Rubber forest self-adaptive light supplementing system based on environmental perception

Abstract

The invention relates to the technical field of agricultural automation and intellectualization, in particular to a rubber forest self-adaptive light supplementing system based on environmental perception. The system is composed of an environment sensing module, a data processing and control module, a light supplementing execution module and a communication module. According to the rubber forest self-adaptive light supplementing system and method based on environment perception, through deep coupling of multi-source environment perception and intelligent decision, the static regulation and control bottleneck of a traditional light supplementing system is broken through, dynamic collaborative optimization of light supplementing intensity, spectrum composition and space-time distribution is achieved, and the photosynthetic efficiency and latex yield of rubber trees are remarkably improved.

Description

Technical Field

[0001] The present invention relates to the technical field of agricultural automation and intelligence, and particularly to an adaptive light supplement system for rubber plantations based on environmental perception. Background Art

[0002] As an important natural rubber production base, the yield of rubber plantations is significantly affected by light conditions. Traditional light supplement systems mostly adopt static regulation modes with fixed spectral ratios and constant light intensities, and cannot adapt to the complex and changeable environmental conditions of rubber plantations and the physiological requirements of the dynamic growth of rubber trees.

[0003] Especially in practical applications, the existing technologies have the following core defects: lacking a real-time coupling mechanism for multi-dimensional environmental parameters and the physiological state of rubber trees, resulting in a low spatio-temporal matching degree between the light supplement strategy and the canopy light demand, and being unable to synergistically optimize the photosynthetic photon flux density (PPFD), spectral band ratio, and light field distribution, causing low light energy utilization efficiency and frequent light inhibition effects, and being difficult to dynamically respond to sudden environmental mutations (such as extreme high temperatures and strong wind disturbances). This directly leads to key problems such as lag in light environment regulation, uneven light reception of the canopy, and low energy efficiency ratio of latex synthesis metabolism in the existing light supplement systems, seriously restricting the improvement of rubber plantation production capacity and sustainable development. Summary of the Invention

[0004] The present invention provides an adaptive light supplement system and method for rubber plantations based on environmental perception. Through the deep coupling of multi-source environmental perception and intelligent decision-making, it breaks through the static regulation bottleneck of traditional light supplement systems and realizes the dynamic collaborative optimization of light supplement intensity, spectral composition, and spatio-temporal distribution, significantly improving the photosynthetic efficiency of rubber trees and latex production.

[0005] The technical solution adopted by the present invention is: an adaptive light supplement system for rubber plantations based on environmental perception, which is composed of an environmental perception module, a data processing and control module, a light supplement execution module, and a communication module.

[0006] The environmental perception module is used to collect multi-dimensional environmental data of rubber plantations in real time; the data processing and control module is connected to the environmental perception module and is built-in with an adaptive light supplement algorithm for dynamically generating a light supplement strategy; the light supplement execution module is controlled by the data processing and control module and includes an LED light supplement device array with adjustable spectra and light intensities; the communication module realizes data transmission and remote monitoring among the modules.

[0007] Among them, the system realizes the collaborative optimization of light supplement intensity, spectral composition, and time distribution by integrating multi-source environmental parameters and rubber tree physiological state data.

[0008] As a further improvement of the present invention, the environmental perception module is composed of a canopy layer image acquisition unit, a rubber tree phenological period monitoring unit, and a variety of sensors.

[0009] As a further improvement of the present invention, the canopy image acquisition unit is equipped with a machine learning algorithm to identify chlorophyll content and pest and disease characteristics; the phenological period monitoring unit of the rubber tree detects the suitable tapping period based on the micro-changes of the bark; the sensors include a light intensity sensor, a multi-spectral sensor, a temperature and humidity sensor, a soil nutrient sensor, and a wind speed and direction sensor.

