Method for photovoltaic intercropping of jatropha curcas

Abstract

The application discloses a kind of photovoltaic shade interplanting methods of Amorphophallus konjac, comprising: establishing interplanting area between photovoltaic array row;According to photovoltaic panel projection range, interplanting area is divided into fixed shade area and variable light transmission area, fixed shade area is located in the area directly below photovoltaic panel, and variable light transmission area is located in the light transmission zone between adjacent photovoltaic panel;Planting ditch is opened according to row distance 40-50 centimeters in fixed shade area;Selecting Amorphophallus konjac tuber, planting in planting bed;Variable light transmission area is provided with sunshade net slide rail system driven by stepping motor, light intensity sensor is installed on slide rail, sensor is connected with controller, and controller calculates model according to solar elevation angle and real-time illumination data, and automatically adjusts sunshade net horizontal displacement daily;Establish intelligent drip irrigation system.The application can provide stable shade environment and suitable soil humidity for Amorphophallus konjac, reduce manual operation, improve the automation level of interplanting process, and meet the specific requirements of crops on growth environment.

Description

Technical Field

[0001] This invention relates to the field of agricultural planting technology. More specifically, this invention relates to a method for intercropping Amorphophallus bulbifera with photovoltaic shading. Background Technology

[0002] Against the backdrop of the coordinated development of agriculture and the photovoltaic industry, intercropping under photovoltaic arrays has attracted widespread attention due to its ability to achieve three-dimensional utilization of land resources. Amorphophallus bulbifera, a shade-tolerant crop with high economic value, has specific requirements for environmental parameters such as light intensity and soil moisture, necessitating meticulous cultivation under shading conditions. However, existing photovoltaic intercropping technologies face many unresolved issues regarding the microenvironmental control of Amorphophallus bulbifera.

[0003] In terms of light environment control, traditional photovoltaic intercropping models lack optimization for the shade tolerance characteristics of Amorphophallus viminalis in terms of the height and spacing of photovoltaic panel supports. If the support height is unreasonable or the photovoltaic panel spacing is inappropriate, it will lead to uneven light distribution within the intercropping area: the area directly below the photovoltaic panel may form a fixed shadow zone due to excessive shading, while the light-transmitting strips between adjacent photovoltaic panels may experience excessive fluctuations in light intensity at different times due to a lack of dynamic adjustment. Amorphophallus viminalis has a narrow suitable range of ground light intensity, but current technology struggles to adjust the shading degree of the light-transmitting strips in real time according to changes in the sun's position, often resulting in problems such as leaf scorching due to excessive light or impaired photosynthesis due to insufficient light.

[0004] Regarding soil moisture management, existing drip irrigation systems generally adopt a uniform irrigation strategy, failing to conduct stratified monitoring and differentiated control based on the root distribution characteristics of Amorphophallus bulbifera. Amorphophallus bulbifera requires a certain level of moisture in the topsoil, moderate moisture in the middle soil layer, and avoidance of waterlogging in the deeper soil layers. However, traditional drip irrigation systems rely solely on a single-depth moisture sensor, which cannot accurately obtain moisture content data for different soil layers, easily leading to over- or under-irrigation.

[0005] Therefore, it is necessary to design a technical solution that can overcome the above-mentioned defects. Summary of the Invention

[0006] One objective of this invention is to provide a method for intercropping Amorphophallus bulbifera with photovoltaic shading, which can provide a stable shading environment and suitable soil moisture for Amorphophallus bulbifera, reduce manual operation, improve the automation level of the intercropping process, and meet the specific requirements of the crop for its growth environment.

[0007] To achieve these objectives and other advantages of the present invention, according to one aspect of the present invention, a method for intercropping Amorphophallus bulbifera with photovoltaic shading is provided, comprising: S1: establishing an intercropping area between rows of a photovoltaic array, wherein the height of the photovoltaic panel support is set to 3-6 meters, and the north-south spacing between adjacent photovoltaic panels is set to 2.5-3.5 meters; S2: dividing the intercropping area into a fixed shading area and a variable light transmission area according to the projection range of the photovoltaic panels, wherein the fixed shading area is located directly below the photovoltaic panels, and the variable light transmission area is located in the light transmission zone between adjacent photovoltaic panels, wherein the width of the fixed shading area is 1.1-1.3 times the width of the photovoltaic panels; S3: digging planting trenches in the fixed shading area at a row spacing of 40-50 cm, with a trench depth of 15-20 cm, laying a 5-8 cm thick layer of decomposed straw in each trench, and covering the straw layer with a 2-3 cm thick layer. S4: Select bulbils weighing 15-25 grams each, disinfect them by soaking in 0.1% streptomycin sulfate solution, and then plant them in the planting bed; S5: Set up a shade net sliding rail system driven by a stepper motor in the variable light transmission zone. The shade net has a light transmittance of 50-60%. Install a light intensity sensor on the sliding rail. The sensor is connected to the controller. The controller automatically adjusts the horizontal displacement of the shade net every day according to the solar altitude angle calculation model and real-time light data to maintain the ground light intensity at 3000-5000 lux; S6: Establish an intelligent drip irrigation system. The drip irrigation tape is laid along the planting row. The spacing between drippers is consistent with the plant spacing. Each dripper is connected to an independent solenoid valve. A humidity sensor is buried in the soil. The humidity sensor is connected to the controller. When the soil moisture content is lower than 18%, the drip irrigation will start automatically.

