Applicable UV sterilization methods and platforms based on nighttime optical phenotypes for facility agriculture

Abstract

This invention discloses a method and platform for ultraviolet sterilization based on nighttime optical phenotypes applicable to facility agriculture. The method includes: synchronously collecting near-infrared reflectance phenotypes, ultraviolet-induced fluorescence phenotypes, short-wave infrared phenotypes, and weak-light texture features of crops at night as multimodal optical phenotype data; decoupling and feature extraction of the collected multimodal optical phenotype data, and outputting four types of quantitative indices based on disease control needs, crop growth characteristics, and optical matrix dose control requirements; constructing a three-in-one adaptive decision model of "spectral matching - precise dose control - efficiency optimization" based on the four types of quantitative indices to drive the ultraviolet LED sterilizing lamp to achieve full-dimensional precise control; providing real-time optical feedback during the ultraviolet LED sterilizing lamp sterilization process to establish a dual protection mechanism of closed-loop equipment operation and crop protection; and uploading the entire chain of data to a big data cloud platform to optimize the adaptive decision model, thereby achieving precise sterilization.

Description

Technical Field

[0001] This invention relates to the field of smart agriculture technology, specifically to a method and platform for ultraviolet sterilization based on nighttime optical phenotypes applicable to facility agriculture. Background Technology

[0002] In facility agriculture, nighttime ultraviolet (UV) sterilization has become a core physical control method for reducing the use of chemical pesticides and lowering pesticide residues in agricultural products. Its core principle is to use UV light to destroy the nucleic acid molecular structure of microorganisms, inhibiting their reproduction and thus controlling diseases. However, current UV sterilization technology combined with dosage control has many significant shortcomings in practical applications, making it difficult to meet the intelligent and precise control needs of facility agriculture. These shortcomings are as follows:

[0003] 1. Low efficiency and lack of dosage control in ultraviolet disinfection: Traditional ultraviolet mercury lamps have low luminous efficiency, high environmental and safety risks, high energy consumption, and a large number of ineffective redundant components in the spectrum. Although ultraviolet LED lamps are being gradually promoted, they have not been combined with optical matrices to achieve precise dosage control, nor have they been accurately adapted to the type of disease and the growth status of crops, resulting in both energy waste and incomplete disinfection.

[0004] 2. Rigid control mode and poor spatial adaptability: Existing technologies mostly adopt open-loop control with fixed dose and fixed path, without differentiated dose adjustment logic of one-dimensional / two-dimensional / three-dimensional optical matrix, which can easily cause ultraviolet damage to healthy plants, or cause incomplete disinfection in diseased plant areas due to insufficient dose, and a single irradiation structure cannot achieve uniform disinfection of crops on multiple sides.

[0005] 3. Phenotypic sensing has defects and is disconnected from dosage control: Daytime phenotypic acquisition is easily interfered with by factors such as natural light, shadow, and high ambient temperature, resulting in low data acquisition accuracy and poor stability; Nighttime crop phenotypic acquisition lighting not only interferes with the crop dark cycle and affects normal crop growth, but also reduces the accuracy of phenotypic sensing. Furthermore, there is a lack of effective coordination between phenotypic sensing results and optical matrix dosage control, which cannot provide reliable support for precise dosage control.

[0006] 4. Insufficient integration of technologies: Optical phenotype, AI decision-making and optical matrix dose control are only superficially combined, without achieving deep collaboration in data acquisition, decision analysis and dose execution. Furthermore, the lack of precise control over core equipment parameters results in insufficient practicality of the technology and makes it difficult to implement and promote.

[0007] 5. Poor adaptability of irradiation band control: The ultraviolet band is not precisely adapted to fungal / bacterial diseases of field crops in facility agriculture. The canopy penetration is insufficient, the uniformity of disinfection is poor, and the band control is disconnected from phenotypic perception, which cannot meet the needs of large-scale and precise disinfection in field.

[0008] 6. High complexity of optical matrix control: Existing two-dimensional / three-dimensional UV sterilization optical matrices lack simplified control schemes, the control logic for different planting scenarios has not been adapted, the hardware and software development costs are high, and it is difficult to scale up and apply them.

[0009] The invention patent with patent application publication number CN120353124A discloses a synergistic control method for filtration and ultraviolet sterilization in a purely physical ballast water management system. This patent is aimed at continuous water bodies in ballast water systems. The dosage control revolves around water quality parameters, ultraviolet transmittance, flow rate and residence time. Essentially, it belongs to the overall operating condition control of the ultraviolet sterilization unit. The dosage execution mainly depends on the adjustment of ultraviolet lamp power and water flow rate. Its core is to meet the overall sterilization requirements by changing the residence time of the fluid in the ultraviolet unit.

[0010] Currently, there is no UV sterilization control method that can achieve a closed-loop control of "sensing-decision-precise dosage control-disinfection-feedback" while taking into account accuracy, efficiency, scenario adaptability and multi-faceted disinfection needs. This method cannot meet the actual needs of green and intelligent prevention and control in facility agriculture, and a new technical solution is urgently needed to fill the gap in this field. Summary of the Invention

[0011] The technical problem to be solved by this invention is that: at present, there is no method applicable to facility agriculture where ultraviolet disinfection is not combined with an optical matrix to achieve precise dose control.

[0012] To solve the above-mentioned technical problems, the present invention provides the following technical solution: A UV sterilization method based on nighttime optical phenotypes applicable to facility agriculture includes: Based on the near-infrared reflectance phenotype, ultraviolet-induced fluorescence phenotype, short-wave infrared phenotype, and weak light texture features of crops collected synchronously at night, multimodal optical phenotypic data were obtained. The collected multimodal optical phenotypic data are decoupled and feature extracted. Combined with the needs of disease control, crop growth characteristics, and optical matrix dose control requirements, four types of quantitative indices are output, including disease type identification index, disease severity quantitative index, lamp efficiency optimization index, and optical matrix adaptation index. Based on four types of quantitative indices, an adaptive decision-making model integrating "spectral matching, precise dose control, and efficiency optimization" is constructed to drive the ultraviolet LED germicidal lamp to achieve full-dimensional precise control. Real-time optical feedback is provided during the ultraviolet LED germicidal process to establish a dual protection mechanism that combines equipment operation with crop protection. The entire data chain is uploaded to a big data cloud platform to optimize the adaptive decision-making model.

