A fertilization decision system fusing soil and rice growth information data acquisition and analysis

Abstract

The present application relates to the technical field of agricultural intelligent prediction, and discloses a fertilization decision system fusing soil and rice growth information data collection and analysis, comprising: a microcontroller applies a preset length of irrigation pulse by controlling an irrigation valve, a soil moisture sensor is used to monitor the time interval from the peak value to the baseline of water, the interval is used to inverse the crop transpiration rate and predict the nutrient requirement, the irrigation behavior is reconstructed as an active diagnostic tool, the traditional chemical index detection is replaced by physical process dynamic monitoring, the mode is changed from lagging adjustment to pre-prediction, and the precise decision system resistant to environmental interference is constructed on the basis of low-cost hardware by combining acoustic verification and differential correction mechanism.

Description

Technical Field

[0001] This invention relates to a fertilization decision-making system that integrates the collection and analysis of soil and rice growth information data, belonging to the field of agricultural intelligent prediction technology. Background Technology

[0002] Current fertilization decision-making technologies mainly rely on two types of methods: one is static detection based on soil nutrients, such as NPK sensors, which directly measure soil inventory through chemical analysis; the other is remote sensing monitoring based on crop phenotypes, such as leaf color and chlorophyll index, which indirectly infer nutrient status through spectral characteristics. Although these technologies are widely used, they have inherent limitations.

[0003] During the typical rice jointing and booting stage, existing systems only trigger fertilization when soil nutrients are depleted to a threshold or when crops show visible nutrient deficiency symptoms. This creates a vicious cycle of delayed detection and delayed remediation. Especially during the midday heat, crops enter a dormant state due to the closure of physiological stomata, resulting in a sharp reduction in transpiration. If fertilization is applied based on morning monitoring data at this time, a large amount of water and fertilizer will be lost due to the misalignment of the absorption window. A deeper contradiction lies in the fact that existing technologies only capture static chemical indicators or apparent biological characteristics, but cannot quantify the dynamic changes in the transpiration rate, the core engine that determines nutrient absorption efficiency. The industry has tried to optimize by fusing multi-source data or increasing sensor density, but because it does not address the fundamental issue of real-time monitoring of physiological processes, it not only exacerbates the complexity of the system but also makes it difficult to avoid derivative problems such as soil evaporation interference and sensor drift.

[0004] Specifically, existing technologies suffer from three main bottlenecks: 1. They rely on historical state detection, making it impossible to predict changes in crop metabolic rhythms in the next few hours; 2. High-temperature evaporation and soil heterogeneity lead to measurement distortion, and compensation algorithms are costly; 3. They sever the causal link between transpiration-driven nutrient transport mechanisms and fertilization decisions. Therefore, how to construct a fertilization prediction system based on real-time inversion of crop physiological processes and resistant to environmental disturbances, thereby transitioning from remedial regulation to proactive supply, is the technical problem this invention aims to solve. Summary of the Invention

[0005] This invention provides a fertilization decision-making system that integrates soil and rice growth information data collection and analysis. Its main purpose is to solve the problems of misaligned fertilization timing and low resource utilization caused by delayed monitoring and noise interference in existing technologies.

[0006] To achieve the above objectives, the present invention provides a fertilization decision-making system that integrates soil and rice growth information data collection and analysis, comprising:

[0007] The irrigation valve is configured to apply an irrigation pulse to the rice root zone for a preset duration.

[0008] A soil moisture sensor is configured to monitor the time interval required for the soil moisture in the rice root zone to fall from a peak level back to the baseline level before the irrigation pulse was applied after the irrigation pulse has ended.

[0009] The microcontroller, connected to the irrigation valve and soil moisture sensor, is configured to: determine the transpiration rate of rice based on the time interval monitored by the soil moisture sensor, wherein the transpiration rate is inversely proportional to the time interval; determine the nutrient requirement level of rice based on the transpiration rate and a preset lookup table method; and generate and output a fertilizer application level to guide the determination of the fertilizer application level based on the nutrient requirement level.

[0010] Preferably, the microcontroller determines the transpiration rate of rice based on the following formula: ,in, Indicates the transpiration rate. This represents a preset constant. This indicates the time interval required for soil moisture to fall from its peak level back to the baseline level, and the time interval is measured in minutes.

[0011] Preferably, the generation of fertilizer application levels is also adjusted based on information about the growth stage of rice.

[0012] Preferably, the system also includes a photoresistor, which is configured to be placed on the ground below the rice canopy to sense the light transmittance through the rice canopy; the microcontroller is configured to: record the initial transmittance baseline value at the beginning of rice transplanting; continuously track the decreasing trend of transmittance, and make a judgment based on the ratio of transmittance to the initial transmittance baseline value: when the ratio is less than or equal to 70%, it is identified as the jointing and booting stage; when the ratio is less than or equal to 30%, it is identified as the heading and grain-filling stage.

[0013] Preferably, the system further includes an acoustic sensor configured to be in close contact with the irrigation valve or a nearby water pipe; the microcontroller is configured to: store the acoustic signal characteristics generated when the irrigation valve is normally open as a reference fingerprint; listen to the real-time acoustic signal received by the acoustic sensor each time an irrigation pulse is applied, obtain the characteristics of the real-time acoustic signal; and compare the characteristics of the real-time acoustic signal with the reference fingerprint, and if they do not match, output an indication that the irrigation pulse is invalid.