[0010] As a further improvement of the present invention, the adaptive supplementary lighting algorithm adopts a three-layer decision-making architecture to achieve dynamic regulation:

[0011] Data fusion layer: Through feature vector extraction and time series analysis, the light intensity, chlorophyll fluorescence value, canopy morphological parameters, and phenological period status are encoded into a multi-dimensional decision matrix.

[0012] Decision optimization layer: Based on the deep reinforcement learning framework, a light environment response model is constructed. Combining the light saturation point and light compensation point of the rubber tree, the optimal photosynthetic photon flux density (PPFD) is dynamically calculated, and a genetic algorithm is used to optimize the spectral ratio scheme, where the proportion of the blue light (450 - 470 nm) and far-red light (730 - 750 nm) bands is non-linearly regulated with the phenological period.

[0013] Execution correction layer: A closed-loop feedback mechanism based on fuzzy PID control is established. The uniformity of the light field is corrected in real time through the distributed light intensity detection unit of the supplementary lighting device array to ensure that the PPFD deviation rate of the illuminated surface of the canopy is ≤ 8%.

[0014] The algorithm updates the supplementary lighting strategy every 15 minutes, and automatically triggers the red light band attenuation protection mechanism when the canopy temperature is detected to exceed 32°C for three consecutive cycles.

[0015] As a further improvement of the present invention, the light environment response model integrates a rubber tree variety-specific parameter library, sets different light morphogenesis response curves for different strains, and introduces a latex yield prediction module to optimize economic benefits through regression analysis of the content of non-structural carbohydrates (NSC) in leaves and the duration of supplementary lighting.

[0016] As a further improvement of the present invention, the supplementary lighting execution module adopts a flexible support structure imitating the sun's trajectory, and its LED array includes a six-channel mixed light unit that can be independently controlled, supporting continuous dimming of 200 - 800 μmol·m-2·s-1, and the spectrum covers the range of 380 - 850 nm, where the 660 nm red light and 730 nm far-red light channels have a wavelength fine-tuning function of ±5 nm.

[0017] As a further improvement of the present invention, the flexible bracket is configured with a three-dimensional pan-tilt mechanism, which can dynamically adjust the pitch angle (0-60°) and azimuth angle (0-360°) according to the growth of the canopy height, and cooperate with the pulse width modulation (PWM) technology to achieve the light intensity gradient distribution at different sites of the canopy.

[0018] As a further improvement of the present invention, the communication module adopts a 5G / LoRaWAN dual-mode communication protocol and is built-in with an Internet of Things (IoT) platform interface.

[0019] A working method of a rubber forest adaptive supplementary lighting system based on environmental perception includes the following steps:

[0020] S1, System initialization stage: Load the strain-specific light response parameter library of rubber trees through the communication module, and initialize the reference spectral ratio of the LED supplementary lighting device array and the initial positioning angle of the three-dimensional pan-tilt of the bracket.

[0021] S2, Multi-source data synchronous acquisition: The canopy image acquisition unit captures the morphological characteristics of the canopy at a frequency of 20 Hz. At the same time, the phenological period monitoring unit scans the density of bark microcracks, combines with a multispectral sensor to obtain the chlorophyll fluorescence value, and the temperature and humidity sensor and the light intensity sensor upload environmental data at 1-minute intervals.

[0022] S3, Data fusion and feature encoding: After the collected raw data is processed by wavelet denoising, extract the canopy projection area, leaf inclination distribution, and chlorophyll spatial heterogeneity index as feature vectors, and generate a 16-dimensional time series decision matrix in combination with the phenological period state encoding.

[0023] S4, Dynamic light demand calculation: In the decision optimization layer, call the deep reinforcement learning model to predict the light quantum flux demand in the next 2 hours, combine with the real-time canopy temperature-photosynthetic rate response surface, dynamically correct the PPFD target value, and iteratively optimize the spectral weight ratio of red, blue, and far-red light through the genetic algorithm.

