A decision support method and system for efficient irrigation and fertilization of greenhouse tomatoes

Abstract

The present invention discloses a decision support method and system for efficient irrigation and fertilization of greenhouse tomatoes, which specifically relates to the field of agricultural precision irrigation technology. The method and system are used to solve the problem that the existing balanced fertilizer supply mode cannot respond to the short-term fertilizer demand peak during the fruit expansion period, resulting in supply lag and resource waste. The method obtains the fruit diameter and generates a phosphorus pulse start signal when the expansion threshold is reached; the thickness change gradient of the ion diffusion boundary layer on the root surface is synchronously measured, and the boundary layer correction coefficient is marked according to the steep change rate; the rhizosphere microscopic phase field is simulated to calculate the curvature distribution of the phosphate diffusion path, and the phase field topological entropy value is generated; based on the phosphorus pulse start signal, the phase field topological entropy value and the boundary layer correction coefficient, the optimal injection time point of the phosphorus fertilizer is calculated by querying a conversion table through the quotient value; and at the optimal time point, a phosphorus fertilizer injection instruction containing the fertilizer code is output to a drip irrigation execution device, effectively reducing the loss of soil nutrient immobilization.

Description

Technical Field

[0001] The present invention relates to the technical field of agricultural precision irrigation, and more specifically, to a method and system for supporting decision-making for efficient irrigation and fertilization of facility tomatoes. Background Art

[0002] In facility tomato cultivation, integrated water and fertilizer systems have been widely used in irrigation and fertilization management; the current mainstream decision-making method relies on a preset fertilization plan for the growth period, and adopts a daily average balanced supply model to regulate nutrient ratios and dosages; this model is based on soil basic fertility indicators and the theoretical full-cycle fertilizer requirements of crops, and achieves stable infusion of fertilizer solution through a drip irrigation system.

[0003] However, the above model has defects in dealing with the fertilizer demand characteristics of tomato fruit expansion: the continuous and uniform fertilizer supply logic cannot respond to the short-term concentrated peak demand for nutrients, which leads to physiological disorders due to delayed supply during the critical development window of the fruit, and the redundant supply during the non-peak period aggravates the loss of soil immobilization, causing the decision-making system to deviate from biological reality, resulting in a double loss of resource efficiency and output quality. Summary of the Invention

[0004] In order to overcome the above-mentioned defects of the prior art, the present invention provides a method and system for supporting decision-making on efficient irrigation and fertilization of greenhouse tomatoes to solve the problems raised in the above-mentioned background technology.

[0005] To achieve the above object, the present invention provides the following technical solutions:

[0006] A decision support method for efficient irrigation and fertilization of greenhouse tomatoes, comprising the following steps:

[0007] S1. Obtaining the fruit diameter of a target tomato plant in the fruit expansion stage;

[0008] S2. When the fruit diameter reaches the set expansion threshold, a phosphorus pulse start signal is generated;

[0009] S3. When a phosphorus pulse start signal is generated, the thickness gradient of the ion diffusion boundary layer on the root surface of the target tomato plant is synchronously measured. When the steepness of the thickness gradient exceeds the empirical threshold of the corresponding soil type, the corresponding steepness is marked as the boundary layer correction coefficient.

[0010] S4. By simulating the three-dimensional microscopic phase field of the rhizosphere of the target tomato plant, the curvature distribution characteristics of the phosphate ion diffusion path set are calculated, and the phase field topological entropy value representing the disorder of the effective diffusion flux is generated;

[0011] S5. Calculating the optimal injection time of phosphate fertilizer according to the phosphate pulse starting signal, the phase field topological entropy value, and the boundary layer correction coefficient;

[0012] S6. Outputting a phosphate fertilizer injection instruction to the drip irrigation execution device at the optimal phosphate fertilizer injection time point.

[0013] Furthermore, obtaining the fruit diameter of the target tomato plant in the fruit expansion stage includes:

[0014] Capturing fruit images of target tomato plants through a camera device fixed to a facility support;

[0015] Color space conversion and threshold segmentation are used to separate the fruit area from the fruit image;

[0016] Extract the minimum circumscribed circle boundary of the fruit area;

[0017] The diameter of the smallest circumscribed circle was taken as the measurement of the fruit diameter.

[0018] Furthermore, the color space conversion adopts the RGB to HSV conversion mode, the threshold segmentation is set based on the hue value range corresponding to the fruit maturity, and the minimum circumscribed circle boundary is generated by iteratively fitting the contour point set.

[0019] Furthermore, when the fruit diameter reaches a set swelling threshold, a phosphorus pulse start signal is generated, including:

[0020] Compare the fruit diameter obtained in real time with the set swelling threshold;

[0021] When the fruit diameter is greater than or equal to the set swelling threshold for the first time, the timer is triggered to start recording the duration;

[0022] If the fruit diameter continues to be greater than or equal to the set swelling threshold value for more than a first preset time period, a phosphorus pulse start signal is generated.

[0023] Furthermore, the swelling threshold is pre-stored in a database according to the characteristics of the tomato variety, and the first preset time duration is dynamically adjusted according to the ambient temperature and light intensity of the greenhouse where the target tomato plant is located. The ambient temperature and light intensity are obtained in real time through sensors deployed in the greenhouse.

[0024] Furthermore, when a phosphorus pulse start signal is generated, the thickness gradient of the ion diffusion boundary layer on the root surface of the target tomato plant is synchronously measured. When the steepness of the thickness gradient exceeds the empirical threshold of the corresponding soil type, the corresponding steepness is marked as the boundary layer correction coefficient, including:

[0025] The microelectrode array measurement is started at the moment when the phosphorus pulse start signal is generated;

[0026] The phosphate ion concentration value is continuously captured by a microelectrode array arranged within a specified distance range on the root surface of the target tomato plant;

[0027] The thickness change of the ion diffusion boundary layer per unit time is calculated based on the concentration difference between adjacent time points;

[0028] The rate of change of thickness variation with time is defined as the steepness rate of thickness gradient;

[0029] Retrieve the pre-calibrated empirical threshold corresponding to the soil type;

[0030] When the steep change rate is greater than the empirical threshold of the current soil type, the current steep change rate is marked as the boundary layer correction coefficient.