[0008] Furthermore, the shading net sliding rail system integrates a GPS module and an infrared sensor. The controller has a built-in solar azimuth angle calculation model and performs the following operations every 30 minutes: calculate the theoretical solar altitude angle θ and azimuth angle φ based on geographical coordinates; obtain the actual position coordinates (x, y) of the shading net through the infrared sensor; calculate the shadow offset compensation amount ΔL=H·tanθ·sin(φ-α), where H is the height of the photovoltaic panel support and α is the azimuth angle of the photovoltaic panel; and drive the stepper motor to adjust the displacement of the shading net so that the actual irradiance of the light-transmitting area deviates from the target value by ≤±5%.

[0009] Furthermore, the stepper motor is equipped with a rotary encoder, and the controller executes a dynamic compensation algorithm: every 5 minutes, the encoder pulse count N is read, the actual displacement ΔL'=N×δ is calculated, where δ is the lead screw accuracy; the theoretical displacement ΔL is compared with ΔL', and if |ΔL-ΔL'|>0.02H, then the compensation command is output according to ΔL_correct=ΔL+0.5(ΔL-ΔL').

[0010] Furthermore, the stepper motor is equipped with a photovoltaic current sensor; the controller synchronously collects the output current I of the photovoltaic panel, and when it detects current fluctuation... At that time, adjust the position of the shade net in the opposite direction until the fluctuation rate is ≤3%.

[0011] Furthermore, the controller calculates the current change gradient ΔI / Δt. When ΔI / Δt > 5% / min, the reverse movement speed of the shade net is increased to 1.5 times the base speed. When the volatility drops to 5%, the position fine-tuning mode is activated, and exponential decay displacement compensation is performed according to the formula ΔL_fine = 0.2 × H × (1 - e^(-t / τ)), where τ = 3 minutes and t is the adjustment time. When the volatility is still > 3% after 3 consecutive adjustments, the movement of the shade net is paused.

[0012] Furthermore, in S6, the drip irrigation tape is connected to a three-dimensional electrode matrix sensor, and the controller executes a stratified irrigation strategy: when the surface soil moisture content WS < 18% and the middle soil moisture content WM ≥ 20%, the drip irrigation flow rate is calculated according to the formula Q = 1.2 + 0.3 × (18 - WS); when WM < 19%, regardless of the WS value, continuous irrigation at a flow rate of 1.8 L / h is started until WM recovers to 21%; if the deep soil moisture content WD > 25% is detected during irrigation, irrigation is immediately stopped and the drainage program is started.

[0013] Furthermore, the three-dimensional electrode matrix sensor integrates a soil permeability detection unit. When the controller starts the drainage program, it executes a dynamic drainage control strategy: when the deep water content WD > 25% is detected, irrigation is paused and soil permeability detection is initiated to obtain the real-time permeability K value; if K < 0.4 cm / h, the pulse drainage mode is activated, and drainage is alternately performed for 5 minutes and resting for 10 minutes at a cycle of T = 20 / K minutes until WD ≤ 22%; if 0.4 ≤ K ≤ 0.8 cm / h, the gradient drainage mode is activated, with an initial drainage flow rate Q = 1.2 L / h, increasing by 0.3 L / h every 10 minutes until WD ≤ 23%; if K > 0.8 cm / h, the rapid drainage mode is activated, continuously draining at a fixed flow rate of 2.0 L / h, while collecting the WD value every 2 minutes. When WD ≤ 24%, the system switches to drip irrigation compensation mode, replenishing water at a flow rate of 0.8 L / h until WD = 22%.

[0014] Furthermore, it also includes: S7: Install multispectral cameras in the planting area to collect images of bulbils morphology daily, and use image recognition algorithms to determine the development status of bulbil abscission layers. When an abscission layer with a diameter of 5-8 mm is detected, the controller triggers a harvest warning signal.

[0015] Furthermore, in the S7, the multispectral camera is equipped with a motorized rotating polarizing filter group, and the image recognition algorithm performs the following steps: automatically switching the polarization direction under morning diffused light conditions, acquiring three sets of bulbil images with different polarization angles of 0°, 60°, and 120°; and performing polarization differential processing on the 680nm band image. Eliminate specular reflection noise; calculate the delamination development index LDI=∫(550-680nm)dλ / ∫(680-800nm)dλ, where λ is the wavelength, and trigger the harvesting signal when LDI>2.2 and morphological diameter≥5mm.

[0016] Furthermore, the electrically rotating polarization filter group includes a near-infrared band interference filter, and the image recognition algorithm further performs the following: morphological opening operation is performed on the polarization difference image to eliminate noise with a diameter <0.5mm; the radius of curvature R of the bulbil edge contour is extracted, and when R≤0.8mm, it is determined as the delamination fracture initiation point; combined with the visible light-near infrared reflectance NIR / VIS=∫(700-900nm)dλ / ∫(400-700nm)dλ, when LDI>2.2 and 0.75≤NIR / VIS≤0.85, the confidence level of the acquired signal is increased to over 95%.