[0024] S5, Supplementary lighting space adaptation: Drive the three-dimensional pan-tilt mechanism of the flexible bracket according to the canopy height growth model, calculate the optimal pitch angle (5°-55°) and azimuth angle deviation compensation amount of each supplementary lighting unit, and synchronously generate a PWM duty cycle gradient mapping table.

[0025] S6, Closed-loop light field regulation: After the supplementary lighting is executed, collect the actual PPFD distribution of the illuminated surface of the canopy through the distributed light intensity detection unit, and use the fuzzy PID algorithm to dynamically adjust the drive current of each LED channel to converge the light intensity deviation rate to a preset threshold.

[0026] S7, Protection mechanism trigger: When the canopy temperature > 32°C and the relative humidity < 60% are monitored for three consecutive cycles, the red light band attenuation program is automatically started, reducing the intensity of the 660nm channel to 30%-50% of the reference value, and at the same time increasing the far-red light ratio by 10%-15%;

[0027] S8, Strategy iterative update: Re-evaluate the environmental parameters and device status every 15 minutes. When sudden strong winds (wind speed > 8m / s) or device attitude instability (tilt error > 5°) are detected, the supplementary lighting is immediately interrupted and the safety retraction mechanism is started.

[0028] Advantages of the present invention: (1) By integrating multi-source environmental parameters (light, temperature and humidity, wind speed, soil nutrients, etc.) and physiological state data of rubber trees (chlorophyll content, canopy morphology, phenological period, etc.), and combining deep reinforcement learning and genetic algorithms, the present invention realizes the dynamic optimization of supplementary lighting intensity, spectral band ratio, and time-space distribution, accurately matching the photosynthetic photon flux density (PPFD) with the photosynthetic demand of rubber trees, effectively improving the light energy utilization rate, and significantly reducing the incidence of photoinhibition effect.

[0029] (2) Based on the flexible bracket and three-dimensional pan-tilt mechanism imitating the sun's trajectory, and cooperating with the fuzzy PID closed-loop control technology, the present invention realizes a lower PPFD deviation rate on the light-receiving surface of the canopy, ensures the dynamic adaptation of the light intensity gradient distribution to the canopy morphology, improves the photosynthetic rate, and enhances the efficiency of latex synthesis metabolism.

[0030] (3) Through the real-time feedback mechanism and protection strategy in the three-layer decision-making architecture, the present invention can respond to sudden environmental changes such as extreme high temperature within 15 minutes, trigger red light band attenuation or device safety retraction, reduce the risk of light damage and equipment failure rate, and improve system stability.

[0031] (4) The present invention adopts a six-channel adjustable LED array and ±5nm wavelength fine-tuning technology, combined with the phenological period-driven non-linear ratio strategy of blue light and far-red light, which can dynamically adjust the photomorphogenesis signal, promote the directional distribution of photosynthetic products of rubber trees to latex synthesis, and significantly improve the latex yield under the same energy consumption; through the regression analysis of the leaf NSC content and the supplementary lighting duration, the dynamic balance between supplementary lighting cost and economic benefits is achieved. Description of the drawings

[0032] Figure 1 is the system block diagram of an adaptive supplementary lighting system for rubber forests based on environmental perception according to the present invention;

[0033] Figure 2 is the working method flow chart of an adaptive supplementary lighting system for rubber forests based on environmental perception according to the present invention. Detailed implementation manners

[0034] In order to make the technical problems, technical solutions and beneficial effects to be solved by this application more clear and understandable, the following further details this application in combination with the accompanying drawings and embodiments. It should be understood that the embodiments described herein are only used to explain this application and are not used to limit this application.

[0035] The present invention provides an adaptive supplementary lighting system for rubber plantations based on environmental perception, which consists of an environmental perception module, a data processing and control module, a supplementary lighting execution module, and a communication module.

