding sampling time of that sampling point is discarded.

9. The method for detecting the field uniformity of applied granules according to claim 5, characterized in that, unit time The number of particles is detected by the internal beam. The determination process is as follows: Based on the characteristics of the effective blocking signal, a blocking event decision is made to obtain a blocking event sequence. ,in, This is indicated as a valid blocking event. This is indicated as an invalid blocking event; The time-segmented integration algorithm is used to divide the occlusion event sequence. According to unit time Divide the time into segments and determine each unit of time. Total number of effective blocking events within The unit time is obtained using the following formula. The number of particles is detected by the internal beam. : , ; Correspondingly, real-time particle application rate The conversion formula is as follows: , This represents the average mass parameter of a single particle.

10. The method for detecting the field uniformity of applied granules according to claim 6, characterized in that, The process for calculating the coefficient of variation of the application rate is as follows: Each sampling point based on the same detection time Real-time granular application rate The mean amount of granules applied at all sampling points was calculated using the following formula. : ； Based on mean The standard deviation of the particle application rate at all sampling points was calculated using the following formula. : ； The formula for calculating the coefficient of variation of the application rate is as follows: 。

Images