[0013] Technical Effects: Addressing the core technical pain points of existing technologies, such as low efficiency of ultraviolet disinfection, lack of dosage control, significant interference from phenotypic perception, insufficient technological integration, complex optical matrix control, and poor scene adaptability, this invention deeply integrates nighttime multimodal optical phenotypic perception, adaptive decision-making, and one-dimensional / two-dimensional / three-dimensional ultraviolet bactericidal optical matrix dosage control. It provides an adaptive ultraviolet bactericidal optical matrix dosage control method based on nighttime optical phenotypic characteristics, achieving a closed-loop control of "perception—decision—precise dosage control—disinfection—feedback." This method balances accuracy, efficiency, scene adaptability, and multifaceted disinfection needs, achieving an organic unity of precise disinfection, high energy efficiency, crop protection, refined dosage control, and multi-scene adaptability. It fills a gap in existing technologies, meets the practical needs of intelligent green pest control in facility agriculture, and promotes the green and sustainable development of facility agriculture.

[0014] In this embodiment, a three-in-one adaptive decision-making model of "spectral matching - precise dosage control - efficiency optimization" is constructed to drive the ultraviolet LED germicidal lamp to achieve full-dimensional precise control, including: Spectral matching: Based on the disease type identification index, the different wavelengths of the ultraviolet LED germicidal lamp are precisely matched; Precise dose control of optical matrix: Based on the quantitative index of disease severity, the corresponding target bactericidal dose is determined. Through the coordinated adjustment of the one-dimensional / two-dimensional / three-dimensional ultraviolet LED optical matrix structure design and parameters in the ultraviolet LED germicidal lamp, the total ultraviolet bactericidal dose actually received by a single crop is close to or reaches the target bactericidal dose, adapting to the differentiated disinfection needs of different plants. Equipment efficiency optimization: Given that the target fungicide dose for a single plant has been determined, the driving parameters and structural parameters of the one-dimensional / two-dimensional / three-dimensional ultraviolet LED optical matrix structure are synergistically optimized by combining the plant canopy width, height, leaf orientation, and spatial distribution of lesions.

[0015] In this embodiment, the target bactericidal dose is determined based on the disease severity quantification index, and the mapping relationship between the two is as follows: ; In the formula, A quantitative index for the severity of disease and The larger the value, the higher the severity of the disease. For the target bactericidal dose, The minimum effective bactericidal dose, The maximum permissible bactericidal dose within the safety threshold range for crops.

[0016] In this embodiment, the total UV sterilization dose actually received by a single crop is expressed by the following formula: ; In the formula, This represents the total amount of UV disinfectant actually received by a single crop plant. For the first The light intensity conversion coefficient of each ultraviolet LED is used to characterize the optical properties of the corresponding LED. For the first The output power of each ultraviolet LED For the first The on / off state of each ultraviolet LED is set to 1 when it is on and 0 when it is off. The effective luminous length of the ultraviolet LED germicidal lamp. The vertical distance between the plant and the one-dimensional / two-dimensional / three-dimensional ultraviolet LED optical matrix structure. To improve the moving speed of the device equipped with ultraviolet LED germicidal lamps, This represents the total number of UV LED beads involved in the dosage calculation.

[0017] In this embodiment, the one-dimensional / two-dimensional / three-dimensional ultraviolet LED optical matrix structure is as follows: the one-dimensional structure is a horizontally uniform arrangement of ultraviolet LED light strips; the two-dimensional structure is a spatial array of ultraviolet LED light strips parallel to the forward direction and perpendicular to the forward direction; and the three-dimensional structure is multiple foldable ultraviolet LED light panels forming a gate-shaped or semi-circular structure.

[0018] In this embodiment, the different pest control needs of different plants are adapted, including: Segmented control: Based on the UV LED switch status, only the LEDs along the plant's path are turned on, while the LEDs in blank areas without plants are turned off. The dosage is controlled segmentally by adjusting the number of LEDs turned on. Continuous adjustment: Based on the fact that the moving speed of the device equipped with the ultraviolet LED germicidal lamp is negatively correlated with the disinfection dosage, the formula is used to adjust the speed. The movement speed is adjusted to match the target dosage; in areas with high disease severity and high dosage requirement, the movement speed is reduced, while in areas with low disease severity and low dosage requirement, the movement speed is increased. To improve the moving speed of the device equipped with ultraviolet LED germicidal lamps, The light intensity conversion coefficient of the ultraviolet LED. This refers to the output power of the ultraviolet LED. The effective luminous length of the ultraviolet LED germicidal lamp. To determine the number of LEDs to turn on, The vertical distance between the plant and the one-dimensional / two-dimensional / three-dimensional ultraviolet LED optical matrix structure. The target bactericidal dose; Fine-tuning: Based on the first Output power of each ultraviolet LED Each UV LED can independently adjust its output power. Based on the LED's on / off state and the moving speed of the UV LED sterilization device, the LED power can be finely adjusted to ensure that the total UV sterilization dose actually received by a single crop is precisely matched with the target sterilization dose.

[0019] In this embodiment, based on the disease type identification index, different wavelengths of the ultraviolet LED germicidal lamp are precisely adapted, including: when the disease type identification index identifies a bacterial infection area, the ultraviolet LED germicidal lamp outputs a disinfection band adapted to bacteria; when the disease type identification index identifies a fungal infection area, the ultraviolet LED germicidal lamp outputs a disinfection band adapted to fungi.

