Cotton irrigation intelligent control method and system based on stem flow data

Abstract

The present application relates to the technical field of agricultural irrigation control, in particular to a cotton irrigation intelligent control method and system based on stem flow data, comprising the following steps: collecting data to calculate stem flow rate and build a monitoring data set, extracting morning data to calculate morning light flow conversion coefficient, deducing noon theoretical potential stem flow rate, comparing the theoretical value with the actual value to calculate instantaneous deficit depth, calculating deficit width based on the length of time that is continuously lower than the theoretical value, screening the maximum concave depth, multiplying the maximum concave depth by the deficit width to calculate the stomatal closure resistance index, and building an opening instruction when the index is greater than the threshold value. In the present application, the potential rate is deduced by using the morning conversion relationship, the deficit depth and width are quantified, the stomatal closure resistance index is calculated to evaluate water stress, the instruction is triggered based on the physiological response, the water requirement diagnosis is realized, the timely water replenishment is ensured, the irrigation precision is improved, and the photosynthetic efficiency is guaranteed.

Description

Technical Field

[0001] This invention relates to the field of agricultural irrigation control technology, and in particular to a smart control method and system for cotton irrigation based on stem flow data. Background Technology

[0002] The field of agricultural irrigation control technology mainly involves engineering techniques for regulating and managing the water environment of farmland using automated equipment and monitoring instruments, aiming to meet the needs of crop growth and development by artificially supplementing water. Traditional intelligent control methods for cotton irrigation involve burying soil moisture sensors in the soil layer of the cotton planting area. The collected soil moisture content electrical signals are transmitted to a central control box via data transmission cables. The programmable logic controller (PLC) in the control box reads the signal value and compares it with a preset humidity threshold. When the detected value is lower than the set lower limit, the controller outputs a drive signal to open a solenoid valve or water pump connected to the water supply network, thereby implementing irrigation.

[0003] Existing technologies rely on soil moisture sensors buried underground to obtain soil moisture content data, which only reflects the dryness and wetness of the external soil environment and cannot directly characterize the actual physiological water demand of cotton plants. Soil moisture changes lag behind the instantaneous transpiration demand of plants caused by changes in environmental light. Relying on fixed soil thresholds ignores the biological regulation mechanisms of crops under different light conditions, making it difficult to accurately match water supply operations with the rhythm of crop physiological activities. This leads to low water resource utilization efficiency or loss of yield due to delayed response to water stress, missing the optimal growth and development window of crops. Summary of the Invention

[0004] The purpose of this invention is to overcome the shortcomings of the existing technology and propose a smart control method and system for cotton irrigation based on stem flow data.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: a smart control method for cotton irrigation based on stem flow data, comprising the following steps:

[0006] S1: Collect thermoelectric potential signals and ambient light intensity using a thermal diffusion stem flow sensor and a photosynthetically active radiation sensor. Calculate stem flow rate based on the thermoelectric potential signals. Construct a cotton physiological and ecological monitoring dataset by associating stem flow rate, ambient light intensity, and collection timestamp.

[0007] S2: Extract the stem flow rate and the ambient light intensity of the morning time window from the cotton physiological and ecological monitoring dataset, calculate the average ratio to generate the morning light flow conversion coefficient, and multiply the ambient light intensity at noon by the morning light flow conversion coefficient to calculate the theoretical potential stem flow rate.

[0008] S3: Extract the stem flow rate at noon from the cotton physiological and ecological monitoring dataset as the actual stem flow rate, subtract the actual stem flow rate from the theoretical potential stem flow rate to calculate the instantaneous deficit depth, calculate the deficit width based on the duration during which the actual stem flow rate is continuously lower than the theoretical potential stem flow rate, and filter the maximum instantaneous deficit depth to generate the maximum depression depth.

[0009] S4: Calculate the pore closure resistance index by multiplying the maximum indentation depth by the defect width, compare the pore closure resistance index with the benchmark threshold, and construct an opening command when the pore closure resistance index is greater than the benchmark threshold.

[0010] As a further aspect of the present invention, the calculation of the stem flow rate and the construction of the cotton physiological and ecological monitoring dataset specifically include:

[0011] The system receives the raw analog voltage transmitted by the thermal diffusion stem flow sensor, uses an analog-to-digital converter to resolve the analog voltage into a value characterizing the temperature difference between probes, combines it with the pre-calibrated maximum temperature difference reference value under zero flow conditions, calculates the liquid flow density through a nonlinear mapping relationship, and combines it with the water-conducting area of ​​the cotton stem to generate the stem flow rate.