[0031] Furthermore, by simulating the three-dimensional microscopic phase field of the rhizosphere of the target tomato plant, the curvature distribution characteristics of the phosphate ion diffusion path set were calculated, and the phase field topological entropy value representing the disorder of the effective diffusion flux was generated, including:

[0032] The rhizosphere geometric model was constructed based on the three-dimensional spatial coordinate data of the target tomato plant root system reconstructed by X-ray tomography;

[0033] Mapping soil porosity distribution data to the geometric model to form solid-liquid-gas three-phase boundary conditions;

[0034] The Cahn-Hilliard phase field equation was used to simulate the diffusion process of phosphate ions in the rhizosphere.

[0035] Extract the instantaneous curvature values ​​of all phosphate ion motion trajectories in the simulation results;

[0036] Calculate the probability density distribution of the instantaneous curvature value in the preset angle range;

[0037] Calculate the phase field topological entropy value based on the probability density distribution:

[0038] The phase field topological entropy is equal to the negative value of the sum of the product of the probability density of each angle interval and the natural logarithm of the corresponding probability density;

[0039] The preset angle interval covers the range from zero radians to 2π radians, and the instantaneous curvature value is solved through the differential geometric characteristics of the trajectory curve.

[0040] Furthermore, the optimal injection time of phosphate fertilizer is calculated based on the phosphate pulse starting signal, the phase field topological entropy value and the boundary layer correction coefficient, including:

[0041] Obtaining the generation time of the phosphorus pulse start signal as the reference time point;

[0042] Calculate the quotient obtained by dividing the phase field topological entropy by the boundary layer correction coefficient;

[0043] The preset conversion relationship table is queried according to the quotient value to obtain the time offset;

[0044] The optimal injection time of phosphate fertilizer is obtained by adding the reference time point and the time offset;

[0045] The preset conversion relationship table is generated based on historical cultivation data training and stored in the local database.

[0046] Furthermore, outputting a phosphate fertilizer injection instruction to the drip irrigation execution device at the optimal phosphate fertilizer injection time point includes:

[0047] Continuously obtain the current system time and compare it with the optimal injection time of phosphate fertilizer in real time;

[0048] When the current system time reaches the optimal injection time point of phosphate fertilizer for the first time, a phosphate fertilizer injection instruction including a timestamp and a fertilizer identifier is generated;

[0049] transmitting the phosphate fertilizer injection instruction to the control unit of the drip irrigation execution device through the wired communication interface;

[0050] The phosphate fertilizer injection instruction includes the execution time point and the potassium dihydrogen phosphate fertilizer type code. The control unit activates the fertilizer injection pump in the drip irrigation pipeline according to the instruction.

[0051] In another aspect, the present invention provides a decision support system for efficient irrigation and fertilization of greenhouse tomatoes, comprising the following modules:

[0052] a diameter detection module for obtaining the fruit diameter of a target tomato plant in the fruit expansion stage;

[0053] A signal trigger module is used to generate a phosphorus pulse start signal when the fruit diameter reaches a set swelling threshold;

[0054] The gradient monitoring module is used to synchronously measure the thickness gradient of the ion diffusion boundary layer on the root surface of the target tomato plant when the phosphorus pulse start signal is generated. When the steepness of the thickness gradient exceeds the empirical threshold of the corresponding soil type, the corresponding steepness is marked as the boundary layer correction coefficient;

[0055] The entropy analysis module is used to simulate the three-dimensional microscopic phase field of the rhizosphere of the target tomato plant, calculate the curvature distribution characteristics of the phosphate ion diffusion path set, and generate the phase field topological entropy value that represents the disorder of the effective diffusion flux;

[0056] The decision-making operation module is used to calculate the optimal injection time of phosphate fertilizer based on the phosphate pulse start signal, the phase field topological entropy value and the boundary layer correction coefficient;

[0057] The execution output module is used to output the phosphate fertilizer injection instruction to the drip irrigation execution device at the optimal phosphate fertilizer injection time point.

[0058] Compared with the prior art, the present invention has the following beneficial effects:

[0059] 1. The present invention captures the biological signal of the fruit diameter exceeding the expansion threshold in real time, accurately triggers the phosphorus pulse start-up mechanism, changes the static decision-making logic of traditional daily fertilizer supply, synchronizes the timing of phosphorus fertilizer supply with the critical period of fruit cell division, and implements targeted and enhanced supply during the peak window of root absorption activity in response to the short-term surge in phosphorus demand during the expansion period, eliminating physiological obstacles such as fruit deformity and insufficient dry matter accumulation caused by delayed nutrient supply from the source.

[0060] 2. By integrating the dynamic coupling analysis of the root surface ion diffusion boundary layer gradient change and the rhizosphere micro-region phase field entropy value, a decision-making model with soil environment adaptability is constructed. The boundary layer correction coefficient reflects the blocking effect of soil texture differences on ion migration in real time, and the phase field topological entropy value quantifies the effective diffusion efficiency of phosphate at the root-soil interface. The time offset calculation driven by the two synergistically significantly improves the targeting of phosphate fertilizer migration, breaking through the traditional regulation limitations of the soil-crop system separation, so that the injection timing simultaneously meets the dual constraints of root absorption dynamics and soil ion transport efficiency, thereby reducing phosphate immobilization loss; the conversion relationship table trained based on historical data realizes the self-evolution of decision rules. By continuously accumulating fertilization response data under different soil types and climatic conditions, the time offset mapping rules are dynamically optimized. Under the premise of ensuring the demand for fruit enlargement, the frequency of unnecessary fertilization is reduced, and a dual leap in resource efficiency and fruit quality is simultaneously achieved. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] Figure 1 This is a flow chart of a decision support method for efficient irrigation and fertilization of facility tomatoes according to the present invention;

[0062] Figure 2 This is a structural schematic diagram of a decision support system for efficient irrigation and fertilization of facility tomatoes according to the present invention. DETAILED DESCRIPTION

[0063] The following will provide a clear and complete description of the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0064] Example 1: Figure 1 The present invention provides a decision support method for efficient irrigation and fertilization of greenhouse tomatoes, which includes the following steps:

[0065] S1. Obtaining the fruit diameter of a target tomato plant in the fruit expansion stage;

[0066] S2. When the fruit diameter reaches the set expansion threshold, a phosphorus pulse start signal is generated;

[0067] S3. When a phosphorus pulse start signal is generated, the thickness gradient of the ion diffusion boundary layer on the root surface of the target tomato plant is synchronously measured. When the steepness of the thickness gradient exceeds the empirical threshold of the corresponding soil type, the corresponding steepness is marked as the boundary layer correction coefficient.