[0017] The present invention has at least the following beneficial effects:

[0018] This invention optimizes the parameters of photovoltaic panel support height and spacing, combined with a sensor-equipped dynamic adjustment system for shading nets. This allows the fixed shading area to provide a stable low-light environment, while the variable light-transmitting area, through a solar altitude angle calculation model linked with real-time light data, precisely maintains ground light intensity between 3000-5000 lux. It can dynamically adjust drip irrigation flow based on different soil moisture contents and designs a permeability-based drainage mode to address the risk of deep water accumulation, avoiding water waste or root rot problems caused by traditional uniform irrigation. Through multispectral image recognition and morphological algorithms, it achieves non-contact intelligent detection of bulbil abscission development, effectively improving bulbil integrity and marketability.

[0019] Other advantages, objectives and features of the present invention will become apparent in part from the following description, and in part from those skilled in the art through study and practice of the invention. Attached Figure Description

[0020] Figure 1 This is a flowchart of one embodiment of this application. Detailed Implementation

[0021] The present invention will now be described in further detail with reference to the accompanying drawings, so that those skilled in the art can implement it based on the description.

[0022] It should be understood that terms such as "having," "comprising," and "including" used in the embodiments of this application do not exclude the presence or addition of one or more other elements or combinations thereof. All directional indications (such as up, down, left, right, front, back, etc.) in the embodiments of this application are only used to explain the relative positional relationship and movement of components in a specific posture. If the specific posture changes, the directional indication will also change accordingly. When an element is referred to as "fixed to" or "set on" another element, it can be directly on the other element or may have an intervening element present. When an element is referred to as "connected to" another element, it can be directly connected to the other element or indirectly connected to the other element through an intervening element. Descriptions involving "first," "second," etc., in the embodiments of this application are for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of indicated technical features. Therefore, a feature defined with "first" or "second" may explicitly or implicitly include at least one of those features.

[0023] It should be noted that the technical solutions of the various embodiments of this application can be combined with each other, but only if they are based on the ability of those skilled in the art to implement them. When the combination of technical solutions is contradictory or cannot be implemented, it should be considered that such combination of technical solutions does not exist and is not within the scope of protection claimed by this application.

[0024] like Figure 1 As shown, an embodiment of this application provides a method for intercropping Amorphophallus bulbifera with photovoltaic shading, comprising: delineating an intercropping area between rows of photovoltaic arrays, setting the height of the photovoltaic panel support to 3-6 meters, and setting the north-south spacing between adjacent photovoltaic panels to 2.5-3.5 meters; dividing the intercropping area into a fixed shading area directly below the photovoltaic panel and a variable light-transmitting area between adjacent photovoltaic panels according to the projection range of the photovoltaic panel, wherein the width of the fixed shading area is 1.1-1.3 times the width of the photovoltaic panel; digging planting trenches 15-20 cm deep in the fixed shading area at a row spacing of 40-50 cm, laying a 5-8 cm thick layer of decomposed straw in each trench, and then covering the straw layer with a 2-3 cm thick layer of humus soil to form a planting bed; selecting single grains weighing 15-25 grams. The bulbils of *Amorphophallus konjac* were disinfected by soaking in a 0.1% streptomycin sulfate solution before being transplanted into the planting bed. A stepper motor-driven shade net rail system was installed in the variable light transmission zone. A shade net with a light transmittance of 50%-60% was selected. A light intensity sensor was installed on the rail and connected to the controller. The controller automatically adjusted the horizontal displacement of the shade net every day based on the solar altitude angle calculation model and real-time light data to maintain the ground light intensity at 3000-5000 lux. An intelligent drip irrigation system was set up, with drip irrigation tape laid along the planting rows. The spacing between drippers was consistent with the plant spacing. Each dripper was connected to an independent solenoid valve. A humidity sensor was buried in the soil and connected to the controller. Drip irrigation was automatically started when the soil moisture content was below 18%.

[0025] For example, the height of the photovoltaic panel support can be selected as 3 meters, 4.5 meters, or 6 meters according to actual needs; the spacing between adjacent photovoltaic panels can be set to 2.5 meters, 3 meters, or 3.5 meters; and the width of the fixed shading area is determined according to 1.1 times, 1.2 times, or 1.3 times the actual width of the photovoltaic panel. The row spacing of the planting trench can be selected as 40 cm, 45 cm, or 50 cm; the trench depth is 15 cm, 18 cm, or 20 cm; the thickness of the decomposed straw layer is 5 cm, 6 cm, 7 cm, or 8 cm; and the thickness of the humus cover is 2 cm or 3 cm. The single tuber weight of the golden konjac bulb is 15 grams, 20 grams, or 25 grams; and the concentration of the streptomycin sulfate solution is strictly controlled at 0.1%. In the shade net sliding rail system, the stepper motor can be the commonly available Dongfang Motor PK264-01A or Leadshine Technology 57byg series. The shade net can be a finished polyethylene net with a light transmittance of 50%, 55%, or 60%. The light intensity sensor can be a silicon photovoltaic cell sensor (such as BH1750) or a thermopile sensor (such as MLX90614), installed 20 cm above the middle of the sliding rail, facing the ground. The controller can be a Siemens S7-200 PLC or an Arduino microcontroller. The drip irrigation tape can be a PE material inlaid patch drip irrigation tape, with the dripper spacing matched to the plant spacing (approximately 30-40 cm). The solenoid valve can be a normally closed SMC VX series solenoid pulse valve. The humidity sensor can be a Decagon EC-5 capacitive soil moisture sensor, with the probe vertically inserted 10-15 cm below the planting bed into the main root distribution area.