[0020] In this embodiment, the driving parameters and structural parameters of the one-dimensional / two-dimensional / three-dimensional ultraviolet LED optical matrix structure are synergistically optimized, including: Based on the location and range of the target irradiation area, the corresponding ultraviolet LED beads are selectively turned on to avoid ineffective irradiation of non-target areas; Based on the target bactericidal dosage requirement In order to satisfy Under these conditions, the driving current of a single ultraviolet LED is preferentially controlled within the high luminous efficiency operating range, and the operation of the ultraviolet LED is adjusted by turning on the UV LED beads. Output power of a single ultraviolet LED bead and the moving speed of the device equipped with ultraviolet LED germicidal lamps Achieve dose compensation; among which, This represents the total amount of UV sterilization actually received by a single crop plant. When the local heat load increases, the junction temperature rise can be suppressed by reducing the driving current of the corresponding LED beads, increasing the speed of the cooling fan, or enhancing the heat exchange capacity of the heat sink. The target sterilization dose can be maintained by increasing the irradiation time or optimizing the combination of ultraviolet LED beads.

[0021] In this embodiment, a dual protection mechanism is established, encompassing both equipment operation and crop protection within a closed loop, including: Closed-loop operation of the equipment: If fluctuations in ultraviolet light intensity, spectral shifts, junction temperature of ultraviolet LED germicidal lamps, or dose deviations are detected to exceed the threshold, the driving parameters of ultraviolet LED germicidal lamps, heat dissipation strategies, and one-dimensional / two-dimensional / three-dimensional ultraviolet LED optical matrix structure control parameters are immediately and automatically adjusted to maintain stable operation of ultraviolet LED germicidal lamps and accurate disinfection dosage. Crop protection closed loop: If abnormal crop phenotypes are detected, the light intensity of the ultraviolet LED germicidal lamp is quickly reduced, the disinfection dosage is adjusted, or the ultraviolet LED germicidal lamp is turned off to avoid ultraviolet damage to the plants. At the same time, abnormal data is recorded to provide support for subsequent adaptive decision-making model iteration and optimization.

[0022] The present invention also provides a mounting platform, including a ground mobile robot and a low-altitude near-canopy drone; the ground mobile robot and the low-altitude near-canopy drone serve as mounting carriers for ultraviolet LED germicidal lamps, and implement the ultraviolet sterilization method based on nighttime optical phenotype applicable to facility agriculture as described in any one of claims 1-9.

[0023] Compared with the prior art, the beneficial effects of the present invention are: This invention deeply integrates nighttime interference-free multimodal multichannel optical phenotyping, adaptive decision-making, and one-dimensional / two-dimensional / three-dimensional ultraviolet sterilization optical matrix dose control. It replaces traditional mercury lamps with an optimized dual-band ultraviolet LED optical matrix system, significantly reducing energy consumption. Through nighttime, interference-free multimodal phenotyping, it accurately acquires crop disease and growth data. AI models are used to decouple and analyze phenotypic data, outputting core quantitative indices to drive the optical matrix for spectral matching, precise dose control, and equipment efficiency optimization. A dual closed-loop protection mechanism for equipment operation and crop protection is established during the sterilization process. All-link data is uploaded to an agricultural big data cloud platform and the adaptive decision-making model is iterated for continuous technology optimization. This invention effectively solves the problems of low sterilization efficiency, lack of dose control, large sensing interference, and complex optical matrix control in existing technologies, achieving a unified approach of precise sterilization, high energy efficiency, crop safety, and multi-scenario adaptability. It is particularly suitable for intelligent green pest control in strawberry cultivation scenarios such as greenhouse ridges and open-air elevated structures, and integrates sterilization with field management, demonstrating high practicality and innovation.

[0024] This invention addresses the differentiated pest control needs of individual crop plants in facility agriculture by allocating dosage control based on their individual pest control requirements. It maps the target dosage to a disease severity quantification index to guide subsequent dosage execution, representing a disease quantification-driven, single-plant target dosage control approach. Furthermore, dosage execution is based on an ultraviolet LED optical matrix, allowing adjustment not only of the device's movement speed but also the on / off status of the LEDs, the number of LEDs activated, and the output power of individual LEDs. When necessary, it can also be spatially adapted using one-dimensional / two-dimensional / three-dimensional matrix structures, representing a spatialized, discretized, and differentiated dosage control approach for individual plants. This invention solves the problem of obtaining different target dosages for individual plants with varying disease severity, and ensures that the optical matrix precisely delivers the dosage to the target plant.

[0025] Precise dosage and crop safety: The actual dose received by a single plant is precisely matched to the target dose, with more accurate spectral matching. The light intensity is obtained by integrating the dose through scanning multiple UV LEDs one by one, ensuring high efficiency and controllability, and minimizing UV damage to plants. While ensuring the effectiveness of disinfection, it effectively avoids damage to healthy plants from UV light, achieving an organic balance between precise disinfection and crop protection, and ensuring normal crop growth.

[0026] Significant and uniform disinfection effect: The three-dimensional optical matrix achieves multi-faceted uniform disinfection on the top and sides of the crop, significantly improving the uniformity of disinfection within the field canopy, enhancing the sterilization effect, reducing energy consumption, and ensuring disinfection effect without dead angles in different cultivation scenarios. It can effectively control the occurrence of diseases in greenhouse crops such as strawberries, reduce the use of chemical pesticides, and reduce pesticide residues in agricultural products.

[0027] High efficiency, energy saving and controllable cost: UV LED lamps have high luminous efficiency and significantly reduce energy consumption compared to traditional UV mercury lamps. The equipment has a long service life of ≥15,000 hours, which greatly reduces the user's equipment investment and operating costs. The simplified control scheme of the two-dimensional optical matrix reduces hardware and software development costs. The simple ground quadruped robot mounting platform enables low-cost deployment, which is in line with the concept of green and energy-saving development.

[0028] Rapid response and stable operation: The equipment parameter adjustment, dosage correction, and dual closed-loop protection mechanisms have short response delays; precise LED on / off and power adjustment; minimal luminous efficiency decay after long-term operation; and the integral dosage control method ensures accurate and controllable total dosage even with LED decay, thanks to the algorithm's periodic self-monitoring and calibration. This ensures a stable and controllable disinfection process, preventing disinfection failure or crop damage caused by abnormal conditions.