[0014] Preferably, the system further includes a second soil moisture sensor, which is the same as the soil moisture sensor and is configured to be deployed in a bare soil area where no rice roots are distributed; the microcontroller is configured to: simultaneously monitor the first time interval measured by the soil moisture sensor and the second time interval measured by the second soil moisture sensor when an irrigation pulse is applied; and correct for soil surface evaporation interference based on the difference between the first time interval and the second time interval to obtain the net evaporation response time, which is used as the basis for judging the evaporation rate.

[0015] Preferably, the soil moisture sensor integrates a thermistor, and the microcontroller is further configured to: apply a constant energy microcurrent pulse to the electrodes of the soil moisture sensor before the monitoring time interval, so as to generate Joule heat in the soil around the probe; monitor the instantaneous temperature rise of the probe through the thermistor; and determine a correction coefficient for correcting subsequent soil moisture monitoring data by querying a preset correction factor lookup table based on the instantaneous temperature rise.

[0016] Preferably, the microcontroller is further configured to: initiate a root function judgment process when the time interval is greater than a preset first diagnostic threshold; the root function judgment process includes: controlling the irrigation valve to apply micro-spraying to the rice canopy for a preset duration; monitoring the rate of change of the surface temperature of the rice canopy leaves after micro-spraying through an infrared temperature sensor; and determining the root health status of the rice based on the comparison result of the temperature change rate and the preset threshold, and adjusting the fertilizer application level according to the root health status.

[0017] Preferably, the system further includes an image sensor configured to acquire an image of the rice canopy; the microcontroller is further configured to: identify the shadow area projected by the rice plants in the image; extract the principal direction angle of the shadow area; calculate the angular standard deviation of the principal direction angle to determine the degree of tilt dispersion of the rice plants; and determine the lodging risk level of the rice based on the comparison result of the angular standard deviation and a preset lodging threshold.

[0018] Preferably, when determining the nutrient requirement level of rice, the microcontroller compares the time interval with a preset time threshold: when the time interval is less than the first time threshold, it is determined to be a high nutrient requirement level; when the time interval is between the first time threshold and the second time threshold, it is determined to be a medium nutrient requirement level; when the time interval is greater than the second time threshold, it is determined to be a low nutrient requirement level.

[0019] Compared with the prior art, the beneficial effects of the present invention are:

[0020] 1. By reconstructing the function of irrigation valves into an active diagnostic tool, and combining it with soil moisture sensors to monitor pulse response time, the system transforms static environmental parameter measurements into dynamic inversion of crop transpiration rates. This time-series analysis based on physical processes allows nutrient demand prediction to no longer rely on lagging chemical index detection, but instead directly captures the rhythmic changes in crop life activities. The coupling mechanism between irrigation behavior and water absorption response addresses the problem of hidden starvation windows in traditional solutions, providing a basis for fertilization decisions. Furthermore, the dual-sensor differential architecture achieves physical isolation of soil evaporation noise at the hardware level by comparing the difference in water drop between the root zone and the bare soil zone. When this mechanism is combined with probe microcurrent self-calibration, the system can still maintain core indicators under extreme conditions. The stability of this system, achieved through a combination of spatial difference and in-situ calibration, frees the predictive model from dependence on absolute sensor accuracy, enabling the construction of a system with high adaptability to field environments based on low-cost components.

[0021] 2. The acoustic fingerprint verification mechanism transforms the invisible valve execution status into a traceable physical signal, ensuring the authenticity and effectiveness of each irrigation pulse trigger. When this verification mechanism is linked with canopy micro-sprinkler diagnosis, a secondary detection is triggered if root function abnormalities occur. The system forms a two-level verification of hardware execution and physiological response: acoustic arbitration ensures the reliability of basic data acquisition, while transpiration lag response analysis further identifies the root cause of physiological abnormalities. This cross-verification between the execution end and the biological end helps reduce decision-making errors caused by actuator failures. Meanwhile, the photoresistor placed under the canopy continuously tracks the light transmittance decay curve, transforming the growth stage identification solidified by human experience into fully automatic physiological rhythm analysis. When this mechanism is combined with shadow direction dispersion analysis, the system synchronously captures changes in the mechanical state of the plant population. The synergistic utilization of light signals and image byproducts achieves an upgrade in agronomic decision-making from growth stage judgment to lodging risk assessment under zero new hardware conditions, avoiding the error chain of manual input.

[0022] 3. By establishing a transpiration response delay By linking the rate of change in canopy temperature across dimensions, the system automatically initiates root function diagnosis when it detects slow water absorption. This two-tiered judgment mechanism of physiological manifestation and pathological root cause helps to quickly distinguish between metabolic inhibition and root lesions. When combined with the analysis of shade direction consistency, the trend of plant mechanical deterioration is captured several days in advance, forming a comprehensive early warning network covering physiological, pathological and physical risks. Attached Figure Description

[0023] Figure 1 This is a framework diagram of the agricultural environment monitoring and fertilization decision-making system of the present invention;

[0024] Figure 2 This is a flowchart of the fertilization decision-making process of the present invention;

[0025] Figure 3 This is a graph showing the relationship between ambient temperature and transpiration response time in this invention.

[0026] Figure 4 This is a timing diagram of the fertilization decision system of the present invention.