[0026] First, install the photovoltaic panel supports according to the design parameters and adjust the spacing to divide the area into fixed shading zones and variable light transmission zones. Next, dig planting trenches, lay down well-rotted straw and humus, and complete the tuber disinfection and planting. Then, install the shading net sliding rail system and drip irrigation system, connecting and debugging the sensors and controller. During system operation, the light intensity sensor provides real-time data feedback, and the controller, combined with the solar altitude angle model, calculates the appropriate position of the shading net and drives the stepper motor for adjustment. The humidity sensor monitors the soil moisture content, triggering drip irrigation when it falls below 18%, thus achieving automatic control of light and water. This method provides a stable shading environment and suitable soil moisture for Amorphophallus bulbifera, reducing manual operation, improving the automation level of the intercropping process, and meeting the specific requirements of the crop's growth environment.

[0027] In another embodiment, the shading net sliding rail system integrates a GPS module and an infrared sensor. The controller has a built-in solar azimuth angle calculation model and performs the following operations every 30 minutes: calculate the theoretical solar altitude angle θ and azimuth angle φ based on geographical coordinates; obtain the actual position coordinates (x, y) of the shading net through the infrared sensor; calculate the shadow offset compensation amount using the formula ΔL=H•tanθ•sin(φ-α) (H is the height of the photovoltaic panel support, and α is the azimuth angle of the photovoltaic panel); and drive the stepper motor to adjust the displacement of the shading net so that the deviation between the actual irradiance of the light-transmitting area and the target value is controlled within ±5%.

[0028] For example, the GPS module can be a u-blox NEO-6M or SIM2003 module to acquire longitude, latitude, and time information in real time; the infrared sensor can be a Sharp GP2Y0A21YKOF infrared ranging sensor or a VL53L0X laser ranging module, installed at both ends and the middle of the slide rail, to detect the position coordinates of the shade net on the x-axis (east-west) and y-axis (north-south). The controller can be an Advantech UNO-2483 industrial control computer or a Raspberry Pi 4B, with a built-in solar azimuth angle calculation model based on astronomical algorithms.

[0029] Solar altitude angle θ = arcsin(sinδ·sinφ + cosδ·cosφ·cosH);

[0030] Azimuth φ = arccos[(sinδ·cosφ-cosδ·sinφ·cosH) / cosθ];

[0031] Where δ is the solar declination (δ=0.3723+23.2567sin(0.9856N-1.4815)-0.1149sin(2×0.9856N-1.9992)-0.1712sin(3×0.9856N-2.4380), N is the number of days in a year), φ is the local latitude, and H is the hour angle (H=15°×(t-12), t is the local time). The infrared sensor is fixed on both sides of the slide rail by a bracket, 10 cm away from the moving plane of the shading net, and provides real-time position data feedback.

[0032] Every 30 minutes, the system obtains the current geographical coordinates and time via GPS, calculates the solar altitude and azimuth angles using astronomical algorithms, and simultaneously detects the current position of the shading net using infrared sensors. The controller calculates the compensation amount ΔL according to a formula and sends pulse signals to the stepper motor to adjust the displacement. This method can track changes in the sun's position in real time, dynamically correct the position of the shading net, reduce light fluctuations caused by solar offset, make the irradiance of the light-transmitting area closer to the target value, improve the accuracy and stability of light control, and create more uniform light conditions for the growth of Amorphophallus bulbifera.

[0033] In another embodiment, the stepper motor is equipped with a rotary encoder, and the controller executes a dynamic compensation algorithm: every 5 minutes, the encoder pulse count N is read, and the actual displacement ΔL' = N × δ (δ is the lead screw accuracy) is calculated; the theoretical displacement ΔL is compared with the actual displacement ΔL', and if the absolute value of the difference between the two exceeds 0.02H (H is the height of the photovoltaic panel support), then a compensation command is output according to ΔL_correct = ΔL + 0.5(ΔL - ΔL').

[0034] For example, the stepper motor can be a Leadshine DM542 stepper motor paired with a 57BYGH250A motor body; the rotary encoder can be an Omron E6B2-CWZ6C incremental encoder (resolution 360 pulses / revolution), coaxially connected to the stepper motor output shaft via a flexible coupling; the lead screw can be a TBI ground-grade ball screw, with a lead accuracy δ commonly 0.5 mm / revolution, 1 mm / revolution, or 2 mm / revolution, selected according to the system accuracy requirements. The controller can be a Panasonic FP0-C32T PLC or a Googol Technology GT-400-SV motion control card, equipped with high-speed pulse counting function. When installing the encoder, ensure it is concentric with the motor shaft to avoid measurement errors caused by eccentricity. The lead screw is fixed on the brackets on both sides of the slide rail, and the sunshade net is connected to the lead screw nut via a slider.

[0035] During operation, the controller reads the encoder pulse count N every 5 minutes, multiplies it by the lead screw pitch accuracy δ to obtain the actual displacement ΔL' (in millimeters), and compares it with the theoretical displacement ΔL (calculated based on the sun's position). If |ΔL - ΔL'| > 0.02H (e.g., when H = 2 meters, the threshold is 40 millimeters), then a compensation command ΔL_correct is calculated. By increasing or decreasing the stepper motor pulse count, the actual displacement of the shading net approaches the theoretical value. This dynamic compensation mechanism effectively corrects displacement errors caused by factors such as lead screw backlash and load changes during stepper motor operation, improves the accuracy of shading net position control, ensures accurate light intensity adjustment, and avoids light instability problems caused by mechanical errors.