[0068] S4. By simulating the three-dimensional microscopic phase field of the rhizosphere of the target tomato plant, the curvature distribution characteristics of the phosphate ion diffusion path set are calculated, and the phase field topological entropy value representing the disorder of the effective diffusion flux is generated;

[0069] S5. Calculating the optimal injection time of phosphate fertilizer according to the phosphate pulse starting signal, the phase field topological entropy value, and the boundary layer correction coefficient;

[0070] S6. Outputting a phosphate fertilizer injection instruction to the drip irrigation execution device at the optimal phosphate fertilizer injection time point.

[0071] S1. Obtaining the fruit diameter of a target tomato plant in the fruit expansion stage, specifically implemented as follows:

[0072] In a greenhouse for cultivating tomatoes, a high-definition camera is fixedly mounted at a predetermined position on a crop growth support. This position ensures that the central axis of the camera lens is horizontally aligned with the main fruit cluster of a target tomato plant. The camera is positioned vertically from the surface of the fruit of the target tomato plant, for example, within a range of 50 to 80 centimeters. The pan / tilt angle is adjusted so that the fruit occupies, for example, more than 60% of the central area of ​​the image frame. The camera captures color images at a frequency of one frame per minute, in a red, green, and blue three-channel 8-bit depth bitmap format, with each frame having a resolution of no less than 1920 pixels by 1080 pixels. The collected red, green and blue three-channel images are input into the color conversion module, which performs mathematical conversion from the red, green and blue color space to the hue, saturation and lightness color space. The conversion process includes: first, normalizing the red channel value, the green channel value and the blue channel value to the range of 0 to 1, and when calculating the hue component, using the inverse cosine function to process the difference between twice the red channel value minus the green channel value minus the blue channel value, divided by the square root of twice the red channel value and the green channel value minus the square root of the product of the red channel value and the blue channel value minus the square root of the product of the green channel value and the blue channel value, and then mapping the calculation result to the range of 0 to 360 degrees.

[0073] Based on the physiological characteristics of the target tomato variety during the fruit expansion stage, the surface reflectance spectrum of the fruit of this variety is measured using a spectrophotometer from flowering and fruit setting to full maturity. The spectral data is converted to hue values ​​and a database of correspondences between hue values ​​and fruit development days is established. The hue value corresponding to the expansion stage is determined to have a lower limit of, for example, 35 degrees and an upper limit of, for example, 85 degrees. During image processing, the converted hue component image is compared pixel by pixel with a preset hue threshold range. If the pixel hue value is between 35 and 85 degrees, the pixel is marked as belonging to the fruit area; otherwise, it is marked as the background area, generating a binary segmentation mask image.

[0074] The connected domains in the binary segmentation mask image are marked, and the connected domain with the largest area is selected as the target fruit area. The boundary tracing algorithm is used to extract the coordinate set of the contour points of the area. The execution process of the boundary tracing algorithm is as follows: scan from the upper left corner of the image until the first foreground pixel is found as the starting point, search for the next contour point clockwise in the eight-neighborhood direction, repeat this process until returning to the starting point to form a closed contour, and record the horizontal and vertical coordinate values ​​of all contour points in the image coordinate system. The minimum circumscribed circle of the obtained contour point set is calculated based on the rotating caliper algorithm: first calculate the convex hull polygon of the contour point set, select the endpoint of the convex hull polygon diameter as the initial baseline, calculate the vertical distance of all contour points to the current baseline and record the maximum point, form a new triangle with the baseline endpoint and the maximum point and update the circumscribed circle, iteratively rotate the baseline direction until the change in the circumscribed circle radius is less than, for example, 0.001 mm or the number of iterations reaches an upper threshold of, for example, 100 times, and finally output the center coordinates and radius value of the minimum circular boundary.

[0075] The minimum circumscribed circle radius is multiplied by 2 to obtain the circular boundary diameter. This diameter is used as the measured fruit diameter of the target tomato plant. The measurement result is stored as a floating-point number and output to the data processing system, with precision retained to one decimal place and in millimeters. The saturation and lightness components of the color conversion process are used to verify image quality. When the saturation component falls below 5%, for example, or the lightness component rises above 95%, for example, the image recapture mechanism is triggered.

[0076] S2. When the fruit diameter reaches the set expansion threshold, a phosphorus pulse start signal is generated, which is specifically implemented as follows:

[0077] A real-time receiving operation is performed to obtain the fruit diameter measurement value of the target tomato plant output from the fruit diameter measurement step. This measurement value is formatted as a floating-point number with precision to one decimal place and is uniformly expressed in millimeters. The set expansion threshold corresponding to the current target tomato plant is retrieved from a pre-generated tomato variety characteristic database. This database is constructed as follows: For each tomato variety, a planting experiment is conducted under standard cultivation conditions. The horizontal and vertical diameters of the fruit are measured using a vernier caliper at a fixed time each day. The fruit volume is calculated using a volume formula and a development curve is plotted. The diameter corresponding to the inflection point where the fruit volume growth rate transitions from linear to exponential growth is used as the set expansion threshold for that variety. For example, the set expansion threshold for the variety Ruixing 818 is set to 28.5 mm, while the set expansion threshold for the variety Jinguan 5 is set to 32.0 mm. The real-time fruit diameter measurement is then numerically compared with the set expansion threshold using a floating-point greater-than-or-equal-to operation and outputting a Boolean logic result.