[0029] Highly adaptable to various scenarios and highly efficient in operation: It supports dual-mode operation of ground mobile robots and low-altitude drones, and can be flexibly adapted to various cultivation scenarios such as greenhouse ridge strawberries and open-air elevated strawberries. The efficiency of drone operation is 3 times higher than that of ground equipment, and the ground mobile robot can adapt to complex terrain between ridges, meeting the disease prevention and control needs of large-scale facility agriculture.

[0030] Significant added value: High accuracy in early warning of abnormal growth; core data across the entire chain helps build an agricultural big data system, realizing the integration of "sterilization + field management", providing data support for the whole-cycle growth regulation of crops, and promoting the development and implementation of smart agriculture. Attached Figure Description

[0031] Figure 1 This is a flowchart of an ultraviolet sterilization method based on nighttime optical phenotype applicable to facility agriculture, according to an embodiment of the present invention.

[0032] Figure 2 This is a schematic diagram of an ultraviolet LED lamp and crops according to an embodiment of the present invention. Detailed Implementation

[0033] To facilitate understanding of the technical solution of the present invention by those skilled in the art, the technical solution of the present invention will now be further described in conjunction with the accompanying drawings.

[0034] The terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this application, "multiple" means two or more, unless otherwise explicitly specified.

[0035] Example 1 Please see Figure 1 As shown, the present invention provides a UV sterilization method based on nighttime optical phenotype applicable to facility agriculture, comprising: S10 is based on near-infrared reflectance phenotype, ultraviolet-induced fluorescence phenotype, short-wave infrared phenotype, and weak light texture features of crops collected synchronously at night, serving as multimodal optical phenotypic data.

[0036] In one embodiment of the invention, the platform serves as a multimodal, multi-channel optical acquisition module and an ultraviolet LED germicidal lamp. It can operate along a preset path or achieve fully autonomous navigation, adapting to the operational needs of different crops in facility agriculture settings such as greenhouses and open fields, including strawberry cultivation. The platform can be a ground-based mobile robot (a simple quadruped robot) or a low-altitude, near-canopy drone. The drone's operating altitude is strictly controlled within 0.5–1.5 m above the crop canopy to ensure ultraviolet irradiation uniformity ≥90%, avoiding uneven disinfection, insufficient local dosage, or excessive irradiation.

[0037] In this embodiment, the UV LED germicidal lamp uses UV LED instead of the traditional UV mercury lamp and has dual-band output capability. Specifically, it is configured with a UV-C 250-280nm core disinfection band and a UV-A 315-365nm fungal targeting auxiliary band, with a luminous efficiency of ≥55%, a half-width at half-maximum (WHM) of ≤15nm, a lamp junction temperature controlled at ≤85℃, and a lifespan of ≥15000h. The energy consumption is reduced by more than 40% compared with the traditional UV mercury lamp.

[0038] like Figure 2 As shown, in this embodiment, the ultraviolet LED germicidal lamp integrates a one-dimensional / two-dimensional / three-dimensional ultraviolet LED optical matrix structure, specifically including a one-dimensional structure of horizontally uniformly arranged ultraviolet LED strips, with the spacing between the LED beads marked. Total length The two-dimensional structure consists of a spatial array of ultraviolet LED light strips parallel to and perpendicular to the direction of travel; the three-dimensional structure consists of multiple foldable ultraviolet LED light panels forming a gate-shaped or semi-circular structure that fits the outline of the object to achieve multi-faceted illumination from the top and sides.

[0039] In this embodiment, under the premise of no visible light interference and without disrupting the crop's dark cycle, a multimodal, multi-channel optical acquisition module (spectral resolution ≤10nm) is used to simultaneously acquire near-infrared reflectance phenotypes, ultraviolet-induced fluorescence phenotypes, short-wave infrared phenotypes, and low-light texture features. The near-infrared reflectance phenotype is used to assist in adjusting light intensity and dosage; the ultraviolet-induced fluorescence phenotype is used to identify disease type and infection degree; the short-wave infrared phenotype is used to capture changes in moisture and chemical composition caused by diseases, and a thermal infrared imaging module can be optionally added to assist in detecting local temperature anomalies caused by diseases; the low-light texture features are used to characterize detailed information such as leaf surface roughness, vein continuity, lesion edge morphology, spot distribution density, local wrinkles, and surface damage texture. Specifically, the multimodal, multi-channel optical acquisition module includes a near-infrared LED, an ultraviolet-excited LED, a short-wave infrared LED (1100–1700 nm), and a low-light imaging submodule. The low-light imaging submodule includes a high-sensitivity low-illuminance industrial camera, an imaging lens, and a non-visible auxiliary lighting unit. The auxiliary lighting unit preferably employs near-infrared narrowband illumination to simultaneously image leaves, fruit surfaces, stems, and diseased areas without introducing visible light interference or disrupting the crop's dark cycle. During acquisition, the control system synchronously triggers camera exposure and auxiliary lighting to obtain grayscale texture images or near-infrared reflective texture images of the crop surface. This precise acquisition of core data related to crop disease types, infection levels, and growth status provides accurate and reliable data support for subsequent AI-based adaptive decision-making models.

[0040] In this embodiment, the advantages of nighttime operation are as follows: Sunlight (especially blue light) triggers the fungal repair mechanism, thereby eliminating the damage caused by UV-C ultraviolet rays. Disinfection in the dark can prevent this repair, improving the disinfection effect. Nighttime operation utilizes a multi-modal, multi-channel optical acquisition module that integrates phenotypic acquisition methods, eliminating visible light interference, shadows, and high light saturation. This results in more accurate and stable field detection without disrupting the crop's dark cycle, enabling non-destructive, high-throughput detection at night and providing reliable data for precise dosage control.