[0027] The objectives, features, and advantages of this invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0028] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be described in detail below. Obviously, the described embodiments are only some embodiments of this invention, and not all embodiments. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0029] This invention discloses a fertilization decision-making system that integrates soil and rice growth information data acquisition and analysis. Its overall architecture revolves around a microcontroller as the core of control and analysis. This microcontroller is networked with a set of sensors and actuators deployed in the field, forming a closed-loop physiological information sensing and precise intervention system. This system mainly includes a core diagnostic unit for dynamically inverting crop transpiration rates, a signal purification unit for adaptively correcting environmental interference, and a comprehensive judgment unit for multi-dimensionally assessing crop growth status. The collaborative operation of these units aims to transform fertilization decisions from remedial adjustments based on lagging chemical indicators to supply based on real-time physiological process predictions. To address the challenge of the mismatch between nutrient supply and the actual absorption window of crops in traditional fertilization strategies, this solution reconstructs irrigation behavior as an active diagnostic tool. Its procedure begins with the microcontroller sending a command to irrigation valves deployed in the rice root zone, applying an irrigation pulse with a preset duration. The duration of this pulse is set according to soil type and field management experience, aiming to provide the root zone soil with a... A sufficient and standardized moisture input can be clearly captured by the sensor. Simultaneously with the issuance of the irrigation pulse command, an acoustic sensor installed close to the irrigation valve or nearby water pipes is activated to monitor the acoustic signal characteristics generated when the valve opens and closes. The microcontroller compares the real-time acquired acoustic signal with a pre-stored reference fingerprint. This reference fingerprint is established during initial system calibration by recording the vibration and water flow acoustic waveforms generated by the valve under normal water pressure and fully open conditions. If the real-time signal characteristics do not match the reference fingerprint, for example, due to valve jamming or pipe blockage causing abnormal energy in the key frequency bands of the sound spectrum, the system determines that the irrigation pulse is invalid and outputs an indication message, thus avoiding the risk of making subsequent decisions based on invalid physical inputs. After an irrigation pulse confirmed as valid by the acoustic indicator ends, a soil moisture sensor located in the rice root zone immediately begins high-frequency monitoring of soil moisture changes. Its task is to accurately record the time interval taken for the soil moisture to naturally fall from the peak level caused by the pulse to the baseline level before irrigation. This time interval directly reflects the overall rate of water consumption in the area, forming the basis for subsequent analysis.

[0030] Given that moisture evaporation from exposed soil surfaces is a factor affecting To address the main source of interference in measurement accuracy, the system incorporates a differential correction mechanism to physically eliminate this environmental noise. Specifically, within the planting area, a second soil moisture sensor of identical model and specifications is deployed in parallel with the first soil moisture sensor monitoring the root zone. However, its location is deliberately chosen to be adjacent to a bare soil area where no rice roots are distributed. When an irrigation pulse is applied, the microcontroller synchronously records the total time elapsed as measured by the first sensor. and the pure evaporation time measured by the second sensor Subsequently, the microcontroller performs interpolation. A net evaporation response time, after eliminating background evaporation interference, was obtained; this was the corrected... This is used as the ultimate basis for determining the transpiration rate of rice. Furthermore, to address potential sensor measurement drift caused by soil temperature changes, the soil moisture sensor also integrates a thermistor. Before initiating each irrigation pulse and moisture monitoring cycle, the microcontroller applies a constant-energy micro-current pulse to the metal electrode of the soil moisture sensor. This pulse is designed to generate a measurable temperature rise without damaging the crop roots. The integrated thermistor then monitors the instantaneous temperature rise of the probe caused by this micro-Joule heating effect. Based on this temperature rise, the microcontroller determines a correction coefficient for this measurement cycle by consulting a preset correction factor lookup table and applies this coefficient to all subsequent readings of raw moisture data. Through this real-time self-calibration process, the system actively compensates for measurement deviations caused by soil temperature changes at the hardware level. This dual purification mechanism of spatial differential and in-situ calibration jointly ensures the accuracy of core indicators. Stability and reliability in complex field environments, achieving purified net transpiration response time. Then, the microcontroller makes hierarchical decisions based on this core parameter. First, it uses an inverse proportional function relationship. To invert the transpiration rate of rice ,in, This is a preset constant derived from field trials calibrated based on soil physical properties and rice variety transpiration models. The transpiration rate is directly related to water and fertilizer absorption efficiency; therefore, the microcontroller will further... The rice's current nutrient requirements are compared with a set of preset time thresholds to determine its current nutrient requirements, for example, when... When the water demand is below the first time threshold, it indicates extremely rapid water consumption and vigorous transpiration, and is thus classified as a high nutrient requirement level; when... When the nutrient requirement level falls between the first and second time thresholds, it is determined to be at the medium nutrient requirement level; while when... When the transpiration rate exceeds the second time threshold, it indicates that transpiration is weak and is classified as a low nutrient requirement level. This threshold is derived from a large amount of experimental data at a specific growth stage and is a critical value that can effectively distinguish different physiological states. This nutrient requirement level is the basis for generating the final fertilization recommendations.