[0078] When the monitored fruit diameter is greater than or equal to the set expansion threshold for the first time, the timer starts recording its duration. The timer initialization process includes resetting the internal time counter to zero and activating a high-precision clock source, using the 1-millisecond timestamp accuracy provided by the greenhouse environmental monitoring system. During the timer's operation, a real-time comparison of the fruit diameter with the set expansion threshold is performed once per minute. If the fruit diameter is less than the set expansion threshold in any comparison, the timer is immediately reset and the time counter is cleared.

[0079] The dynamic adjustment of the first preset duration is based on the ambient temperature and light intensity parameters. The specific implementation process is as follows: the ambient temperature measurement value is obtained in degrees Celsius through a temperature sensor deployed 30 cm above the canopy of the target tomato plant. At the same time, the light intensity measurement value is obtained in micromoles per square meter per second through a photosynthetic active radiation sensor installed horizontally at the same position. Establish a first preset duration adjustment model: input the ambient temperature value into the temperature response function, and the implementation method of the temperature response function is that when the temperature is in the range of 20 degrees Celsius to 25 degrees Celsius, the output temperature coefficient is 1.0, when the temperature is above 25 degrees Celsius, the temperature coefficient decreases by 0.05 for every 1 degree Celsius increase, and when the temperature is below 20 degrees Celsius, the temperature coefficient increases by 0.03 for every 1 degree Celsius decrease; input the light intensity value into the light response function, and the implementation method of the light response function is that when the light intensity is 500 micromoles per square meter per second, the output light coefficient is 1.0, when the light intensity is lower than this value, the light coefficient is calculated according to the ratio of the actual measured value to 500 micromoles per square meter per second, and when the light intensity is higher than this value, the light coefficient remains unchanged at 1.0. The temperature coefficient and the light coefficient are multiplied to obtain a comprehensive correction factor, and the comprehensive correction factor is multiplied by the reference duration value to obtain the first preset duration after dynamic adjustment, wherein the reference duration value is set according to the maturity of the tomato variety. For example, the reference duration of early-maturing varieties is set to 40 minutes, while the reference duration of late-maturing varieties is set to 60 minutes.

[0080] If the time counter value recorded by the timer is continuously greater than the first preset time length after dynamic adjustment, and during this period all fruit diameter comparison results remain greater than or equal to the set swelling threshold, a phosphorus pulse start signal is generated. This signal is a high-level valid digital signal with a signal level standard of 3.3 volts. The moment the signal is generated is recorded in the year-month-day-hour-minute-second format of the current system time. After the signal is generated, the timer reset operation is executed and the time counting is stopped until the next time the fruit diameter exceeds the set swelling threshold for the first time, and the timing process is retriggered. The acquisition frequency of ambient temperature and light intensity is set to 1 time per minute. When the sensor measurement value exceeds the reasonable range, a data substitution mechanism is adopted. The reasonable range is defined as an ambient temperature of 0 degrees Celsius to 50 degrees Celsius and a light intensity of 0 micromoles per square meter per second to 2000 micromoles per square meter per second. The data substitution mechanism specifically takes the arithmetic average of the last 10 valid measurement values ​​as the current value.

[0081] S3. When a phosphorus pulse start signal is generated, the thickness gradient of the ion diffusion boundary layer on the root surface of the target tomato plant is synchronously measured. When the steepness of the thickness gradient exceeds the empirical threshold of the corresponding soil type, the corresponding steepness is marked as the boundary layer correction coefficient. The specific implementation is as follows:

[0082] The microelectrode array measurement operation is triggered immediately upon receiving the phosphorus pulse start signal. This operation is performed by deploying a microelectrode array within a specified distance range from the root surface of the target tomato plant, where the specified distance range is defined as a vertical distance of 0 mm to 5 mm from the outside of the root surface. The microelectrode array is composed of multiple ion-selective electrodes, which are arranged in a matrix around the root surface. The center spacing between adjacent electrodes is, for example, 0.5 mm, and the electrode surface is covered with a phosphate ion sensitive membrane. After the measurement is started, the phosphate ion concentration value is continuously captured at a fixed sampling interval. The sampling interval is set to no more than 5 seconds, for example, a 3-second sampling period. Each sampling is synchronously recorded with the coordinates of the position of each electrode and the corresponding timestamp. The concentration measurement accuracy is 0.1 mg per liter.

[0083] The thickness change of the ion diffusion boundary layer per unit time is calculated based on the difference in phosphate ion concentration at adjacent time points. The specific calculation process is as follows: the concentration values ​​of two adjacent samples at the same electrode position are selected, and a physical model of concentration gradient and boundary layer thickness is established based on Fick's second diffusion law. This model states that under steady-state diffusion conditions, the boundary layer thickness is inversely proportional to the concentration difference. When calculating the thickness change, the mathematical relationship of the diffusion coefficient multiplied by the concentration difference divided by the concentration gradient is used. The diffusion coefficient value is corrected according to the soil temperature. For example, at 25 degrees Celsius, the diffusion coefficient of phosphate ions in soil solution is 0.89×10 -9 Square meters per second. The unit time is defined as the time difference between adjacent sampling points. For example, when the sampling interval is 3 seconds, the unit time is 3 seconds. The final output thickness change unit is microns.

[0084] The rate of change of thickness over time is defined as the steepness of the thickness gradient. The steepness is calculated using the central difference method: Three consecutive thickness change measurements are taken. The difference is then subtracted from the sum of the two subsequent thickness changes. The difference is then divided by twice the square of the time interval, where the time interval is the sampling interval. For example, if the sampling interval is 3 seconds, the steepness is expressed in microns per square second. If the measurement sequence is less than three times, the calculation condition is met by waiting for subsequent samples.

[0085] Pre-calibrated empirical thresholds corresponding to soil types are retrieved. This threshold calibration method involves preparing potted test samples of different soil types under identical environmental conditions. Irrigation levels are controlled to induce changes in the root surface boundary layer. A high-speed microscope camera directly measures the rate of change in boundary layer thickness. The rate of change corresponding to the critical point of boundary layer instability is recorded as the empirical threshold for that soil type. For example, the critical empirical threshold for sandy loam is 0.15 microns per square second, while that for clay is 0.08 microns per square second. Soil type determination is based on a soil texture analysis report for the target cultivation area.