[0041] In this embodiment, the platform is further equipped with a high-precision light intensity monitoring module, a spectrum monitoring module, and a dose monitoring module to collect data such as ultraviolet light intensity, spectral stability, disinfection dose, and equipment operating status in real time, providing reliable data support for subsequent closed-loop adjustment and ensuring that the disinfection dose is accurate and controllable.

[0042] S20 decouples and extracts features from the collected multimodal optical phenotypic data, and outputs four types of quantitative indices based on disease control requirements, crop growth characteristics, and optical matrix dosage control requirements: disease type identification index, disease severity quantitative index, lamp efficiency optimization index, and optical matrix adaptation index.

[0043] In one embodiment of the present invention, multimodal optical phenotypic data is decoupled and feature extracted to output a network model with four types of quantization indices. The present invention is not limited to this; specifically, a lightweight CNN network based on the MobileNet-V3 architecture can be used.

[0044] In this embodiment, the collected multimodal optical phenotypic data undergoes decoupling analysis and feature extraction. Combining disease control needs, crop growth characteristics, and optical matrix dosage control requirements, four core quantitative indices are output: disease type identification index, disease severity quantification index, lighting efficiency optimization index, and optical matrix adaptation index. Simultaneously, growth monitoring indices (including plant health index and abnormal growth early warning index) are also output, achieving synergistic advancement of disease control and crop growth monitoring. The disease type identification index can accurately distinguish between different types of diseases such as fungi and bacteria, providing a direct basis for accurate matching of ultraviolet spectra and setting of initial parameters for optical matrices.

[0045] The disease severity quantification index, with a value range of 0 to 1, can accurately quantify the depth and severity of disease infection. Moreover, this index is positively correlated with the required ultraviolet dose and light intensity, providing a core basis for the precise control of optical matrix dose.

[0046] The luminaire efficiency optimization index, based on environmental factors such as ambient temperature and humidity, provides a basis for dynamically adjusting the luminaire drive current and heat dissipation strategy, maintaining the efficient and stable operation of the luminaire, and avoiding energy waste and equipment damage.

[0047] The optical matrix adaptation index outputs initial adaptation parameters such as optical matrix type (one-dimensional / two-dimensional / three-dimensional), number of LEDs turned on, and moving speed of the mounting platform, based on crop plant type, planting density, and arrangement, thereby improving the adaptability of the optical matrix to the cultivation scenario.

[0048] S30, based on four types of quantitative indices, constructs a three-in-one adaptive decision-making model of "spectral matching, precise dose control, and efficiency optimization" to drive the ultraviolet LED germicidal lamp to achieve full-dimensional precise control.

[0049] In one embodiment of the present invention, a three-in-one adaptive decision-making model of "spectral matching-precise dosage control-efficiency optimization" is constructed to drive the ultraviolet LED germicidal lamp to achieve full-dimensional precise control, including: Spectral matching: Based on the disease type identification index, the different wavelengths of the ultraviolet LED germicidal lamp are precisely matched. The peak wavelength of the ultraviolet LED lamp is precisely matched according to the disease type identification index, balancing the disinfection effect with the safety of strawberry plants.

[0050] In this embodiment, when the disease type identification index identifies a bacterial infection area, the ultraviolet LED germicidal lamp outputs a disinfection wavelength adapted to the bacteria. Specifically, for bacterial infection areas (such as strawberry bacterial wilt), the UV-C 250-280nm wavelength is used (this wavelength has the strongest bactericidal ability and can quickly destroy the bacterial nucleic acid structure, achieving efficient disinfection).

[0051] In this embodiment, when the disease type identification index identifies a fungal infection area, the ultraviolet LED germicidal lamp outputs a disinfection wave adapted to the fungus. The fungal infection area (such as strawberry powdery mildew or gray mold) is combined with the UV-A 315-365nm band (this band has stronger penetration, reaching deep into the strawberry canopy crevices while causing less damage to the strawberry plant; it inhibits fungal spore germination, interferes with mycelial growth, and induces the plant's own defense response, thereby achieving safe and effective integrated control). Spectral matching optimization significantly improves the targeting and efficiency of disinfection, while avoiding plant damage or incomplete disinfection caused by improper band selection.

[0052] Precise dose control of optical matrix: Based on the quantitative index of disease severity, the corresponding target bactericidal dose is determined. Through the coordinated adjustment of the one-dimensional / two-dimensional / three-dimensional ultraviolet LED optical matrix structure design and parameters in the ultraviolet LED germicidal lamp, the total ultraviolet bactericidal dose actually received by a single crop is close to or reaches the target bactericidal dose, adapting to the differentiated disinfection needs of different plants.

[0053] In this embodiment, the target bactericidal dose is determined based on the disease severity quantification index, and the mapping relationship between the two is as follows: ; In the formula, A quantitative index for the severity of disease and The larger the value, the higher the severity of the disease. For the target bactericidal dose, The minimum effective bactericidal dose, The maximum permissible bactericidal dose within the safety threshold range for crops.

[0054] In this embodiment, the total amount of ultraviolet disinfection actually received by a single crop is expressed by the following formula: ; In the formula, This represents the total amount of UV disinfectant actually received by a single crop plant. For the first The light intensity conversion coefficient of each ultraviolet LED is used to characterize the optical properties of the corresponding LED. For the first The output power of each ultraviolet LED For the first The on / off state of each ultraviolet LED is set to 1 when it is on and 0 when it is off. The effective luminous length of the ultraviolet LED germicidal lamp. The vertical distance between the plant and the one-dimensional / two-dimensional / three-dimensional ultraviolet LED optical matrix structure. To improve the moving speed of the device equipped with ultraviolet LED germicidal lamps, This represents the total number of UV LED beads involved in the dosage calculation.

[0055] In this embodiment, the different pest control needs of different plants are adapted, including: Segmented control: Based on the UV LED switch status, only the LEDs along the plant's path are turned on, while the LEDs in blank areas without plants are turned off. By adjusting the number of LEDs turned on, the dosage can be controlled in segments, reducing ineffective energy consumption.