[0012] The real-time photon flux reading uploaded by the photosynthetically active radiation sensor is acquired synchronously, and the reading is subjected to moving average filtering to eliminate instantaneous high-frequency noise caused by cloud cover, thereby generating the ambient light intensity.

[0013] A time-series database with the collection timestamp as the unique index is established. The stem flow rate after time-series alignment and the ambient light intensity are written into the corresponding fields of the database and associated with the current crop growth stage identifier to construct the cotton physiological and ecological monitoring dataset.

[0014] As a further aspect of the present invention, the process of generating the morning optical flow conversion coefficient and calculating the theoretical potential stem flow rate specifically includes:

[0015] The cotton physiological and ecological monitoring dataset is traversed, and a specific time period from sunrise to noon is locked as the morning time window. The stem flow rate and the ambient light intensity within the window are extracted point by point. Invalid data points with light intensity lower than the preset effective photosynthetic starting point are removed to generate an effective morning data sequence.

[0016] For each group of effective morning data sequences, a division operation is performed between the stem flow rate and the ambient light intensity to obtain a series of instantaneous conversion ratios. An arithmetic mean operation is performed on all instantaneous conversion ratios after removing outliers to establish the morning light flow conversion coefficient that can characterize the photosynthetic driving capacity of crops under water stress.

[0017] The ambient light intensity during the midday high-radiation period is extracted and directly multiplied by the morning light flow conversion coefficient to simulate the upper limit of the transpiration rate that crops should reach under ideal water supply conditions, thus generating the theoretical potential stem flow rate.

[0018] As a further aspect of the present invention, the process of generating the instantaneous defect depth, defect width, and maximum depression depth specifically includes:

[0019] The theoretical potential stem flow rate and the actual stem flow rate at the same timestamp are differentially calculated. When the difference is positive, the difference is retained as the effective water deficit. When the difference is negative or zero, the deficit is set to zero, thereby generating the instantaneous deficit depth that changes continuously with time.

[0020] Identify the continuous time interval where the instantaneous deficit depth is continuously greater than zero, count the total number of time steps contained in the interval, and multiply the total number by the time interval of a single sampling period to quantify the duration of continuous water stress experienced by the crop and generate the deficit width.

[0021] Within the aforementioned continuous time interval, an extreme value search is performed on the instantaneous deficit depth at all times to screen out the peak difference in the water deficit process, which characterizes the strongest water supply and demand imbalance experienced by the crop during that period, and generates the maximum depression depth.

[0022] As a further aspect of the present invention, the process of calculating the pore closure resistance index and constructing the opening command specifically includes:

[0023] The maximum indentation depth and deficit width are obtained, and a multiplication operation is performed to construct a two-dimensional rectangular area model that can reflect the cumulative effect of water deficit intensity and duration. The value of the area model is defined as a quantitative index characterizing the degree of crop stomatal regulation fatigue, and the stomatal closure resistance index is generated.

[0024] The preset irrigation decision benchmark parameter is invoked. This parameter is set based on the drought resistance characteristics of the cotton at the current growth stage. The stomatal closure resistance index is compared with the benchmark parameter to determine whether the current crop is in an irreversible stomatal closure critical state.

[0025] When the stomatal closure resistance index exceeds the benchmark parameter, the crop is determined to be in a state of water stress that urgently needs water replenishment. A control signal containing the target irrigation amount and execution priority is immediately generated to construct an opening command.

[0026] As a further aspect of the present invention, the calculation process of the stem flow rate specifically includes:

[0027] Obtain the maximum temperature difference formed by the heat diffusion probe at night due to the cessation of evaporation, calculate the ratio of the relative difference between the current measured temperature difference and the maximum temperature difference, and use this ratio to construct a dimensionless coefficient reflecting the degree of heat dissipation.

[0028] An empirical power function model is used to perform a nonlinear transformation on the dimensionless coefficient, calculate the liquid flow velocity per unit sapwood area, and multiply the liquid flow velocity by the area of ​​the sapwood region with water transport function in the cross-section of the cotton stalk to generate the stalk flow rate.