[0086] The calculated steepness rate is compared in real time with an empirical threshold for the current soil type. When the steepness rate exceeds the threshold, the current steepness rate is marked as a boundary layer correction factor, which is stored and output as a floating-point number. If the marking operation is not triggered after 10 consecutive calculations, the measurement process is automatically terminated and a default boundary layer correction factor of 1.00 is output. The microelectrode array monitors the electrode potential stability in real time during the measurement. If the potential drift of any electrode exceeds 50 millivolts, data from adjacent electrodes are interpolated using an inverse distance weighted algorithm based on spatial distance.

[0087] S4. By simulating the three-dimensional microscopic phase field of the rhizosphere of the target tomato plant, the curvature distribution characteristics of the phosphate ion diffusion path set are calculated, and the phase field topological entropy value representing the disorder of the effective diffusion flux is generated. The specific implementation is as follows:

[0088] X-ray tomography was used to obtain three-dimensional spatial coordinate data of the target tomato plant's root system. This scanning operation was performed using a microfocus X-ray source at a tube voltage of 50 kilovolts and a tube current of 800 microamperes, with a scan slice thickness of 10 microns. The resulting three-dimensional mesh data was reconstructed with a voxel resolution of 10 microns by 10 microns by 10 microns. A geometric model of the rhizosphere region was constructed based on this 3D mesh data. The rhizosphere region was defined as a spatial region extending 2 mm outward from the root surface. The geometric model was spatially discretized using tetrahedral elements with a maximum cell size of no more than 50 microns. Soil porosity distribution data was mapped onto the geometric model to form solid-liquid-gas three-phase boundary conditions. Soil porosity data was derived from simultaneously collected undisturbed rhizosphere soil samples. Pore size distribution curves were obtained using mercury intrusion porosimetry. In the model, regions with porosity greater than 30% were designated as the gas phase, those with porosity between 10% and 30% as the liquid phase, and those with porosity less than 10% as the solid phase. The boundary conditions were assigned using a spatial interpolation algorithm.

[0089] The Cahn-Hilliard phase-field equation was used to simulate the diffusion of phosphate ions in the rhizosphere. The numerical solution of this equation involved setting the initial conditions to a root-surface phosphate concentration equal to the plant's physiological uptake concentration, for example, 0.2 mmol / L, and a bulk soil solution concentration equal to the background phosphorus concentration, for example, 0.05 mmol / L. The mobility parameter in the governing equation was determined using the Stokes-Einstein relation, which states that the mobility equals the diffusion coefficient divided by the Boltzmann constant divided by the absolute temperature. The time step was set to 0.1 seconds, and the total simulation duration was set to 600 seconds. The finite element method was used to iteratively solve the evolution of the phase field variables. The iteration was terminated when the Euclidean norm of the concentration field change between two consecutive steps was less than 10 to the power of -6 mmol / L.

[0090] The instantaneous curvature values ​​of all phosphate ion trajectories in the simulation results were extracted. The specific operation process was as follows: all ion trajectory curves were identified in the simulation output file, a trajectory point was sampled every 0.1 seconds along the trajectory line, and the instantaneous curvature value at the trajectory point was calculated. The instantaneous curvature value was solved using the differential geometry characteristics of the trajectory curve. The calculation method was to take the coordinates of three consecutive trajectory points within a time interval of 0.05 seconds before and after the trajectory point, and calculate the curvature value based on the inverse of the arc radius determined by the three points. The curvature unit was per millimeter. The probability density distribution of the instantaneous curvature values ​​of all sampling points in the preset angle range was calculated. The preset angle range covers the range of 0 radians to 2π radians. The angle range was divided into 36 intervals, each spanning 10 degrees. The frequency of curvature values ​​falling into each angle interval was calculated and divided by the total number of sampling points to obtain the normalized probability density.

[0091] The phase field topological entropy is calculated based on the probability density distribution. The calculation formula states that the phase field topological entropy is equal to the negative of the sum of the product of the probability density of each angular interval and the natural logarithm of the corresponding probability density. The specific calculation process is as follows: Each of the 36 angular intervals is traversed sequentially. When the probability density of an interval is greater than zero, the probability density value of that interval is multiplied by the natural logarithm of the probability density value. The sum of these products for all intervals is then negated to obtain the dimensionless phase field topological entropy. If an interval has a probability density of zero, its contribution to the entropy is zero. The entropy calculation result is stored as a floating-point number with a precision of 4 decimal places.

[0092] S5. Calculate the optimal injection time of phosphate fertilizer according to the phosphorus pulse starting signal, the phase field topological entropy value and the boundary layer correction coefficient, which is specifically implemented as follows:

[0093] The generation time of the phosphorus pulse start signal is obtained as the reference time point. The signal is automatically triggered by the rhizosphere microenvironment monitoring system when it detects that the available phosphorus concentration in the soil is lower than the dynamically set threshold. The dynamically set threshold is adjusted according to the growth stage of the tomato plant. For example, it is set to 15 mg / kg in the seedling stage and 25 mg / kg in the fruiting stage. The signal generation time is recorded in the coordinated universal time format and is accurate to the millisecond level. For example, the recording example of 8:30:45.123 seconds on May 12, 2023 is marked as 2023-05-12T08:30:45.123Z.

[0094] Calculate the quotient of the phase field topological entropy value divided by the boundary layer correction coefficient. The phase field topological entropy value comes from the floating-point value calculated and output in step S4, and the boundary layer correction coefficient comes from the floating-point value marked and output in step S3. Perform a data validity check before performing the division calculation: when the boundary layer correction coefficient is less than or equal to 0.1, it is forcibly reset to 1.0 to avoid division by zero errors. The calculation result retains four significant digits after the decimal point. The calculation process adopts the IEEE 754 floating-point arithmetic standard.