[0031] A precise fertilization plan not only needs to respond to current physiological needs but also must be adapted to the crop's growth stage. To this end, the system integrates an automatic growth stage identification mechanism based on light signals. Specifically, a photoresistor is placed on the ground below the rice canopy to continuously sense the light transmittance through the leaf canopy. In the early stages of rice transplanting, the system automatically records and stores an initial baseline value of light transmittance before the canopy has formed. As the rice tillers and grows, the canopy gradually closes, and the light intensity received by the photoresistor tends to decrease. The microcontroller continuously calculates the ratio of the current light transmittance to the initial baseline value to determine the transition point in the growth stage. When this ratio first drops to less than or equal to 100%, the system detects the transition point. When the ratio reaches 70%, the system automatically identifies the crop growth stage as the jointing and booting stage, which is sensitive to nitrogen fertilizer requirements. When this ratio further decreases to less than or equal to 30%, it is identified as the heading and grain-filling stage, which has higher requirements for phosphorus and potassium fertilizers. The final fertilizer application level output by the microcontroller is the result of weighted adjustment based on the aforementioned nutrient requirement levels and the current growth stage information. For example, under the same high nutrient requirement level, the system will recommend a higher nitrogen fertilizer application ratio for crops in the jointing and booting stage, thereby achieving intelligent integration of agronomic knowledge and real-time physiological data. In addition, the system also has the ability to proactively warn of potential growth risks. When the microcontroller detects... An abnormally prolonged duration, exceeding a preset first diagnostic threshold, is defined as the longest duration of healthy plants under low transpiration requirements, such as at night or on cloudy days. Records may indicate two distinct situations: one, the crop enters physiological dormancy due to environmental stress; two, the crop's root system suffers from disease or functional decline. For accurate identification, the system automatically initiates a root function assessment process. In this process, the microcontroller instructs the irrigation system to apply a pre-set duration of micro-spraying to the rice canopy. Simultaneously, infrared temperature sensors deployed above the canopy begin monitoring the rate of temperature change on the leaf surface. Healthy plants exhibit rapid stomata response to the micro-spraying, significantly cooling through water evaporation; their rate of temperature change will exceed a pre-set health threshold. Conversely, if root function is impaired, the leaf stomatal regulation ability decreases, and the rate of temperature change will be very slow. Based on this... By comparing the rate of temperature change with a threshold, the system can determine the root health status of rice and adjust the final fertilization level accordingly. For example, if abnormal root function is detected, the system will suppress or suspend fertilization recommendations and issue an early warning. Simultaneously, the system can use an image sensor to acquire an overhead image of the rice canopy, identify the shaded areas of the plants and extract their principal direction angles through image processing algorithms, and then calculate the angular standard deviation of these angles. This angular standard deviation quantifies the tilt dispersion of the rice plant population. When this value exceeds a preset lodging threshold, the system will output the corresponding lodging risk level to guide subsequent nitrogen fertilizer application, so as to avoid exacerbating the lodging risk due to excessive vegetative growth.

[0032] Example 1: In a specific application scenario, the technical solution of the present invention operates as follows: The scenario involves rice entering the critical jointing and booting stage, which demands nitrogen, and experiencing several consecutive days of high midday temperatures and strong sunlight. Under these conditions, systems relying on soil chemical sensors or fixed irrigation schedules generally face a technical dilemma: the soil exhibits signs of water and fertilizer deficiency due to high-temperature evaporation, while the crop itself, by activating its physiological stomatal closure mechanism to cope with high-temperature stress, has its transpiration and nutrient absorption capacity actually reduced to their lowest level during midday. In this scenario, the system initiates a routine diagnostic and decision-making cycle at 2 PM. The microcontroller first issues an instruction to execute an irrigation pulse verified throughout by acoustic sensors. Then, by processing the differential data from the root zone soil moisture sensor and the second soil moisture sensor in the bare soil area, the system calculates a net transpiration response time after eliminating the impact of high-temperature evaporation. Its value was significantly higher than that recorded in the early morning, from which the transpiration rate was deduced. The result, close to zero, directly quantifies the fact that the crop has entered a state of physiological dormancy. At the same time, the transmittance measured by the photoresistor deployed under the canopy identifies the current crop growth stage as the jointing and heading stage. The agronomic knowledge base at this stage pre-sets a recommendation for the application of high nitrogen fertilizer. At this moment, a decision conflict based on different information dimensions is formed within the system. The macro-level agronomic timing judgment requires the application of nitrogen fertilizer to promote growth, while the micro-level real-time physiological monitoring indicates that the crop's absorption channels have been basically closed.

[0033] The built-in procedures of this scheme are activated under such conflicts. Instead of performing a weighted average of multi-source data, they establish a priority-based decision arbitration mechanism, specifically, the real-time inverted evaporation rate. This is set as a prerequisite for determining whether the nutrient absorption window is open; it only applies when... When the growth stage exceeds a preset basal metabolic threshold, the fertilization recommendation corresponding to the growth stage information determined by the photoresistor will be executed. Therefore, although the growth stage identification unit sends a signal indicating a high nitrogen demand, due to constraints from core physiological indicators, the microcontroller does not output a fertilization command at that time, but instead suspends the decision and enters a higher-frequency monitoring mode. Furthermore, in this cycle, when the acoustic sensor verifies the irrigation pulse, although the real-time acoustic signal characteristics recorded by it generally match the reference fingerprint, there is a slight deviation in the energy ratio in a specific frequency band. This information is recorded by the system as a potential hardware status anomaly log. When it is compared with the subsequently measured ultra-long When combined, the system can distinguish between slow water absorption caused by different reasons. For example, the primary cause of slow water absorption might be crop physiological inhibition, but it could be accompanied by secondary factors such as insufficient irrigation system pressure. This diagnostic mechanism, formed by acoustic hardware verification and physiological response monitoring, enables the differentiation of complex, multi-factor coupled field conditions. By 4 PM that day, as light intensity and temperature decreased, the system calculated [the following data] in a new monitoring cycle. The value shortened significantly. The return to normal levels indicates that the crop's physiological absorption window has reopened. The microcontroller then activates the suspended decision-making process, combining the information on the jointing and booting stages previously determined by the photoresistor, to generate and output a fertilization level that matches the needs of this growth stage and execute the fertilization operation. The entire process resolves the contradiction of the time mismatch between supply and absorption in traditional fertilization management by shifting the focus of decision-making from what nutrients the soil is currently lacking to whether the crop can efficiently absorb nutrients at this moment. Ultimately, without human intervention, this decision-making process autonomously avoids water and fertilizer loss caused by high temperature stress and delivers nitrogen nutrients within the crop's effective absorption window.