[0056] Continuous adjustment: Based on the fact that the moving speed of the device equipped with the ultraviolet LED germicidal lamp is negatively correlated with the disinfection dosage, the formula is used to adjust the speed. The movement speed is adjusted to match the target dosage by reverse calculation. In areas with high disease severity and high dosage requirement, the movement speed is reduced; in areas with low disease severity and low dosage requirement, the movement speed is increased. Among these, To improve the moving speed of the device equipped with ultraviolet LED germicidal lamps, The light intensity conversion coefficient of the ultraviolet LED. This refers to the output power of the ultraviolet LED. The effective luminous length of the ultraviolet LED germicidal lamp. To determine the number of LEDs to turn on, The vertical distance between the plant and the one-dimensional / two-dimensional / three-dimensional ultraviolet LED optical matrix structure. The target bactericidal dose.

[0057] Fine-tuning: Based on the first Output power of each ultraviolet LED Each UV LED can independently adjust its output power. Based on the LED's on / off state and the moving speed of the UV LED sterilization device, the LED power can be finely adjusted to ensure that the total UV sterilization dose actually received by a single crop is precisely matched with the target sterilization dose.

[0058] Equipment efficiency optimization: Given that the target fungicide dose for a single plant has been determined, the driving parameters and structural parameters of the one-dimensional / two-dimensional / three-dimensional ultraviolet LED optical matrix structure are synergistically optimized by combining the plant canopy width, height, leaf orientation, and spatial distribution of lesions.

[0059] In this embodiment, the specific steps include: selectively activating corresponding LEDs based on the location and extent of the target irradiation area to avoid ineffective irradiation of non-target areas. Based on the target dose requirement, while meeting... Under these conditions, the driving current of a single ultraviolet LED is preferentially controlled within the high luminous efficiency operating range, and the operation of the ultraviolet LED is adjusted by turning on the UV LED beads. Output power of a single ultraviolet LED bead and the moving speed of the device equipped with ultraviolet LED germicidal lamps Dosage compensation is achieved. When local heat load increases, the junction temperature rise is suppressed by reducing the corresponding LED drive current, increasing the cooling fan speed, or enhancing the heat exchange capacity of the heat sink. The target bactericidal dose is maintained by increasing the irradiation time or optimizing the combination of UV LEDs. For the three-dimensional optical matrix, the deployment angle and irradiation direction of the lateral array can be further adjusted according to the plant canopy structure to match the irradiation range with the plant's spatial morphology. This reduces redundant irradiation, lowers system energy consumption, and extends the lifespan of the UV LEDs while ensuring effective irradiation of the lesion area and achieving the target dose.

[0060] The S40 provides real-time optical feedback during the UV LED sterilization process, establishing a dual protection mechanism that combines equipment operation with crop protection.

[0061] In one embodiment of the present invention, during the ultraviolet disinfection process, a high-precision monitoring module continuously collects ultraviolet light intensity, spectral stability, lamp junction temperature, disinfection dosage, and crop phenotypic characteristics to establish a dual protection mechanism of equipment operation closed loop and crop protection closed loop, ensuring that the disinfection process is stable, safe, and controllable, specifically including: Closed-loop equipment operation: If fluctuations in ultraviolet light intensity, spectral shifts, junction temperature of ultraviolet LED germicidal lamps, or dose deviations are detected to exceed the threshold, the driving parameters of ultraviolet LED germicidal lamps, heat dissipation strategies, and one-dimensional / two-dimensional / three-dimensional ultraviolet LED optical matrix structure control parameters are immediately and automatically adjusted to maintain stable operation of ultraviolet LED germicidal lamps and accurate disinfection dosage.

[0062] Crop protection closed loop: If abnormal crop phenotypes are detected, the light intensity of the ultraviolet LED germicidal lamp is quickly reduced, the disinfection dosage is adjusted, or the ultraviolet LED germicidal lamp is turned off to avoid ultraviolet damage to the plants. At the same time, abnormal data is recorded to provide support for subsequent adaptive decision-making model iteration and optimization.

[0063] The S50 uploads end-to-end data to a big data cloud platform to optimize the adaptive decision-making model.

[0064] In one embodiment of the present invention, the core data of the entire chain (including phenotypic collection data, disease index data, optical matrix control parameters, disinfection dosage data, equipment operation data, and disinfection effect data) is uploaded to a big data cloud platform. The adaptive decision model is iteratively optimized using cross-validation method to further improve the spectral matching strategy, optical matrix dosage control logic, and equipment configuration parameters. At the same time, crop growth data is accumulated simultaneously to help build an agricultural big data system, realize continuous upgrading and optimization of the technology, and improve the scenario adaptability and practical applicability of the technology.

[0065] Example 2 The present invention also provides a mounting platform, including a ground mobile robot and a low-altitude near-canopy drone, wherein the ground mobile robot and the low-altitude near-canopy drone serve as mounting carriers for ultraviolet LED germicidal lamps to perform the ultraviolet sterilization method based on nighttime optical phenotypes applicable to facility agriculture as described in Example 1.

[0066] Specific example 1: Greenhouse strawberry ridge cultivation based on facility agriculture - ground quadruped robot + two-dimensional optical matrix mode.

[0067] Application scenario: Greenhouse strawberry ridge cultivation, with individual plants planted in the open ground and arranged in a single-plant-independent manner, mainly for the prevention and control of powdery mildew (fungus) and bacterial wilt (bacteria).

[0068] System Deployment: The ground quadruped robot is equipped with a dual-band ultraviolet LED disinfection system and a two-dimensional ultraviolet sterilization optical matrix. It adopts a simplified control strategy (synchronous control of LED beads perpendicular to the direction of travel and LED beads parallel to the direction of travel). It is also equipped with a multi-modal multi-channel optical acquisition module, a parameter monitoring module and a dose monitoring module. After the equipment is debugged, it is put into use.

[0069] Nighttime phenotypic data collection: Multimodal optical phenotypic data of strawberry plants are collected through a multimodal multi-channel optical acquisition module. After decoupling analysis and feature extraction by an AI model, areas such as powdery mildew or bacterial wilt are identified, and the disease severity index of each area is output, and the target fungicide dose for a single plant is obtained.