[0095] The time offset is retrieved based on the quotient value by querying a preset conversion table. This table is stored in the conversion relationship data table of the local embedded database and contains two columns: the first is the quotient interval definition column, which stores left-closed and right-open intervals with a step size of 0.1. For example, intervals with values ​​greater than or equal to 1.0 and less than 1.1 are labeled [1.0, 1.1); the second is the time offset column, which stores integer values ​​in seconds. The query operation uses an interval matching algorithm: the calculated quotient is sequentially compared with the lower limit of each interval in the data table. If the quotient value is greater than or equal to the lower limit of the interval and less than the upper limit of the interval, the corresponding time offset is returned. For example, a quotient of 1.25 matches the interval [1.2, 1.3) and returns an offset of 120 seconds.

[0096] The optimal injection time of phosphate fertilizer is obtained by adding the base time point and the time offset. The time addition calculation requires coordination of time zone conversion: first convert the base time point to a local timestamp, then add the time offset in seconds to the timestamp, and convert the calculation result to the ISO 8601 standard time format that can be recognized by the irrigation controller. For example, when the base time point 2023-05-12T08:30:45.123Z and the offset 120 seconds are input, the output optimal time point is 2023-05-12T08:32:45.123Z.

[0097] The preset conversion relationship table is generated based on historical cultivation data training. The historical data collection process includes collecting no less than 200 sets of tomato cultivation case data within a three-year cultivation cycle. Each set of cases includes a complete rhizosphere monitoring data set and harvest yield records. The training process performs the following steps in sequence:

[0098] Effective event extraction operation: Cultivation examples with a yield increase of more than 10% after phosphate fertilizer injection were selected as effective samples. The yield increase calculation formula is (yield after fertilization minus yield before fertilization) divided by the yield before fertilization multiplied by 100%. At the same time, abnormal samples encountered extreme climate were eliminated.

[0099] Feature association analysis operation: Three feature quantities are extracted for each valid sample: quotient feature: the real-time calculation result of the phase field topological entropy value divided by the boundary layer correction coefficient during the fertilization; time offset feature: the difference in seconds between the actual fertilization execution time and the phosphorus pulse start signal generation time; environmental weight factor: including soil type weight coefficient (for example, sandy loam takes 0.8 and clay takes 1.2) and reproductive period weight coefficient (for example, 1.0 for seedling stage and 1.3 for flowering stage).

[0100] Data normalization operation: The time offset characteristics are weighted according to the environmental weight factor. The weighted calculation formula is expressed as: normalized offset = soil type weight coefficient × reproductive period weight coefficient × original offset.

[0101] Relationship table construction operation: discretize the quotient value feature into continuous intervals with a step size of 0.1, calculate the arithmetic mean of the normalized offsets of all samples in the same interval as the final offset of the interval, and generate a conversion relationship table covering the quotient value range of 1.0 to 4.5, containing 36 discrete intervals.

[0102] S6. Outputting a phosphate fertilizer injection instruction to the drip irrigation execution device at the optimal phosphate fertilizer injection time point is specifically implemented as follows:

[0103] The current system time is continuously obtained and compared with the optimal injection time of phosphate fertilizer in real time. The current system time is obtained through the real-time clock interface of the embedded operating system. The time data format is Coordinated Universal Time and is accurate to milliseconds. The comparison operation is executed in a fixed frequency loop of 10 times per second by a dedicated timing thread. Each time the loop is executed, the optimal injection time of phosphate fertilizer calculated and output by step S5 is read from the shared memory area, and the current system time string is accurately matched with the time point string character by character. The matching accuracy requires that the timestamp year, month, day, hour, minute, second and millisecond fields are completely consistent. When it is detected that the lexicographic order of the current system time string is less than the optimal injection time of phosphate fertilizer string, the waiting state is maintained. When the two strings are detected to be exactly the same for the first time, the instruction generation process is triggered.

[0104] When the current system time reaches the optimal injection time for phosphate fertilizer for the first time, a phosphate fertilizer injection instruction containing a timestamp and a fertilizer identifier is generated. The instruction generation process includes data encapsulation and verification calculation: the timestamp field directly copies the ISO 8601 format string of the optimal injection time for phosphate fertilizer, such as 2023-06-18T14:28:30.500Z. The fertilizer identifier field obtains the corresponding code by querying the locally stored fertilizer type mapping table. Potassium dihydrogen phosphate fertilizer corresponds to the fixed character code KH2PO4-01 in this mapping table. A 2-byte checksum field is added to the instruction body. The checksum calculation uses the International Telegraph and Telephone Consultative Committee standard 16-bit polynomial algorithm to perform a cyclic redundancy check on the byte sequence of the timestamp field and the fertilizer identifier field, ultimately generating a fixed-length 64-byte binary instruction message.

[0105] The phosphate fertilizer injection instruction is transmitted to the control unit of the drip irrigation execution device through the wired communication interface. The wired communication interface adopts an industrial serial bus with differential signal transmission. The communication process follows the master-slave serial communication specification. The transmission operation is performed in the following order: the 64-byte instruction message is added with the device address byte 0x01 and the control function byte 0x10 to form a transmission frame, and the data stream is sent at a rate of 19200 bits per second through the serial communication port. After the sending is completed, the timeout timer is started, for example, 500 milliseconds are set to wait for a response. When the confirmation frame containing the same device address and function code returned by the control unit is received and the frame check passes, the transmission is marked as successful. If no valid response is received within the timeout, the retransmission mechanism is started. The upper limit of the retransmission number is set to 3 times, and the adjacent retransmission intervals are set in an increasing sequence of 200 milliseconds, 400 milliseconds, and 800 milliseconds.