[0036] The environmental perception module is used to collect multi-dimensional environmental data of the rubber plantation in real time; the data processing and control module is connected to the environmental perception module and has an adaptive supplementary lighting algorithm built in, which is used to dynamically generate a supplementary lighting strategy; the supplementary lighting execution module is controlled by the data processing and control module and includes an LED supplementary lighting device array with adjustable spectrum and light intensity; the communication module realizes data transmission and remote monitoring among the modules. Among them, the system realizes the collaborative optimization of the supplementary lighting intensity, spectral composition and time distribution by integrating multi-source environmental parameters and rubber tree physiological state data.

[0037] In the present invention, the environmental perception module is composed of a canopy layer image acquisition unit, a rubber tree phenological period monitoring unit and a variety of sensors. The canopy layer image acquisition unit is equipped with a machine learning algorithm to identify chlorophyll content and pest and disease characteristics; the rubber tree phenological period monitoring unit detects the suitable period for rubber tapping based on the micro-changes in the bark; the sensors include a light intensity sensor, a multi-spectral sensor, a temperature and humidity sensor, a soil nutrient sensor, and a wind speed and direction sensor.

[0038] In the present invention, the adaptive supplementary lighting algorithm adopts a three-layer decision-making architecture to achieve dynamic regulation: (1) Data fusion layer: Through feature vector extraction and time series analysis, light intensity, chlorophyll fluorescence value, canopy morphological parameters, and phenological stage status are encoded into a multi-dimensional decision matrix; (2) Decision optimization layer: Based on the deep reinforcement learning framework, a light environment response model is constructed. Combining the light saturation point and light compensation point of rubber trees, the optimal photosynthetic photon flux density (PPFD) is dynamically calculated, and a genetic algorithm is used to optimize the spectral ratio scheme, where the proportion of the blue light (450 - 470 nm) and far-red light (730 - 750 nm) bands is non-linearly regulated with the phenological stage; (3) Execution correction layer: A closed-loop feedback mechanism based on fuzzy PID control is established. The uniformity of the light field is corrected in real time through the distributed light intensity detection unit of the supplementary lighting device array to ensure that the PPFD deviation rate of the illuminated surface of the canopy ≤ 8%. The algorithm updates the supplementary lighting strategy every 15 minutes, and when the canopy temperature exceeds 32 °C for three consecutive cycles, the red light band attenuation protection mechanism is automatically triggered. The light environment response model integrates a rubber tree variety-specific parameter library, sets different light morphogenesis response curves for different strains, and introduces a latex yield prediction module to optimize economic benefits through the regression analysis of the content of leaf non-structural carbohydrates (NSC) and the supplementary lighting duration.

[0039] In the present invention, the supplementary lighting execution module adopts a flexible support structure imitating the sun's trajectory. Its LED array includes a six-channel mixed light unit that can be independently controlled, supports continuous dimming of 200 - 800 μmol·m-2·s-1, and the spectrum covers the range of 380 - 850 nm. Among them, the 660 nm red light and 730 nm far-red light channels have a wavelength fine-tuning function of ±5 nm. The flexible support is configured with a three-dimensional pan-tilt mechanism, which can dynamically adjust the pitch angle (0 - 60°) and azimuth angle (0 - 360°) according to the growth of the canopy height, and cooperate with the pulse width modulation (PWM) technology to realize the light intensity gradient distribution at different positions of the canopy.

[0040] In the present invention, the communication module adopts a 5G / LoRaWAN dual-mode communication protocol and is built with an Internet of Things (IoT) platform interface.

[0041] A working method of a rubber forest adaptive supplementary lighting system based on environmental perception includes the following steps:

[0042] S1, System initialization stage: Load the rubber tree strain-specific light response parameter library through the communication module, and initialize the reference spectral ratio of the LED supplementary lighting device array and the initial positioning angle of the three-dimensional pan-tilt of the support.