[0036] In another embodiment, the stepper motor is equipped with a photovoltaic current sensor, and the controller synchronously collects the output current I of the photovoltaic panel. When current fluctuation is detected... At time (I0 is the current value at the previous moment), adjust the position of the shading net in the opposite direction until the fluctuation rate is ≤3%.

[0037] For example, the photovoltaic current sensor can be a LEM LA25-NP Hall current sensor or an ACS712 current detection module, connected in series in the DC output circuit of the photovoltaic panel for real-time acquisition of the output current; the controller can be a Mitsubishi FX3U-32MR PLC or an STM32 microcontroller, equipped with an analog input channel (12-bit resolution) and a sampling frequency of 1 time / second. When installing the current sensor, pay attention to the positive and negative polarity, ensure good shielding of the signal transmission line, and avoid electromagnetic interference. The current fluctuation rate is calculated as follows: When the volatility exceeds 8%, it indicates that the shading net may be excessively blocking the photovoltaic panels. The controller sends a reverse movement command to the stepper motor (e.g., if it was originally moving east, it will be moving west). Each adjustment is 5 centimeters, until the volatility drops to below 3%.

[0038] During operation, the system continuously monitors changes in the output current of the photovoltaic panels while adjusting the position of the shading net to meet the crop's light requirements. If an abnormally large increase in current fluctuation is detected, it indicates that the shading net is obstructing the photovoltaic panels and affecting power generation efficiency. By adjusting the position of the shading net in reverse to reduce the obstructed area, the system maintains stable operation of the photovoltaic system while ensuring sufficient light intensity for the Amorphophallus bulbifera. This coordinated control method achieves a balance between "crop growth needs" and "photovoltaic power generation efficiency," avoiding a decline in the performance of one objective due to adjustment of a single objective, and improving the overall efficiency of the photovoltaic intercropping system.

[0039] In another embodiment, the controller calculates the current change gradient ΔI / Δt. When ΔI / Δt > 5% / min, the reverse movement speed of the shade net is increased to 1.5 times the base speed. When the volatility drops to 5%, the position fine-tuning mode is activated, and exponential decay displacement compensation is performed according to the formula ΔL_fine = 0.2 × H × (1 - e^(-t / τ)) (τ = 3 minutes, t is the adjustment time). If the volatility is still > 3% after 3 consecutive adjustments, the movement of the shade net is paused.

[0040] The specific parameters and control logic are as follows: The base speed is determined based on the rated speed of the stepper motor and the lead screw. For example, if the motor speed is 100 rpm and the lead screw lead is 10 mm / rpm, the base speed is 1000 mm / min. The current change gradient ΔI / Δt is calculated using the current values ​​at two adjacent moments (1 minute apart), in units of % / min. When ΔI / Δt > 5% / min, it indicates that the current fluctuation is aggravated. By increasing the stepper motor pulse frequency, the moving speed of the shading net is increased to 1.5 times the base speed (i.e., 1500 mm / min), quickly reducing shading. When the fluctuation rate drops to 5%, the fine-tuning mode is entered. The displacement compensation decreases exponentially with time t. For example, when t=0, ΔL_fine=0.2H, and when t=3 minutes, ΔL_fine=0.2H×(1-e^(-1))≈0.126H, gradually reducing the adjustment amplitude to avoid system oscillation. If the volatility still exceeds 3% after three consecutive adjustments (each with a 5-minute interval), the controller sends a stop command to prevent the stepper motor from frequently operating and causing mechanical wear.

[0041] In practical applications, the system first judges the severity of fluctuations by the current change gradient. When rapid changes occur, the adjustment speed is increased to respond promptly to abnormal photovoltaic panel output. When the fluctuations subside, it switches to fine-tuning mode to precisely control the position of the shading net. If multiple adjustments are ineffective, the system pauses operation to avoid excessive intervention. This tiered adjustment strategy balances adjustment speed and precision, making shading net position adjustments more stable and efficient, reducing the impact on the photovoltaic system and crop growth environment, and improving the system's robustness.

[0042] In another embodiment, the drip irrigation tape is connected to a three-dimensional electrode matrix sensor, and the controller executes a stratified irrigation strategy: when the surface soil moisture content WS < 18% and the middle soil moisture content WM ≥ 20%, the drip irrigation flow rate (unit: L / h) is calculated according to Q = 1.2 + 0.3 × (18 - WS); when WM < 19%, regardless of the WS value, continuous irrigation at a flow rate of 1.8 L / h is started until WM recovers to 21%; if the deep soil moisture content WD > 25% is detected during irrigation, irrigation is immediately stopped and the drainage program is started.

[0043] For example, the three-dimensional electrode matrix sensor consists of three sets of resistive moisture sensors: a surface electrode (detecting the 0-10 cm soil layer), a middle electrode (detecting the 15-25 cm soil layer), and a deep electrode (detecting the 30-40 cm soil layer). The electrodes are made of stainless steel (1 mm in diameter and 5 cm in length), and each set of electrodes is spaced 10 cm apart, arranged in a matrix on both sides of the planting row. The drip irrigation tape is an inlaid cylindrical drip irrigation tape (PE material, 0.4 mm wall thickness), with a dripper spacing of 30 cm and a flow coefficient Kv=0.8. Each dripper is connected to an independent solenoid valve (ASCO 8262 series normally closed valve) via a φ16 mm PE branch pipe. The controller uses a Delta DVP-32ES2 PLC with 4 analog input channels, acquiring sensor voltage signals and converting them into moisture content (%). When the surface layer is dry (WS < 18%) but the middle layer is moist (WM ≥ 20%), calculate the flow rate according to the formula. For example, if WS = 15%, then Q = 1.2 + 0.3 × (18 - 15) = 2.1 L / h. When the middle layer is dry (WM < 19%), irrigate directly at a flow rate of 1.8 L / h until WM reaches 21%. If the deep layer moisture content exceeds 25%, immediately close all solenoid valves and start the drainage pump.