[0106] The phosphate fertilizer injection instruction includes the execution time point and the potassium dihydrogen phosphate fertilizer type code. The control unit activates the fertilizer injection pump in the drip irrigation pipeline according to the instruction. The control unit response process includes instruction parsing and execution control: after receiving the complete data frame, the check code is first verified. After the verification is passed, the execution time point field is extracted and a secondary time comparison is performed with the local clock of the control unit. The time deviation is allowed to be within the range of ±5 minutes. The fertilizer type code KH2PO4-01 is extracted to query the fertilizer parameter database built into the control unit to obtain the potassium dihydrogen phosphate injection control parameters. The parameters include the target molar concentration of 0.1 mol per liter and the injection flow rate per unit time of 5 ml per second. The speed of the fertilizer injection pump motor is controlled by the pulse width modulation circuit according to the parameter value. At the same time, the electromagnetic flowmeter connected to the pipeline monitors the accumulated flow in real time. When the accumulated flow value reaches the preset threshold, such as 200 ml, the power supply of the motor drive circuit is cut off and the execution completion status code is sent to the host computer.

[0107] In the field of greenhouse tomato water and fertilizer management, existing technologies generally adopt a static fertilizer supply model based on the growth period. In this embodiment, steps S1 to S6 establish a collaborative decision-making mechanism: First, the fruit diameter biosignal (S1) is cross-scale coupled with the physical response of the root-soil interface (S3) and the microscopic diffusion dynamics (S4). This decision-making logic, which dynamically links macroscopic morphological indicators, mesoscopic boundary layer gradients, and microscopic ion transport characteristics, breaks away from the traditional mindset of analyzing the soil-crop system in isolation. Second, the steepness rate marking mechanism (S3) of the boundary layer correction coefficient does not simply use a fixed threshold. Instead, a dynamic empirical threshold system is established based on different soil types. This transforms the differentiated retardation effect of soil texture on ion migration into a quantifiable parameter, overcoming the inherent limitations of homogenized soil models. Furthermore, the phase field topological entropy (S4) is used to quantify the disorder of the curvature distribution of the phosphate diffusion path, transforming the difficult-to-observe microscopic diffusion process into an actionable decision factor, forming a multidimensional constraint with the boundary layer parameters. Finally, the conversion relationship table (S5) trained based on historical data maps complex biophysical parameters into time offsets, so that the timing of phosphate fertilizer injection can simultaneously meet the dual optimization of root absorption activity window and soil migration efficiency, forming a closed-loop decision-making chain, significantly improving phosphate utilization efficiency and overcoming the supply lag problem.

[0108] Example 2: Figure 2 The present invention provides a structural diagram of a high-efficiency irrigation and fertilization decision support system for greenhouse tomatoes, which includes the following modules:

[0109] a diameter detection module for obtaining the fruit diameter of a target tomato plant in the fruit expansion stage;

[0110] A signal trigger module is used to generate a phosphorus pulse start signal when the fruit diameter reaches a set swelling threshold;

[0111] The gradient monitoring module is used to synchronously measure the thickness gradient of the ion diffusion boundary layer on the root surface of the target tomato plant when the phosphorus pulse start signal is generated. When the steepness of the thickness gradient exceeds the empirical threshold of the corresponding soil type, the corresponding steepness is marked as the boundary layer correction coefficient;

[0112] The entropy analysis module is used to simulate the three-dimensional microscopic phase field of the rhizosphere of the target tomato plant, calculate the curvature distribution characteristics of the phosphate ion diffusion path set, and generate the phase field topological entropy value that represents the disorder of the effective diffusion flux;

[0113] The decision-making operation module is used to calculate the optimal injection time of phosphate fertilizer based on the phosphate pulse start signal, the phase field topological entropy value and the boundary layer correction coefficient;

[0114] The execution output module is used to output the phosphate fertilizer injection instruction to the drip irrigation execution device at the optimal phosphate fertilizer injection time point.

[0115] The calculations involved in the embodiments are all dimensionless numerical calculations, and the preset parameters and thresholds in the calculations are set by those skilled in the art according to actual conditions.

[0116] The above embodiments may be implemented in whole or in part through software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments may be implemented in whole or in part in the form of a computer program product.

[0117] Those skilled in the art will appreciate that the modules and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application of the technical solution and the invention constraints. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0118] In addition, each functional module in each embodiment of the present application may be integrated into one processing module, or each module may exist physically separately, or two or more modules may be integrated into one module.

[0119] In the several embodiments provided in this application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are merely schematic. For example, the division of the modules is only a logical function division. In actual implementation, there may be other division methods, such as multiple modules or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or modules, which can be electrical, mechanical or other forms.

[0120] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

[0121] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A decision support method for efficient irrigation and fertilization of greenhouse tomatoes, characterized in that: The steps include: S1. Obtaining the fruit diameter of a target tomato plant in the fruit expansion stage; S2. When the fruit diameter reaches the set expansion threshold, a phosphorus pulse start signal is generated; S3. When a phosphorus pulse start signal is generated, the thickness gradient of the ion diffusion boundary layer on the root surface of the target tomato plant is synchronously measured. When the steepness of the thickness gradient exceeds the empirical threshold of the corresponding soil type, the corresponding steepness is marked as the boundary layer correction coefficient. S4. By simulating the three-dimensional microscopic phase field of the rhizosphere of the target tomato plant, the curvature distribution characteristics of the phosphate ion diffusion path set are calculated, and the phase field topological entropy value representing the disorder of the effective diffusion flux is generated; S5. Calculating the optimal injection time of phosphate fertilizer according to the phosphate pulse starting signal, the phase field topological entropy value, and the boundary layer correction coefficient; S6. Outputting a phosphate fertilizer injection instruction to the drip irrigation execution device at the optimal phosphate fertilizer injection time point.

2. A decision support method for efficient irrigation and fertilization of greenhouse tomatoes according to claim 1, characterized in that: Obtain the fruit diameter of the target tomato plant at the fruit expansion stage, including: Capturing fruit images of target tomato plants through a camera device fixed to a facility support; Color space conversion and threshold segmentation are used to separate the fruit area from the fruit image; Extract the minimum circumscribed circle boundary of the fruit area; The diameter of the smallest circumscribed circle was taken as the measurement of the fruit diameter.

3. The method for supporting efficient irrigation and fertilization decision of greenhouse tomatoes according to claim 2, characterized in that: The color space conversion adopts the RGB to HSV conversion mode, the threshold segmentation is set based on the hue value range corresponding to the fruit maturity, and the minimum circumscribed circle boundary is generated by iteratively fitting the contour point set.