[0043] S2, Multi-source data synchronous acquisition: The canopy image acquisition unit captures the morphological characteristics of the canopy at a frequency of 20 Hz. Meanwhile, the phenological period monitoring unit scans the density of bark micro-cracks, combines with a multi-spectral sensor to obtain chlorophyll fluorescence values, and the temperature and humidity sensors and light intensity sensors upload environmental data at 1-minute intervals;

[0044] S3, Data fusion and feature encoding: After wavelet denoising processing of the collected raw data, extract the canopy projection area, leaf inclination distribution, and chlorophyll spatial heterogeneity index as feature vectors, and generate a 16-dimensional time series decision matrix in combination with the phenological period state encoding;

[0045] S4, Dynamic light demand calculation: In the decision optimization layer, call the deep reinforcement learning model to predict the light quantum flux demand in the next 2 hours, combine with the real-time canopy temperature - photosynthetic rate response surface, dynamically correct the PPFD target value, and iteratively optimize the spectral weight ratio of red, blue, and far-red light through the genetic algorithm;

[0046] S5, Supplementary light space adaptation: Drive the three-dimensional pan-tilt mechanism of the flexible bracket according to the canopy height growth model, calculate the optimal pitch angle (5° - 55°) and azimuth angle deviation compensation amount of each supplementary light unit, and synchronously generate a PWM duty cycle gradient mapping table;

[0047] S6, Closed-loop light field regulation: After supplementary lighting, collect the actual PPFD distribution of the illuminated surface of the canopy through the distributed light intensity detection unit, and use the fuzzy PID algorithm to dynamically adjust the drive current of each LED channel to converge the light intensity deviation rate to a preset threshold;

[0048] S7, Protection mechanism trigger: When it is continuously monitored for 3 cycles that the canopy temperature > 32 °C and the relative humidity < 60%, automatically start the red light band attenuation program, reduce the intensity of the 660 nm channel to 30% - 50% of the reference value, and at the same time increase the proportion of far-red light by 10% - 15%;

[0049] S8, Policy iterative update: Re-evaluate the environmental parameters and device status every 15 minutes. When sudden strong winds (wind speed > 8 m / s) or device attitude instability (tilt error > 5°) are detected, immediately interrupt the supplementary lighting and start the safety withdrawal mechanism.

[0050] Embodiment:

[0051] This embodiment takes a rubber plantation base in Hainan Province as the implementation scenario to specifically illustrate the application process and technical effects of this system.

[0052] (I) System deployment

[0053] Deploy this system in a 10-hectare rubber forest test area, and the specific configuration is as follows:

[0054] Environmental perception module: Install a canopy image acquisition unit (equipped with a 5-megapixel CMOS camera and an infrared fill light) at intervals of 50 m, install a phenological period monitoring unit (integrated with a macro camera and a laser rangefinder) on the trunk of each rubber tree, and evenly arrange 20 groups of sensor nodes in the forest (including a light intensity sensor, a multispectral sensor, a temperature and humidity sensor, a soil nutrient sensor, and a wind speed and direction sensor).

[0055] Light compensation execution module: Adopt a flexible bracket structure imitating the sun's trajectory, with the brackets arranged in a grid pattern at intervals of 8 m × 8 m. Each bracket is equipped with a six-channel LED array (red light 660 nm ± 5 nm, blue light 450 nm ± 5 nm, far red light 730 nm ± 5 nm, white light, UV-A 380 nm, IR 850 nm), the single lamp power is 200 W, the PPFD adjustment range is 200 - 800 μmol·m-2·s-1, and the three-dimensional pan-tilt of the bracket is driven by a stepper motor (angle resolution 0.1°).

[0056] Communication module: Adopt a hybrid networking of 5G base stations and LoRaWAN gateways, with the data transmission delay ≤ 200 ms to ensure that control instructions are issued in real time.

[0057] (2) System operation process

[0058] Step 1, initialization and parameter loading:

[0059] Import the light response parameter library of the rubber tree variety "Reyan 7-33-97" through the IoT platform, including: (1) Light saturation point: 1200 μmol·m-2·s-1; (2) Light compensation point: 50 μmol·m-2·s-1; (3) Phenological period - spectral ratio mapping table (such as the blue light proportion is 35% - 45% during the leaf emergence period, and the far red light proportion is 20% - 30% during the tapping period).