[0034] Example 2: To objectively verify the technical effectiveness of the present invention in quantifying crop physiological states and responding to environmental disturbances, this comparative experiment was established. The purpose of the experiment was to conduct a parallel comparison between the system of the present invention and a traditional decision-making system based on a high-precision soil chemical sensor under a controlled environment simulating high temperature and strong evaporation, and to verify the engineering feasibility of the present invention in terms of fertilization decision accuracy in the form of data. For this purpose, the experimental platform was constructed in a glass greenhouse with a controllable environment. Two groups of rice plants of the same species and with uniform growth were cultivated in pots. The control group adopted a traditional precision agriculture scheme, whose decision-making was based on a commercial high-precision NPK sensor and soil tensiometer deployed in the root system. Fertilization was triggered when the soil moisture content or NPK reading was lower than a preset threshold. The experimental group deployed the complete system of the present invention, including a soil moisture sensor in the root zone, a second soil moisture sensor in the bare soil zone for differential correction, and corresponding microcontrollers and irrigation valves.

[0035] In the experiment, the setting of key parameters within the system of this invention, taking the duration of the irrigation pulse as an example, aims to balance the relationship between signal strength and water resource utilization. A pulse that is too short may result in slight changes in soil moisture, making it difficult for the sensor to capture a clear attenuation curve, thus affecting... The signal-to-noise ratio was calculated, but excessively long pulses could cause unnecessary water leakage; therefore, the procedure was as follows: an offline calibration was used to determine the shortest pulse duration that would allow the soil moisture in the root zone to reach 150% to 200% of its baseline level after the pulse ended. Under the loam soil conditions used in this experiment, this duration was determined to be 90 seconds. Regarding the soil moisture sampling period, the trade-off was between data resolution and system energy consumption. Excessively long sampling intervals might miss the precise moment when the moisture returned to the baseline, introducing… The measurement error is high, while excessively short intervals increase the data processing load on the microcontroller. Therefore, the setting rule is that the sampling frequency should be higher than twice the estimated fastest water consumption rate. Specifically, this is determined through preliminary experiments under the most extreme transpiration conditions. The sampling period is set to be no less than 15 minutes, with a 10-second sampling period to ensure sufficient data collection points throughout the attenuation process, thereby optimizing system resource usage while maintaining accuracy.

[0036] The experiment was conducted within a simulated summer sunny day cycle. The greenhouse environment gradually increased in temperature and light from early morning, reaching its peak in the afternoon. Both systems operated autonomously according to their built-in logic. At 9:00 AM, the ambient temperature was moderate and evaporation was low. Neither system triggered irrigation or fertilization operations, and their states were basically the same. Starting at 11:00 AM, as the ambient temperature rapidly increased, the soil tensiometer reading of the control group dropped rapidly. The system determined this to be water stress and triggered a fertilization operation at 12:30 PM. Correspondingly, during the same time period, the soil moisture sensor deployed in the root zone of the present invention also detected a decreasing trend in moisture. However, by performing data differential analysis with the second soil moisture sensor in the bare soil area, the system identified that the water consumption was mainly caused by intense evaporation from the soil surface. Table 1 shows a comparison of the key parameters of the two decision-making systems at different times. Referring to Table 1, the data in this table records a series of key parameters that led the two systems to make different decisions.

[0037] Table 1: A comparison record of key parameters of the two decision-making systems at different times.

[0038]

[0039] Table 1 shows that at 12:30, the control group made fertilization decisions because they failed to distinguish between physical evaporation and physiological water absorption, while the system of this invention, through a dual-sensor differential architecture, measured the net transpiration response time. It only took 5.3 minutes, calculated in reverse. The value was at an extremely low level, indicating that the crop had entered a state of transpiration inhibition due to high temperature stress, resulting in weak nutrient absorption capacity. The system therefore autonomously suspended fertilization decisions. Until 4:00 PM, when the ambient temperature dropped, the measured value... Restored to 22.1 minutes. Once the soil temperature returns to normal, the system determines that the crop's physiological functions have recovered before triggering fertilization. This process shifts the decision-making basis from indirect and easily affected soil chemical indicators to transpiration rate, which directly reflects the crop's true physiological state.

[0040] Example 3: This example combines Figures 1 to 3 This section describes a fertilization decision-making system that integrates soil and rice growth data collection and analysis. Figure 1 As shown in the diagram, the system first includes an acoustic sensor for monitoring and verifying acoustic signals in the irrigation system. A soil moisture sensor in the root zone and a second soil moisture sensor in the bare soil zone are used to monitor soil moisture changes in the rice root zone and the bare soil zone, respectively. The net transpiration response time is then calculated through a differential correction mechanism. The microcontroller, as the core of control and analysis, receives sensor data, calculates the transpiration rate through an inversion formula, and determines the nutrient requirement level of the rice based on this rate, combined with growth stage information and a preset lookup table method. In addition, the system also includes components such as photoresistor transmittance monitoring, infrared temperature sensor for root health diagnosis, and image sensor for lodging risk assessment. Through collaborative work, these components form a fertilization decision output, ultimately generating a fertilization amount level and executing the corresponding fertilization operation.