4. The method for supporting efficient irrigation and fertilization decision-making for greenhouse tomatoes according to claim 2, characterized in that: When the fruit diameter reaches the set expansion threshold, a phosphorus pulse start signal is generated, including: Compare the fruit diameter obtained in real time with the set swelling threshold; When the fruit diameter is greater than or equal to the set swelling threshold for the first time, the timer is triggered to start recording the duration; If the fruit diameter continues to be greater than or equal to the set swelling threshold value for more than a first preset time period, a phosphorus pulse start signal is generated.

5. The method for supporting efficient irrigation and fertilization decision-making for greenhouse tomatoes according to claim 4, characterized in that: The expansion threshold is pre-stored in the database according to the characteristics of the tomato variety, and the first preset duration is dynamically adjusted according to the ambient temperature and light intensity of the greenhouse where the target tomato plant is located. The ambient temperature and light intensity are obtained in real time through sensors deployed in the greenhouse.

6. The method for supporting efficient irrigation and fertilization decision-making for greenhouse tomatoes according to claim 4, characterized in that: When a phosphorus pulse start signal is generated, the thickness gradient of the ion diffusion boundary layer on the root surface of the target tomato plant is measured synchronously. When the steepness of the thickness gradient exceeds the empirical threshold of the corresponding soil type, the corresponding steepness is marked as the boundary layer correction coefficient, including: The microelectrode array measurement is started at the moment when the phosphorus pulse start signal is generated; The phosphate ion concentration value is continuously captured by a microelectrode array arranged within a specified distance range on the root surface of the target tomato plant; The thickness change of the ion diffusion boundary layer per unit time is calculated based on the concentration difference between adjacent time points; The rate of change of thickness variation with time is defined as the steepness rate of thickness gradient; Retrieve the pre-calibrated empirical threshold corresponding to the soil type; When the steep change rate is greater than the empirical threshold of the current soil type, the current steep change rate is marked as the boundary layer correction coefficient.

7. The method for supporting efficient irrigation and fertilization decision-making for greenhouse tomatoes according to claim 6, characterized in that: By simulating the three-dimensional microscopic phase field of the rhizosphere of the target tomato plant, the curvature distribution characteristics of the phosphate ion diffusion path set are calculated, and the phase field topological entropy value that characterizes the disorder of the effective diffusion flux is generated, including: The rhizosphere geometric model was constructed based on the three-dimensional spatial coordinate data of the target tomato plant root system reconstructed by X-ray tomography; Mapping soil porosity distribution data to the geometric model to form solid-liquid-gas three-phase boundary conditions; The Cahn-Hilliard phase field equation was used to simulate the diffusion process of phosphate ions in the rhizosphere. Extract the instantaneous curvature values ​​of all phosphate ion motion trajectories in the simulation results; Calculate the probability density distribution of the instantaneous curvature value in the preset angle range; Calculate the phase field topological entropy value based on the probability density distribution: The phase field topological entropy is equal to the negative value of the sum of the product of the probability density of each angle interval and the natural logarithm of the corresponding probability density; The preset angle interval covers the range from zero radians to 2π radians, and the instantaneous curvature value is solved through the differential geometric characteristics of the trajectory curve.

8. The method for supporting efficient irrigation and fertilization decision-making for greenhouse tomatoes according to claim 7, characterized in that: The optimal injection time of phosphate fertilizer is calculated based on the phosphate pulse start signal, phase field topological entropy value and boundary layer correction coefficient, including: Obtaining the generation time of the phosphorus pulse start signal as the reference time point; Calculate the quotient obtained by dividing the phase field topological entropy by the boundary layer correction coefficient; The preset conversion relationship table is queried according to the quotient value to obtain the time offset; The optimal injection time of phosphate fertilizer is obtained by adding the reference time point and the time offset; The preset conversion relationship table is generated based on historical cultivation data training and stored in the local database.

9. The method for supporting efficient irrigation and fertilization decision-making for greenhouse tomatoes according to claim 8, characterized in that: Outputting a phosphate fertilizer injection instruction to a drip irrigation execution device at an optimal phosphate fertilizer injection time point includes: Continuously obtain the current system time and compare it with the optimal injection time of phosphate fertilizer in real time; When the current system time reaches the optimal injection time point of phosphate fertilizer for the first time, a phosphate fertilizer injection instruction including a timestamp and a fertilizer identifier is generated; transmitting the phosphate fertilizer injection instruction to the control unit of the drip irrigation execution device through the wired communication interface; The phosphate fertilizer injection instruction includes the execution time point and the potassium dihydrogen phosphate fertilizer type code. The control unit activates the fertilizer injection pump in the drip irrigation pipeline according to the instruction.

10. A high-efficiency irrigation and fertilization decision support system for greenhouse tomatoes, used to implement the high-efficiency irrigation and fertilization decision support method for greenhouse tomatoes according to any one of claims 1 to 9, characterized in that: Includes the following modules: a diameter detection module for obtaining the fruit diameter of a target tomato plant in the fruit expansion stage; A signal trigger module is used to generate a phosphorus pulse start signal when the fruit diameter reaches a set swelling threshold; The gradient monitoring module is used to synchronously measure the thickness gradient of the ion diffusion boundary layer on the root surface of the target tomato plant when the phosphorus pulse start signal is generated. When the steepness of the thickness gradient exceeds the empirical threshold of the corresponding soil type, the corresponding steepness is marked as the boundary layer correction coefficient; The entropy analysis module is used to simulate the three-dimensional microscopic phase field of the rhizosphere of the target tomato plant, calculate the curvature distribution characteristics of the phosphate ion diffusion path set, and generate the phase field topological entropy value that represents the disorder of the effective diffusion flux; The decision-making operation module is used to calculate the optimal injection time of phosphate fertilizer based on the phosphate pulse start signal, the phase field topological entropy value and the boundary layer correction coefficient; The execution output module is used to output the phosphate fertilizer injection instruction to the drip irrigation execution device at the optimal phosphate fertilizer injection time point.

Images