[0060] Initialize the reference spectrum of the LED array as 40% blue light / 45% red light / 15% far red light, and the initial pitch angle of the bracket pan-tilt is 30° and the azimuth angle is 180°.

[0061] Step 2, multi-source data acquisition:

[0062] The canopy image acquisition unit captures canopy images at a frequency of 20 Hz, and calculates the chlorophyll content (SPAD value) and the proportion of the area of pest and disease patches in real time through the ResNet-18 convolutional neural network (recognition accuracy > 92%).

[0063] The phenological period monitoring unit detects that the density of microcracks on the bark reaches 0.8 cracks / cm 2 (threshold 0.5 cracks / cm 2 ), and determines that the best tapping period has entered.

[0064] The chlorophyll fluorescence parameter Fv / Fm measured by the multispectral sensor is 0.78 (normal range 0.75 - 0.85), the canopy temperature is 29.5°C, and the light intensity is natural light PPFD 550 μmol·m-2·s-1.

[0065] Step 3, Data fusion and decision generation:

[0066] After wavelet denoising the above data, the following feature vectors are extracted: (1) Canopy projection area: 12.6 m 2 ; (2) Average leaf inclination angle: 42° ± 8°; (3) Chlorophyll spatial heterogeneity index: 0.15.

[0067] Combined with the phenological period coding to generate a 16-dimensional decision matrix, and input it into the deep reinforcement learning model (DRL-A3C algorithm) to predict the light demand in the next 2 hours. The target PPFD output by the model is 680 μmol·m-2·s-1, and the spectral ratio is optimized by the genetic algorithm to be 38% blue light / 50% red light / 12% far red light.

[0068] Step 4, Supplementary light space adaptation:

[0069] According to the canopy height growth model (current canopy height 8.2 m), calculate that the optimal pitch angle of the bracket pan-tilt is 48° ± 2°, and the azimuth compensation amount is +15°. Generate a PWM duty cycle gradient mapping table to increase the light intensity in the canopy edge area by 12% (duty cycle 75%), and the duty cycle in the center area is 60%.

[0070] Step 5, Closed-loop light field regulation:

[0071] After the supplementary light is started, the distributed light intensity detection unit measures the actual PPFD distribution to be 665 - 720 μmol·m-2·s-1, and the maximum deviation rate is 7.3%. The fuzzy PID controller dynamically adjusts the LED drive current, and after 3 iterations (consuming 45 seconds), the deviation rate converges to 5.8%, meeting the requirement of ≤8%.

[0072] Step 6, Sudden environment response:

[0073] When continuous strong wind is detected (wind speed 9.2 m / s, lasting for 2 minutes), the system immediately interrupts the supplementary light, and the bracket pan-tilt automatically retracts to a safe posture (pitch angle 0°), and at the same time starts the mechanical locking device.

[0074] Implementation effect verification

[0075] After three months of continuous operation testing, the comparison with the traditional supplementary light system is as follows in the table:

[0076] Index The system of the present invention Traditional system Improvement rate Light energy utilization rate 82％ 58％ +41.4％ PPFD deviation rate ≤8％ 25％-40％ -68％ Latex yield (kg / plant / month) 6.8 5.7 +19.3％ Energy consumption (kWh / ha / day) 42 65 -35.4％ Equipment failure rate 0.7 times / month 2.5 times / month -72％

[0077] Typical scenario application

[0078] When the canopy temperature rises to 33.2 °C at 14:00 in the afternoon (exceeding the standard for 3 consecutive cycles), the system automatically triggers the red light attenuation mechanism: (1) the intensity of the 660 nm red light channel drops to 40% of the reference value; (2) the proportion of 730 nm far-red light increases to 18%; (3) the spray cooling system is synchronously started (linked with the supplementary lighting module).