[0070] Adaptive decision-making and optical matrix control: The adaptive decision-making model automatically switches the ultraviolet spectrum based on the disease type identification index; the two-dimensional optical matrix synchronously controls the on / off state and output power of the LEDs perpendicular / parallel to the direction of travel according to a simplified control strategy. In areas with high disease severity, the robot's moving speed is reduced and the LED power is increased, while in areas with low disease severity, the moving speed is increased and the power is reduced, thus achieving precise dosage control for single-row strawberry plants.

[0071] Closed-loop monitoring and disinfection: During the disinfection process, ultraviolet light intensity, spectral stability, disinfection dosage, and strawberry phenotypic characteristics are collected in real time. If there are no abnormal parameters or phenotypic conditions, the disinfection operation is completed according to the calculated dosage.

[0072] Furthermore, a three-dimensional matrix model can be adopted based on the ridge topography on site.

[0073] Specific example 2: Open-air elevated strawberry cultivation based on facility agriculture - low-altitude drone + three-dimensional foldable optical matrix mode.

[0074] Application scenario: Open-air elevated strawberry cultivation, planted on long, raised trellises, arranged one trellis at a time, and planted individually, mainly to prevent and control powdery mildew, gray mold, etc.

[0075] System deployment: The low-altitude UAV flies at a fixed altitude of 0.8m and the initial flight speed is set at 1.2m / s. It is equipped with an ultraviolet LED disinfection system and a three-dimensional gate-shaped foldable optical matrix (3 foldable long strip LED light panels) to improve the uniformity of ultraviolet irradiation. It is also equipped with a parameter monitoring module and a dose monitoring module to collect equipment operating status and disinfection dose data in real time. Nighttime phenotypic data collection: Through a multimodal, multi-channel optical acquisition module, near-infrared reflectance phenotypic, ultraviolet-induced fluorescence phenotypic, short-wave infrared phenotypic, and weak light texture feature data of elevated strawberry plants are collected across the entire area. After analysis by an AI model, high-risk areas for strawberry fungal diseases are identified, and the target fungicide dose for each area is output.

[0076] Adaptive decision-making and optical matrix control: The adaptive decision-making model drives the three-dimensional foldable optical matrix to automatically unfold into a gate shape, fitting the outline of the plant to achieve multi-faceted uniform irradiation of the top and sides of the elevated strawberry plant; Based on the quantitative index of disease severity, the ultraviolet light intensity and drone flight speed are dynamically adjusted, the light intensity is set to match the standard area, the initial flight speed is maintained, and the output power of the LED beads is finely adjusted to ensure that the actual disinfection dose is accurately matched with the target dose. Closed-loop monitoring and disinfection: During the disinfection process, the drone's flight parameters, optical matrix operation status, disinfection dosage, and strawberry phenotype are monitored in real time. If no abnormalities occur, the disinfection operation is completed according to the control parameters.

[0077] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered illustrative and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention, and no reference numerals in the claims should be construed as limiting the scope of the claims.

[0078] The above-described embodiments are merely examples of implementation methods of the invention. The scope of protection of the present invention is not limited to the above-described embodiments. For those skilled in the art, several modifications and improvements can be made without departing from the concept of the present invention, and these all fall within the scope of protection of the present invention.

Claims

1. A UV sterilization method based on nighttime optical phenotype applicable to facility agriculture, characterized in that, include: Based on the near-infrared reflectance phenotype, ultraviolet-induced fluorescence phenotype, short-wave infrared phenotype, and weak light texture features of crops collected synchronously at night, multimodal optical phenotypic data were obtained. The collected multimodal optical phenotypic data are decoupled and feature extracted. Combined with the needs of disease control, crop growth characteristics, and optical matrix dose control requirements, four types of quantitative indices are output, including disease type identification index, disease severity quantitative index, lamp efficiency optimization index, and optical matrix adaptation index. Based on four types of quantitative indices, an adaptive decision-making model integrating "spectral matching, precise dose control, and efficiency optimization" is constructed to drive the ultraviolet LED germicidal lamp to achieve full-dimensional precise control. Real-time optical feedback is provided during the ultraviolet LED germicidal process to establish a dual protection mechanism that combines equipment operation with crop protection. The entire data chain is uploaded to a big data cloud platform to optimize the adaptive decision-making model.

2. The ultraviolet sterilization method for facility agriculture based on nighttime optical phenotypes according to claim 1, characterized in that, An adaptive decision-making model integrating "spectral matching, precise dosage control, and efficiency optimization" is constructed to drive the UV LED germicidal lamp to achieve full-dimensional precise control, including: Spectral matching: Based on the disease type identification index, the different wavelengths of the ultraviolet LED germicidal lamp are precisely matched; Precise dose control of optical matrix: Based on the quantitative index of disease severity, the corresponding target bactericidal dose is determined. Through the coordinated adjustment of the one-dimensional / two-dimensional / three-dimensional ultraviolet LED optical matrix structure design and parameters in the ultraviolet LED germicidal lamp, the total ultraviolet bactericidal dose actually received by a single crop is close to or reaches the target bactericidal dose, adapting to the differentiated disinfection needs of different plants. Equipment efficiency optimization: Given that the target fungicide dose for a single plant has been determined, the driving parameters and structural parameters of the one-dimensional / two-dimensional / three-dimensional ultraviolet LED optical matrix structure are synergistically optimized by combining the plant canopy width, height, leaf orientation, and spatial distribution of lesions.

3. The ultraviolet sterilization method for facility agriculture based on nighttime optical phenotypes according to claim 2, characterized in that, The target fungicide dosage is determined based on the disease severity quantification index, and the mapping relationship between the two is as follows: ； In the formula, A quantitative index for the severity of disease and The larger the value, the higher the severity of the disease. For the target bactericidal dose, The minimum effective bactericidal dose, The maximum permissible bactericidal dose within the safety threshold range for crops.