[0044] The sensor collects the water content of each layer every 10 minutes, and the controller adjusts the drip irrigation flow or stops irrigation according to preset logic. This tiered irrigation strategy is tailored to the root distribution characteristics of Amorphophallus bulbifera. When the surface layer is dry, water is added as needed to avoid seepage into deeper layers; when the middle layer is dry, priority is given to ensuring water supply to the root layer; and when deep layers are waterlogged, damage is stopped in time to reduce the risk of root rot, providing a more suitable water environment for crop growth and improving water resource utilization efficiency.

[0045] In another embodiment, the three-dimensional electrode matrix sensor integrates a soil permeability detection unit. When the drainage program is started, the controller executes a dynamic drainage control strategy: when the deep water content WD > 25% is detected, irrigation is paused and permeability detection is initiated to obtain the real-time permeability K value; if K < 0.4 cm / h, the pulse drainage mode is activated, and drainage is alternately carried out for 5 minutes and then left to stand for 10 minutes at a cycle of T = 20 / K minutes until WD ≤ 22%; if 0.4 ≤ K ≤ 0.8 cm / h, the gradient drainage mode is activated, with an initial flow rate of 1.2 L / h, increasing by 0.3 L / h every 10 minutes until WD ≤ 23%; if K > 0.8 cm / h, the rapid drainage mode is activated, and drainage is continuously carried out at a flow rate of 2.0 L / h, with WD collected every 2 minutes. When it reaches 24%, the system switches to drip irrigation compensation mode, replenishing water at a rate of 0.8 L / h until WD = 22%.

[0046] For example, the soil permeability testing unit uses a constant-pressure infiltrator (such as the Soilmoisture Equipment model 1500), which calculates the K value by measuring the rate at which water seeps into the soil under a certain pressure. The probe is buried in the deep soil layer (30-40 cm). The drainage equipment includes a centrifugal pump (flow rate 5 L / min) and a drainage pipe (PVC material, inner diameter 25 mm, laid 5 cm below the deep soil layer, with a slope of 0.5%). In the pulse drainage mode, if K = 0.2 cm / h, T = 20 / 0.2 = 100 minutes, that is, drainage for 5 minutes and resting for 95 minutes, to avoid compaction of low-permeability soil due to continuous drainage; in the gradient drainage mode, starting from 1.2 L / h, the flow rate increases by 0.3 L / h every 10 minutes until WD ≤ 23%, which is suitable for medium-permeability soil; the rapid drainage mode is used for high-permeability soil. During continuous drainage, WD is detected at high frequency. When it reaches 24%, drip irrigation is switched to compensate, and water is replenished at 0.8 L / h to the target value of 22% to prevent over-drainage.

[0047] During actual drainage, the system first detects the deep soil moisture content. If it exceeds the standard, irrigation is suspended and permeability detection is initiated. The drainage mode is selected based on the K value. During the drainage process, the WD change is monitored in real time. Once the standard is met, it switches to compensatory irrigation. This dynamic strategy can adapt to different soil conditions, avoid soil structure damage or water imbalance caused by improper drainage, effectively control the deep soil moisture content, ensure root respiration and nutrient absorption, and reduce the occurrence of diseases.

[0048] In another embodiment, a multispectral camera is installed in the planting area to collect images of bulbils daily. The development status of the bulbil abscission layer is determined by an image recognition algorithm. When an abscission layer with a diameter of 5-8 mm is detected, the controller triggers a harvest warning signal.

[0049] For example, a Basler acA2000-50gm camera with a 400-900nm wavelength filter wheel can be used as the multispectral camera. It is mounted on an aluminum alloy bracket 3-4 meters high, with the lens pointing vertically downwards, covering a 2×2 meter planting area. Images are automatically acquired daily at 9:00 AM (under diffused light conditions). The image recognition algorithm is implemented using the OpenCV library. First, the image is converted to grayscale. The Otsu thresholding method is used to extract the bulbil region. Then, contour detection is performed, and the equivalent diameter of each connected region is calculated. When a circular or elliptical region with a diameter between 5-8 mm is detected, and its edges are clear and its color differs significantly from the surrounding tissue, it is determined to be a delamination formation. The controller outputs a warning signal (audio-visual alarm or SMS notification) via a relay.

[0050] The camera collects images daily at set times and transmits them to the industrial computer built into the controller for processing. The algorithm automatically identifies the size of the bulbils' abscission layer, replacing the subjectivity and lag of manual observation. This non-contact detection method can monitor maturity in real time, promptly reminding users of the harvesting time, avoiding excessive growth or shedding of bulbils due to human error, improving the efficiency and accuracy of harvesting management, and providing data support for mechanized harvesting.