[0079] This operation causes the canopy temperature to drop back to 30.5 °C within 15 minutes, and the chlorophyll fluorescence parameter Fv / Fm remains at 0.79, effectively avoiding the photoinhibition effect.

[0080] This embodiment verifies the adaptive regulation ability of the system of the present invention under multi-environment coupling conditions. By dynamically optimizing the light environment parameters, it significantly improves the photosynthetic efficiency and latex yield of the rubber forest, while reducing energy consumption and operation and maintenance costs.

[0081] In summary, an adaptive supplementary lighting system for rubber forests based on environmental perception of the present invention realizes precise regulation of the light environment of rubber forests by integrating advanced environmental perception technologies, intelligent algorithms and efficient supplementary lighting equipment. This system not only improves the photosynthesis efficiency and latex yield of rubber trees, but also significantly reduces energy consumption and equipment failure rates, providing strong support for the sustainable development of the rubber planting industry. In the future, with the continuous progress of technologies such as the Internet of Things and artificial intelligence, the performance and intelligent level of this system are expected to be further improved, opening up new ways for the efficient management and precise operation of rubber forests.

[0082] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. An adaptive light supplement system for rubber plantations based on environmental perception, characterized in that, It consists of an environmental perception module, a data processing and control module, a supplementary lighting execution module, and a communication module. The environmental perception module is used to collect multi-dimensional environmental data of the rubber plantation in real time; the data processing and control module is connected to the environmental perception module and has an adaptive supplementary lighting algorithm built-in, which is used to dynamically generate a supplementary lighting strategy; the supplementary lighting execution module is controlled by the data processing and control module and includes an LED supplementary lighting device array with adjustable spectrum and light intensity; the communication module realizes data transmission and remote monitoring among modules. Among them, the system realizes the coordinated optimization of supplementary lighting intensity, spectral composition, and time distribution by fusing multi-source environmental parameters and rubber tree physiological state data.

2. The adaptive supplementary lighting system for rubber plantations based on environmental perception according to claim 1, wherein, The environmental perception module consists of a canopy layer image acquisition unit, a rubber tree phenological period monitoring unit, and a variety of sensors.

3. The adaptive supplementary lighting system for rubber plantations based on environmental perception according to claim 2, characterized in that, The canopy layer image acquisition unit is equipped with a machine learning algorithm to identify chlorophyll content and pest and disease characteristics; the rubber tree phenological period monitoring unit detects the suitable tapping period based on the micro-changes in the bark; the sensors include a light intensity sensor, a multi-spectral sensor, a temperature and humidity sensor, a soil nutrient sensor, and a wind speed and direction sensor.

4. An adaptive supplementary lighting system for rubber plantations based on environmental perception according to claim 1, wherein, The adaptive supplementary lighting algorithm uses a three-layer decision-making architecture to achieve dynamic regulation: Data fusion layer: By extracting feature vectors and time series analysis, the light intensity, chlorophyll fluorescence value, canopy morphology parameters, and phenological period status are encoded into a multi-dimensional decision matrix. Decision optimization layer: Based on the deep reinforcement learning framework, a light environment response model is constructed. Combining the light saturation point and light compensation point of rubber trees, the optimal photosynthetic photon flux density (PPFD) is dynamically calculated, and the genetic algorithm is used to optimize the spectral ratio scheme. Among them, the proportion of the blue light (450 - 470nm) and far red light (730 - 750nm) bands is non-linearly regulated with the phenological period. Execution correction layer: A closed-loop feedback mechanism based on fuzzy PID control is established. The uniformity of the light field is corrected in real time through the distributed light intensity detection unit of the supplementary lighting device array to ensure that the PPFD deviation rate of the illuminated surface of the canopy ≤ 8%. The algorithm updates the supplementary lighting strategy every 15 minutes, and automatically triggers the red light band attenuation protection mechanism when the canopy temperature is detected to exceed 32°C for three consecutive cycles.