[0041] like Figure 2 As shown, the system first initiates the diagnostic process. The microcontroller controls the irrigation valve to apply an irrigation pulse of a preset duration. This operation uses acoustic sensors to monitor the valve's acoustic signal in real time and compares it with a reference fingerprint for acoustic fingerprint verification, ensuring the irrigation is effective. Subsequently, dual soil moisture sensors deployed in the field begin monitoring the time required for moisture to drop from its peak level back to the baseline level. One set of sensors is located in the rice root zone, while the other set is deployed in bare soil without roots for differential correction, calculating the net transpiration response time. Differential correction calculation of net evapotranspiration response time To eliminate interference from soil surface evaporation, the microcontroller then... Calculate the transpiration rate of crops According to the inversion model Inverted evaporation rate The system obtains dynamic data on water absorption capacity and, based on transpiration rate, determines nutrient requirement levels to differentiate between high, medium, and low nutrient demand states for the current crop. To achieve more precise fertilization decisions, the system also integrates a root health diagnostic infrared temperature sensor to monitor the cooling rate after foliar micro-spraying and assess root function. Combined with a photoresistor to obtain rice canopy transmittance, the system identifies crop growth stages, distinguishing between the jointing and booting stages and the heading and grain-filling stages. An image sensor is used for image recognition, extracting the angular standard deviation of the principal shadow direction to assess rice tilt and perform lodging risk assessment. All of this multi-dimensional information is uniformly input into a comprehensive fertilization adjustment module to optimize fertilization decisions based on physiological and environmental factors. The system outputs warning information or directly issues fertilization commands based on risk and diagnostic results, ensuring precise matching of fertilization amount with the crop's actual absorption capacity and avoiding resource waste and misaligned fertilization timing.

[0042] like Figure 3 As shown in the figure, the ambient temperature and transpiration response time are represented by two different curves, with triangles indicating the change over a day and circles indicating the trend. The horizontal axis represents time, the left side of the vertical axis represents the ambient temperature in °C, and the right side represents the change in transpiration response time in °C. The unit is minutes. The curve in the figure shows that as time goes by, the ambient temperature first rises and then falls, gradually increasing from about 24°C at 9:00 to about 32°C in the afternoon, and then starting to fall after 16:00 in the afternoon. The evaporation response time shows a clear inverse relationship with the trend of ambient temperature change. The evaporation response time gradually increases from the morning and reaches its peak in the afternoon before rapidly decreasing. Especially after the temperature drops, the evaporation response time also recovers rapidly.

[0043] like Figure 4 As shown, the microcontroller first sends an irrigation pulse command for 90 seconds to activate the irrigation valve and monitors the acoustic signal through an acoustic sensor to ensure the effectiveness of the irrigation pulse execution. Next, the soil moisture sensor monitors changes in soil moisture, especially when the moisture level drops back to the baseline value. The soil moisture value is determined through repeated measurements and compared with the time interval between moisture drops back to the baseline. The microcontroller then uses this time interval... The transpiration rate was inverted and calculated using the formula. The system assesses the transpiration rate of rice, providing basic data for fertilization decisions. If the transpiration rate assessment indicates that the rice has high nutrient requirements, the system will generate corresponding fertilization recommendations through the microcontroller and initiate fertilization operations. The entire process is facilitated by the collaboration of various sensors and control units through real-time data interaction and feedback mechanisms, ensuring that each fertilization can be adjusted according to actual needs.

[0044] Example 4: This example details a set of key parameter calibration and model building procedures required in the initial deployment of the system. Its purpose is to adapt a general algorithm model to the soil, climate, and rice variety of a specific field, ensuring that the system's subsequent autonomous operation decisions have a traceable physical basis. The procedure begins with constructing the acoustic sensor reference fingerprint, which serves as the basis for physical execution verification in the system. This process requires that, under the condition that the water pressure in the field pipeline network is stable within its normal operating range, the microcontroller continuously triggers the irrigation valve to perform ten standard opening and closing cycles. During this period, the acoustic sensor, at a sampling rate of 44.1 kHz, completely records the time-domain waveform of the acoustic signal generated in each cycle. The ten sets of time-domain signals collected are then processed by the microcontroller or its connected computing device. The Lie transform converts the signal from the time domain to the frequency domain, thereby obtaining the power spectral density of each signal group. Given that the mechanical vibration and water flow sound during normal valve operation have stable energy distribution characteristics in a specific frequency band, while potential faults will generate abnormal energy in other frequency bands, the data structure of the reference fingerprint is defined as a vector composed of the average power values ​​in pre-selected key frequency bands. The key frequency bands are set as band one from 100 to 250 Hz and band two from 1000 to 1500 Hz. By calculating the mean and standard deviation of the average power values ​​in these key frequency bands for ten normal operations, the system establishes an acoustic reference fingerprint that characterizes the normal operating state. In subsequent operations, if the power value of any real-time signal in these frequency bands deviates from its mean by more than three standard deviations, it can be judged as an execution anomaly.