[0051] In another embodiment, the multispectral camera is equipped with an electrically rotating polarizing filter group, and the image recognition algorithm performs the following steps: automatically switching the polarization direction under morning diffused light conditions, acquiring bulb images with three sets of polarization angles of 0°, 60°, and 120°; and performing polarization difference processing on the 680nm band image. Eliminate specular reflection noise; calculate the delamination development index LDI=∫(550-680nm)dλ / ∫(680-800nm)dλ, and trigger the harvesting signal when LDI>2.2 and morphological diameter≥5mm.

[0052] For example, the electrically driven rotating polarizing filter assembly uses Thorlabs' CVI-Melles Griot polarizing wheel, which includes optical filters with three polarization directions: 0°, 60°, and 120°. It is mounted 5 cm in front of the camera lens and driven by a miniature stepper motor (1 rpm). The 680nm band image is acquired through the camera's band selection function. During polarization differential processing, the three sets of images are first registered, and then the ΔI value for each pixel is calculated to generate an image after de-reflection noise. LDI is calculated by integrating the light intensity values ​​of the 550-680nm band (visible red region) and the 680-800nm ​​band (near-infrared region). When LDI > 2.2 and the morphological diameter ≥ 5 mm, the abscission layer is considered mature, triggering a harvesting signal.

[0053] The system automatically switches polarization direction to acquire images in the morning (when the solar altitude angle is <30° and scattered light is dominant). It uses polarization difference to eliminate specular reflection from the waxy layer on the leaf surface, thus improving image quality. By using both LDI and morphological diameter for determination, it reduces interference from leaf texture, stains, and other factors. Compared with single morphology detection, it improves the accuracy of delamination identification, reduces the probability of misjudgment, and provides a more reliable basis for accurate harvesting.

[0054] In another embodiment, the electrically rotating polarization filter group includes a near-infrared band interference filter, and the image recognition algorithm further performs the following: morphological opening operation (structural element 3×3 pixels) on the polarization difference image to eliminate noise with a diameter <0.5mm; extracting the radius of curvature R of the bulbil edge contour, and determining the delamination fracture initiation point when R≤0.8mm; combining NIR / VIS=∫(700-900nm)dλ / ∫(400-700nm)dλ, when LDI>2.2 and 0.75≤NIR / VIS≤0.85, the confidence level of the acquired signal is increased to over 95%.

[0055] For example, the near-infrared interference filter uses a filter with a center wavelength of 750nm and a bandwidth of 50nm, which is coaxially mounted with the polarization filter and switched via a filter wheel. Morphological opening operations use 3×3 pixel rectangular structural elements, first eroding and then dilating to remove small noise points; edge contour extraction uses the Canny operator, calculating the radius of curvature R for each point on the contour (calculated by fitting a circle to three adjacent points), and taking the minimum value as the curvature feature of the contour. The NIR / VIS reflectance is calculated by integrating the reflected light intensity in the near-infrared (700-900nm) and visible (400-700nm) bands. When LDI > 2.2, NIR / VIS is between 0.75 and 0.85, and R ≤ 0.8mm, the delamination is comprehensively judged to be mature, with a confidence level exceeding 95%, avoiding misjudgment based on a single parameter.

[0056] Based on polarization difference processing, the system purifies images through morphological operations, extracts radius of curvature and reflectance parameters, and performs multi-layer verification by combining multispectral information and morphological features. This multi-dimensional detection method can effectively distinguish between layered and similar morphological tissue defects, improve detection reliability, provide high-precision signals for harvesting decisions, reduce missed or false detections caused by environmental noise or algorithm errors, and ensure the harvesting quality and yield of bulbils.

[0057] Although embodiments of the present invention have been disclosed above, they are not limited to the applications listed in the specification and embodiments. They can be applied to various fields suitable for the present invention. For those skilled in the art, other modifications can be easily made. Therefore, without departing from the general concept defined by the claims and their equivalents, the present invention is not limited to the specific details and illustrations shown and described herein.