4. The ultraviolet sterilization method for facility agriculture based on nighttime optical phenotypes according to claim 2, characterized in that, The total amount of ultraviolet (UV) disinfectant actually received by a single crop plant is expressed by the following formula: ； In the formula, This represents the total amount of UV disinfectant actually received by a single crop plant. For the first The light intensity conversion coefficient of each ultraviolet LED is used to characterize the optical properties of the corresponding LED. For the first The output power of each ultraviolet LED For the first The on / off state of each ultraviolet LED is set to 1 when it is on and 0 when it is off. The effective luminous length of the ultraviolet LED germicidal lamp. The vertical distance between the plant and the one-dimensional / two-dimensional / three-dimensional ultraviolet LED optical matrix structure. To improve the moving speed of the device equipped with ultraviolet LED germicidal lamps, This represents the total number of UV LED beads involved in the dosage calculation.

5. The ultraviolet sterilization method for facility agriculture based on nighttime optical phenotypes according to claim 2, characterized in that, The one-dimensional / two-dimensional / three-dimensional ultraviolet LED optical matrix structure is as follows: one-dimensional is a horizontally uniform arrangement of ultraviolet LED light strips, two-dimensional is a spatial array of ultraviolet LED light strips parallel to the forward direction and perpendicular to the forward direction, and three-dimensional is multiple foldable ultraviolet LED light panels, forming a gate-shaped or semi-circular structure.

6. The ultraviolet sterilization method for facility agriculture based on nighttime optical phenotypes according to claim 2, characterized in that, To meet the diverse pest control needs of different plants, including: Segmented control: Based on the UV LED switch status, only the LEDs along the plant's path are turned on, while the LEDs in blank areas without plants are turned off. The dosage is controlled segmentally by adjusting the number of LEDs turned on. Continuous adjustment: Based on the fact that the moving speed of the device equipped with the ultraviolet LED germicidal lamp is negatively correlated with the disinfection dosage, the formula is used to adjust the speed. The movement speed is adjusted to match the target dosage; in areas with high disease severity and high dosage requirement, the movement speed is reduced, while in areas with low disease severity and low dosage requirement, the movement speed is increased. To improve the moving speed of the device equipped with ultraviolet LED germicidal lamps, The light intensity conversion coefficient of the ultraviolet LED. This refers to the output power of the ultraviolet LED. The effective luminous length of the ultraviolet LED germicidal lamp. To determine the number of LEDs to turn on, The vertical distance between the plant and the one-dimensional / two-dimensional / three-dimensional ultraviolet LED optical matrix structure. The target bactericidal dose; Fine-tuning: Based on the first Output power of each ultraviolet LED Each UV LED can independently adjust its output power. Based on the LED's on / off state and the moving speed of the UV LED sterilization device, the LED power can be finely adjusted to ensure that the total UV sterilization dose actually received by a single crop is precisely matched with the target sterilization dose.

7. The ultraviolet sterilization method for facility agriculture based on nighttime optical phenotypes according to claim 2, characterized in that, Based on the disease type identification index, the UV LED germicidal lamp is precisely adapted to different wavelengths, including: when the disease type identification index identifies a bacterial infection area, the UV LED germicidal lamp outputs a disinfection wavelength adapted to bacteria; when the disease type identification index identifies a fungal infection area, the UV LED germicidal lamp outputs a disinfection wavelength adapted to fungi.

8. The ultraviolet sterilization method for facility agriculture based on nighttime optical phenotypes according to claim 2, characterized in that, Coordinated optimization of driving parameters and structural parameters for one-dimensional / two-dimensional / three-dimensional ultraviolet LED optical matrix structures, including: Based on the location and range of the target irradiation area, the corresponding ultraviolet LED beads are selectively turned on to avoid ineffective irradiation of non-target areas; Based on the target bactericidal dosage requirement In order to satisfy Under these conditions, the driving current of a single ultraviolet LED is preferentially controlled within the high luminous efficiency operating range, and the operation of the ultraviolet LED is adjusted by turning on the UV LED beads. Output power of a single ultraviolet LED bead and the moving speed of the device equipped with ultraviolet LED germicidal lamps Achieve dose compensation; among which, This represents the total amount of UV sterilization actually received by a single crop plant. When the local heat load increases, the junction temperature rise can be suppressed by reducing the driving current of the corresponding LED beads, increasing the speed of the cooling fan, or enhancing the heat exchange capacity of the heat sink. The target sterilization dose can be maintained by increasing the irradiation time or optimizing the combination of ultraviolet LED beads.

9. The ultraviolet sterilization method for facility agriculture based on nighttime optical phenotypes according to claim 1, characterized in that, Establish a dual protection mechanism that combines equipment operation and crop protection in a closed loop, including: Closed-loop operation of the equipment: If fluctuations in ultraviolet light intensity, spectral shifts, junction temperature of ultraviolet LED germicidal lamps, or dose deviations are detected to exceed the threshold, the driving parameters of ultraviolet LED germicidal lamps, heat dissipation strategies, and one-dimensional / two-dimensional / three-dimensional ultraviolet LED optical matrix structure control parameters are immediately and automatically adjusted to maintain stable operation of ultraviolet LED germicidal lamps and accurate disinfection dosage. Crop protection closed loop: If abnormal crop phenotypes are detected, the light intensity of the ultraviolet LED germicidal lamp is quickly reduced, the disinfection dosage is adjusted, or the ultraviolet LED germicidal lamp is turned off to avoid ultraviolet damage to the plants. At the same time, abnormal data is recorded to provide support for subsequent adaptive decision-making model iteration and optimization.

10. A platform comprising a ground mobile robot and a low-altitude near-canopy unmanned aerial vehicle, characterized in that, Ground mobile robots and low-altitude near-canopy drones serve as carriers for ultraviolet LED germicidal lamps, implementing the ultraviolet sterilization method based on nighttime optical phenotypes applicable to facility agriculture as described in any one of claims 1-9.

Images