[0045] Subsequently, the evaporation rate inversion model of the system core, namely... Key constants in The nutrient requirement level lookup method was used for joint calibration. This process was completed in pot experiments using undisturbed soil from the target field and rice seedlings of the same variety. The experimental group was divided into three groups, and different concentrations of nutrient solution were applied to maintain three distinct nitrogen nutrient levels: high, medium, and low. After the rice growth stabilized, the irrigation pulse diagnosis process of this invention was repeated for each pot group, and the net transpiration response time measured by the system under different nutrient levels was recorded. The stable range of the low nitrogen stress group Record as Interval, high nitrogen-sufficient group Record as This establishes the nutrient requirement levels and intervals. The correspondence between measured values ​​is solidified into a lookup table method within the system; based on this, constants are determined. The precise value was obtained by introducing a lysimeter to analyze one group of potted plants. Simultaneously with the measurement, the evapotranspiration rate per unit time was independently determined, i.e., the actual evapotranspiration rate. Through multiple synchronous measurements, a series of ( , The data points are paired, and a linear regression is performed on these data points in this coordinate system. The slope of the resulting straight line is a specific constant for this soil and rice variety. Finally, the image analysis module used in the system to assess lodging risk is calibrated; this procedure is performed during the late jointing stage of the crop. Under windless conditions, the image sensor acquires a set of overhead images of the canopy of healthy, upright rice plant communities. For each image, the microcontroller executes an internal image processing algorithm, which includes grayscale conversion, binarization using Otsu's method to separate plant shadows from the background, principal component analysis of the obtained shadow region pixel set, and the direction vector angle of the first principal component is defined as the principal direction angle of the shadow. By statistically analyzing the principal direction angles of all independent shadow regions in an image, an angular standard deviation is calculated. Calculated from multiple healthy community images Perform statistical analysis and take the average value. with standard deviation Initial warning threshold for lodging risk Based on statistical principles, it is set as Add three times When the standard deviation of the angle measured in the field exceeds this When the tilt dispersion of the plant has deviated from the normal fluctuation range of its healthy upright state, all key parameters and models of the system that need to be localized and adapted have been calibrated, providing the technical conditions for reliable autonomous operation in specific sites.

[0046] Example 5: This example details two key diagnostic and verification logics. The first is the offline calibration of key thresholds in the root function assessment process. This procedure aims to provide an objective quantitative basis for the system to distinguish between physiological transpiration inhibition and pathological root decline. The calibration process involves selecting a group of healthy rice samples in the jointing stage. Under controlled conditions, the samples are induced into a non-stressed low-transpiration state by increasing air humidity and reducing light intensity. Subsequently, canopy micro-spraying is triggered, and an infrared temperature sensor is used to record the rate of temperature change on the leaf surface due to evaporative cooling. This process is repeated multiple times to obtain a distribution of normal cooling rates for a group of healthy plants under low transpiration demand, and the average value is calculated. Ultimately, the threshold for judging the health status of the root system. The value was set to 50 percent of this average, when the rate of change in blade temperature monitored in the field was lower than this. At that time, the system will determine that there is an abnormality in the crop root system function.

[0047] Secondly, there is an autonomous data consistency verification logic designed to prevent long-term sensor drift or latent faults. This logic is executed periodically by the microcontroller by comparing data streams from different sensors and those with inherent agronomical correlations. Taking the determination of growth stage as an example, the system not only relies on the transmittance measured by the photoresistor placed under the canopy, but also compares this data with the accumulated temperature model since the rice transplanting date, as well as the recently continuously measured net transpiration response time. Cross-validation of macroscopic change trends; under normal circumstances, the decrease in light transmittance should be related to the accumulation of accumulated temperature and Under the same illumination conditions, the trend of gradual shortening is synchronized. If the system detects a drastic jump in the transmittance data of the photoresistor in a short period of time that is inconsistent with the seasonal and accumulated temperature models, or if the growth stage it indicates is inconsistent with... The data from the photoresistor showed a significant discrepancy over a long period. The microcontroller then marked the data from this photoresistor as unreliable and suspended the determination of the growth stage based on it. Instead, it adopted a backup model based on historical data and accumulated temperature for decision-making. At the same time, it issued sensor maintenance instructions to the user. This mechanism can effectively avoid long-term erroneous fertilization decisions caused by physical damage or contamination of a single sensor.

[0048] Example 6: This example describes the systematic calibration procedure for the core parameters of the thermal response self-calibration module integrated into the soil moisture sensor for compensating for temperature drift. This procedure aims to determine the specific parameters of a constant-energy micro-current pulse, ensuring it generates a stable and measurable Joule heating effect under different soil moisture conditions without affecting the sensor or root environment. The calibration process is first conducted on an offline experimental platform using undisturbed soil samples from the target field, equipped with a device to precisely control their moisture content. The soil samples are subjected to three moisture states sequentially: air-dried, field-hold, and saturated. Under each moisture state, the soil moisture sensor embedded in the soil sample is... A series of microcurrent pulses with different parameters are applied to the sensor. The pulse parameters vary in terms of voltage and duration. The voltage ranges from 1 volt to 5 volts, and the duration ranges from 50 milliseconds to 200 milliseconds. For each parameter combination, the instantaneous temperature rise is recorded by the thermistor built into the sensor. By analyzing this dataset, a pulse parameter combination is selected. The constraints that this combination must meet are: under all humidity conditions, the instantaneous temperature rise is greater than the minimum reliable detection threshold of 0.1 degrees Celsius, and under any condition, the temperature rise does not exceed the safe upper limit of 2 degrees Celsius. Thus, an effective pulse parameter working matrix is ​​determined for the sensor of this specific soil type.