Claims

1. A method for intercropping Amorphophallus bulbifera with photovoltaic shading, characterized in that, include: S1: Establish intercropping areas between rows of photovoltaic arrays, with the height of photovoltaic panel supports set at 3-6 meters and the north-south spacing between adjacent photovoltaic panels set at 2.5-3.5 meters; S2: Based on the projection range of the photovoltaic panels, the planting area is divided into a fixed shading area and a variable light transmission area. The fixed shading area is located directly below the photovoltaic panels, and the variable light transmission area is located in the light transmission strip between adjacent photovoltaic panels. The width of the fixed shading area is 1.1-1.3 times the width of the photovoltaic panels. S3: In the fixed shaded area, dig planting furrows with a row spacing of 40-50 cm and a furrow depth of 15-20 cm. Lay a layer of well-rotted straw with a thickness of 5-8 cm in each furrow, and cover the straw layer with a layer of humus soil with a thickness of 2-3 cm to form a planting bed. S4: Select bulbils of golden konjac tubers with a single tuber weight of 15-25 grams, disinfect them by soaking in 0.1% streptomycin sulfate solution, and then transplant them into the planting bed. S5: A stepper motor driven shading net slide rail system is set in the variable light transmission zone. The light transmittance of the shading net is 50-60%. A light intensity sensor is installed on the slide rail. The sensor is connected to the controller. The controller automatically adjusts the horizontal displacement of the shading net every day according to the solar altitude angle calculation model and real-time light data to maintain the ground light intensity at 3000-5000 lux. S6: Establish an intelligent drip irrigation system. Drip irrigation tape is laid along the planting row, and the spacing between drippers is consistent with the spacing between plants. Each dripper is connected to an independent solenoid valve. A humidity sensor is buried in the soil and connected to the controller. Drip irrigation is automatically started when the soil moisture content is lower than 18%. The stepper motor is equipped with a photovoltaic current sensor; the controller synchronously collects the output current I of the photovoltaic panel, and when the current fluctuation rate |ΔI / I0|>8% is detected, the position of the shading net is adjusted in the opposite direction until the fluctuation rate is ≤3%; The controller calculates the current change gradient ΔI / Δt. When ΔI / Δt > 5% / min, the reverse movement speed of the shade net increases to 1.5 times the base speed. When the volatility drops to 5%, the position fine-tuning mode is activated, and exponential decay displacement compensation is performed according to the formula ΔL_fine = 0.2 × H × (1 - e^(-t / τ)), where τ = 3 minutes and t is the adjustment time. If the volatility is still > 3% after 3 consecutive adjustments, the movement of the shade net is paused. In S6, the drip irrigation tape is connected to a three-dimensional electrode matrix sensor, and the controller executes a stratified irrigation strategy: when the surface soil moisture content WS < 18% and the middle soil moisture content WM ≥ 20%, the drip irrigation flow rate is calculated according to the formula Q = 1.2 + 0.3 × (18 - WS); when WM < 19%, regardless of the WS value, continuous irrigation at a flow rate of 1.8 L / h is started until WM recovers to 21%; if the deep soil moisture content WD > 25% is detected during irrigation, irrigation is immediately stopped and the drainage program is started. The three-dimensional electrode matrix sensor integrates a soil permeability detection unit. When the drainage program is initiated, the controller executes a dynamic drainage control strategy: when the deep water content (WD) is detected to be greater than 25%, irrigation is paused and soil permeability detection is activated to obtain the real-time permeability K value; if K < 0.4 cm / h, a pulse drainage mode is activated, alternating between drainage for 5 minutes and resting for 10 minutes at a cycle of T = 20 / K minutes until WD ≤ 22%; if 0.4 ≤ K ≤ 0.8 cm / h, a gradient drainage mode is activated, with an initial drainage flow rate Q = 1.2 L / h, increasing by 0.3 L / h every 10 minutes until WD ≤ 23%; if K > 0.8 cm / h, a rapid drainage mode is activated, continuously draining at a fixed flow rate of 2.0 L / h, while simultaneously collecting the WD value every 2 minutes. When WD ≤ 24%, the system switches to drip irrigation compensation mode, replenishing water at a flow rate of 0.8 L / h until WD = 22%. Also includes: S7: Multispectral cameras are set up in the planting area to collect images of bulbils daily. The development status of the bulbils abscission layer is determined by the image recognition algorithm. When an abscission layer with a diameter of 5-8 mm is detected, the controller triggers a harvest warning signal. In the S7, the multispectral camera is equipped with a motorized rotating polarizing filter group. The image recognition algorithm performs the following steps: automatically switching the polarization direction under morning diffused light conditions, acquiring three sets of bulbil images with different polarization angles of 0°, 60°, and 120°; and performing polarization difference processing on the 680nm band image: ΔI=|I0-I 60 |+|I 60 -I 120 | Eliminate specular reflection noise; calculate the delamination development index LDI=∫(550-680nm)dλ / ∫(680-800nm)dλ, where λ is the wavelength, and trigger the harvesting signal when LDI>2.2 and morphological diameter≥5mm.

2. The method for intercropping Amorphophallus bulbifera with photovoltaic shading as described in claim 1, characterized in that, The shade net sliding rail system integrates a GPS module and an infrared sensor. The controller has a built-in solar azimuth angle calculation model and performs the following operations every 30 minutes: The solar altitude angle θ and azimuth angle φ are calculated based on geographic coordinates. The actual position coordinates (x, y) of the shade net are obtained using an infrared sensor; Calculate the shadow offset compensation amount ΔL = H·tanθ·sin(φ-α), where H is the height of the photovoltaic panel support and α is the azimuth angle of the photovoltaic panel; Drive a stepper motor to adjust the displacement of the shading net so that the actual irradiance of the light-transmitting area deviates from the target value by ≤±5%.

3. The method for intercropping Amorphophallus bulbifera with photovoltaic shading as described in claim 1, characterized in that, The stepper motor is equipped with a rotary encoder, and the controller executes a dynamic compensation algorithm: Read the encoder pulse count N every 5 minutes and calculate the actual displacement ΔL'=N×δ, where δ is the lead screw accuracy. Compare the theoretical displacement ΔL with ΔL'. If |ΔL-ΔL'|>0.02H, then output the compensation command as ΔL_correct=ΔL+0.5(ΔL-ΔL').

4. The method for intercropping Amorphophallus bulbifera with photovoltaic shading as described in claim 1, characterized in that, The electrically rotating polarizing filter group includes a near-infrared band interference filter, and the image recognition algorithm further performs: morphological opening operation on the polarization difference image to eliminate noise with a diameter <0.5mm; Extract the radius of curvature R of the bulbil edge contour. When R ≤ 0.8 mm, it is determined as the starting point of delamination fracture. Combining the visible-near-infrared reflectance ratio NIR / VIS=∫(700-900nm)dλ / ∫(400-700nm)dλ, when LDI>2.2 and 0.75≤NIR / VIS≤0.85, the confidence level of the acquired signal can be increased to over 95%.

Images