5. The adaptive supplementary lighting system for rubber plantations based on environmental perception according to claim 4, characterized in that, The light environment response model integrates a rubber tree variety-specific parameter library, sets different light morphogenesis response curves for different strains, and introduces a latex yield prediction module to optimize economic benefits through regression analysis of the content of leaf non-structural carbohydrates (NSC) and supplementary lighting duration.

6. The adaptive supplementary lighting system for rubber plantations based on environmental perception according to claim 1, wherein The supplementary lighting execution module adopts a flexible support structure imitating the sun's trajectory. Its LED array includes a six-channel mixed light unit that can be independently controlled, supports continuous dimming of 200 - 800 μmol·m-2·s-1, and the spectral coverage ranges from 380 - 850nm. Among them, the 660nm red light and 730nm far red light channels have a wavelength fine-tuning function of ±5nm.

7. An environment perception-based adaptive light supplement system for rubber plantations according to claim 6, characterized in that, The flexible support is configured with a three-dimensional pan-tilt mechanism, which can dynamically adjust the pitch angle (0 - 60°) and azimuth angle (0 - 360°) according to the growth of the canopy height, and cooperate with the pulse width modulation (PWM) technology to realize the light intensity gradient distribution at different positions of the canopy.

8. The adaptive supplementary lighting system for rubber plantations based on environmental perception according to claim 1, wherein The communication module adopts a 5G / LoRaWAN dual-mode communication protocol and has an interface for the Internet of Things (IoT) platform built in.

9. The working method of an adaptive light supplement system for rubber plantations based on environmental perception according to any one of claims 1-8, characterized in that, It includes the following steps: S1. System initialization phase: Load the rubber tree strain-specific light response parameter library through the communication module, and initialize the reference spectral ratio of the LED supplementary lighting device array and the initial positioning angle of the three-dimensional cloud platform of the bracket. S2. Multi-source data synchronous acquisition: The canopy image acquisition unit captures the canopy morphological characteristics at a frequency of 20 Hz. At the same time, the phenological period monitoring unit scans the density of bark microcracks, combines with a multispectral sensor to obtain the chlorophyll fluorescence value, and the temperature and humidity sensor and the light intensity sensor upload environmental data at 1-minute intervals. S3. Data fusion and feature encoding: After the collected raw data is processed by wavelet denoising, extract the canopy projection area, leaf inclination distribution, and chlorophyll spatial heterogeneity index as feature vectors, and generate a 16-dimensional time series decision matrix in combination with the phenological period state encoding. S4. Dynamic light demand calculation: In the decision optimization layer, call the deep reinforcement learning model to predict the light quantum flux demand in the next 2 hours, combine with the real-time canopy temperature-photosynthetic rate response surface, dynamically correct the PPFD target value, and iteratively optimize the spectral weight ratio of red, blue, and far red light through the genetic algorithm. S5. Supplementary lighting space adaptation: Drive the three-dimensional cloud platform mechanism of the flexible bracket according to the canopy height growth model, calculate the optimal pitch angle (5°-55°) and azimuth angle deviation compensation amount of each supplementary lighting unit, and synchronously generate a PWM duty cycle gradient mapping table. S6. Closed-loop light field regulation: After supplementary lighting is executed, collect the actual PPFD distribution of the canopy illuminated surface through the distributed light intensity detection unit, and use the fuzzy PID algorithm to dynamically adjust the drive current of each LED channel to make the light intensity deviation rate converge to the preset threshold. S7. Protection mechanism trigger: When it is continuously monitored for 3 cycles that the canopy temperature > 32 °C and the relative humidity < 60%, automatically start the red light band attenuation program, reduce the intensity of the 660 nm channel to 30%-50% of the reference value, and at the same time increase the proportion of far red light by 10%-15%. S8. Policy iteration update: Re-evaluate the environmental parameters and device status every 15 minutes. When sudden strong wind (wind speed > 8 m / s) or device attitude instability (tilt error > 5°) is detected, immediately interrupt the supplementary lighting and start the safety withdrawal mechanism.

Images