[0049] After the system is deployed in the field, before the microcontroller performs its first self-calibration procedure each day, it first measures the initial soil resistivity using the sensor electrodes. This value is directly related to the current soil moisture content. Then, based on this resistivity value, the microcontroller queries and selects a corresponding pulse voltage and duration combination from the previously calibrated parameter working matrix to ensure that the Joule thermal energy injected in this pulse is close to a preset constant value. After applying this dynamically selected microcurrent pulse, the system records the actual instantaneous temperature rise and uses this value as an index to query the preset correction factor lookup table to obtain a real-time correction coefficient for correcting subsequent soil moisture monitoring data. This procedure, which combines offline calibration with online selection, ensures the consistency and reliability of the thermal response self-calibration function in the variable field environment.

[0050] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0051] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A fertilization decision-making system integrating soil and rice growth information data collection and analysis, characterized in that, include: The irrigation valve is configured to apply an irrigation pulse to the rice root zone for a preset duration. A soil moisture sensor is configured to monitor the time interval required for the soil moisture in the rice root zone to fall from a peak level back to the baseline level before the irrigation pulse was applied after the irrigation pulse has ended. The microcontroller, connected to the irrigation valve and soil moisture sensor, is configured to: determine the transpiration rate of rice based on the time interval monitored by the soil moisture sensor, wherein the transpiration rate is inversely proportional to the time interval; determine the nutrient requirement level of rice based on the transpiration rate and a preset lookup table method; and generate and output the fertilization amount level to guide the next fertilization based on the nutrient requirement level. The system also includes an acoustic sensor configured to be in close proximity to the irrigation valve or a nearby water pipe; the microcontroller is configured to: store the acoustic signal characteristics generated when the irrigation valve is normally open as a reference fingerprint; listen to the real-time acoustic signal received by the acoustic sensor each time an irrigation pulse is applied, obtain the characteristics of the real-time acoustic signal; and compare the characteristics of the real-time acoustic signal with the reference fingerprint, and if they do not match, output an indication that the irrigation pulse is invalid. The system also includes a second soil moisture sensor, which is identical to the soil moisture sensor and is configured to be deployed in a bare soil area where no rice roots are distributed. The microcontroller is configured to: simultaneously monitor the first time interval measured by the soil moisture sensor and the second time interval measured by the second soil moisture sensor when an irrigation pulse is applied; and correct for soil surface evaporation interference based on the difference between the first time interval and the second time interval to obtain the net evaporation response time, which is used as the basis for judging the evaporation rate. The soil moisture sensor integrates a thermistor, and the microcontroller is configured to: apply a constant-energy microcurrent pulse to the electrodes of the soil moisture sensor before the monitoring time interval, causing Joule heating in the soil around the probe; monitor the instantaneous temperature rise of the probe through the thermistor; and determine a correction coefficient for correcting subsequent soil moisture monitoring data by querying a preset correction factor lookup table based on the instantaneous temperature rise.

2. The fertilization decision-making system integrating soil and rice growth information data collection and analysis as described in claim 1, characterized in that, The microcontroller determines the transpiration rate of rice based on the following formula: ,in, Indicates the transpiration rate. This represents a preset constant. This indicates the time interval required for soil moisture to fall from its peak level back to the baseline level, and the time interval is measured in minutes.

3. The fertilization decision-making system integrating soil and rice growth information data collection and analysis as described in claim 1, characterized in that, The fertilization level is also adjusted based on information about the rice's growth stage.

4. The fertilization decision-making system integrating soil and rice growth information data collection and analysis as described in claim 3, characterized in that, The system also includes a photoresistor, which is placed on the ground below the rice canopy to sense the light transmittance through the rice canopy. The microcontroller is configured to: record the initial baseline value of light transmittance at the beginning of rice transplanting; continuously track the decreasing trend of light transmittance; and make judgments based on the ratio of light transmittance to the initial baseline value: when the ratio is less than or equal to 70%, it is identified as the jointing and booting stage; when the ratio is less than or equal to 30%, it is identified as the heading and grain-filling stage.

5. The fertilization decision-making system integrating soil and rice growth information data collection and analysis as described in claim 1, characterized in that, The microcontroller is also configured to initiate the root system function judgment process when the time interval is greater than the preset first diagnostic threshold. The root system function assessment process includes: controlling the irrigation valve to apply micro-spraying to the rice canopy for a preset duration; monitoring the rate of change of the surface temperature of the rice canopy leaves after micro-spraying using an infrared temperature sensor; and determining the root health status of the rice based on the comparison between the rate of temperature change and a preset threshold, and adjusting the fertilization level according to the root health status.

6. The fertilization decision-making system integrating soil and rice growth information data collection and analysis as described in claim 1, characterized in that, The system also includes an image sensor configured to acquire images of the rice canopy; the microcontroller is further configured to: identify the shadow areas cast by the rice plants in the image; extract the principal direction angle of the shadow areas; calculate the angular standard deviation of the principal direction angle to determine the degree of tilt dispersion of the rice plants; and determine the lodging risk level of the rice based on the comparison between the angular standard deviation and a preset lodging threshold.

7. The fertilization decision-making system integrating soil and rice growth information data collection and analysis as described in claim 1, characterized in that, When determining the nutrient requirement level of rice, the microcontroller compares the time interval with a preset time threshold: when the time interval is less than the first time threshold, it is determined to be a high nutrient requirement level; when the time interval is between the first time threshold and the second time threshold, it is determined to be a medium nutrient requirement level; when the time interval is greater than the second time threshold, it is determined to be a low nutrient requirement level.

Images