[0029] As a further aspect of the present invention, the extraction process of the stem flow rate and the ambient light intensity during the morning time window specifically includes:

[0030] Obtain local sunrise time data for the day, set the time period between one and three hours after sunrise as the filter interval, monitor the rate of change of light intensity within this interval, and identify and eliminate unstable periods of intense light fluctuations caused by rapid cloud movement.

[0031] Filter the continuous data segment within the interval where the stem flow rate increases synchronously with the ambient light intensity and the correlation coefficient is higher than the preset linear threshold. Ensure that the extracted data reflects the physiological state where the stomata are fully open and not inhibited by midday high temperature. Extract the values ​​within this data segment as the stem flow rate and ambient light intensity in the morning time window.

[0032] As a further aspect of the present invention, the process of generating the maximum indentation depth specifically includes:

[0033] A sliding window of dynamic length is constructed, and the window is used to perform a traversal scan on the time series of the instantaneous deficit depth. The fluctuation variance of the data within each window is calculated to eliminate abrupt spikes caused by sensor random noise.

[0034] In the defect curve after noise interference is eliminated, the inflection point when the slope of the positioning curve changes from positive to negative is located. The amplitude value corresponding to the inflection point is extracted. In the case of multiple inflection points, the vertex with the largest absolute amplitude value is selected by comparison and determined as the maximum depression depth that characterizes the extreme state of this stress event.

[0035] As a further aspect of the present invention, the process of constructing the activation instruction specifically includes:

[0036] Before generating the control signal, the current reading of the soil moisture sensor is read as an auxiliary verification parameter to determine whether the soil moisture content is unsaturated, so as to prevent accidental irrigation caused by the lag in stem flow response when rainfall occurs.

[0037] After confirming that the soil moisture content has not reached the saturation threshold, the pulse duration of the required replenishment water is dynamically calculated based on the proportion of the stomatal closure resistance index exceeding the benchmark threshold. This pulse duration is then encoded into a solenoid valve drive level sequence to generate an opening command.

[0038] A cotton irrigation intelligent control system based on stem flow data, the system being used to implement the aforementioned cotton irrigation intelligent control method based on stem flow data, the system comprising:

[0039] The data acquisition and processing module is used to acquire thermoelectric potential signals and ambient light intensity through a thermal diffusion stem flow sensor and a photosynthetically active radiation sensor, calculate stem flow rate based on the thermoelectric potential signal, and construct the cotton physiological and ecological monitoring dataset by associating stem flow rate, ambient light intensity and acquisition timestamp.

[0040] The potential transpiration prediction module is used to extract the stem flow rate and the ambient light intensity of the morning time window from the cotton physiological and ecological monitoring dataset, calculate the ratio mean to generate the morning light flow conversion coefficient, and multiply the ambient light intensity at noon by the morning light flow conversion coefficient to calculate the theoretical potential stem flow rate.

[0041] The water deficit analysis module is used to extract the stem flow rate at noon from the cotton physiological and ecological monitoring dataset as the actual stem flow rate, subtract the actual stem flow rate from the theoretical potential stem flow rate to calculate the instantaneous deficit depth, calculate the deficit width based on the duration during which the actual stem flow rate is continuously lower than the theoretical potential stem flow rate, filter the maximum instantaneous deficit depth, and generate the maximum depression depth.

[0042] The irrigation decision control module is used to calculate the pore closure resistance index by multiplying the maximum depression depth by the deficit width, compare the pore closure resistance index with a benchmark threshold, and generate an opening command when the pore closure resistance index is greater than the benchmark threshold.

[0043] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0044] In this invention, physiological and ecological monitoring data are constructed by collecting stem flow rate and ambient light intensity. The theoretical potential stem flow rate at noon is deduced by using the morning light flow conversion relationship. The theoretical value is compared with the actual measured value to quantify the instantaneous deficit depth and deficit width. The stomatal closure resistance index is calculated to accurately assess the degree of water stress inside the cotton plant. Irrigation instructions are triggered based on the physiological response characteristics of the crop itself rather than the external soil environment, realizing water demand diagnosis from the perspective of the plant's life itself. This ensures timely water replenishment before excessive stomatal closure hinders photosynthesis, improving irrigation accuracy while ensuring the efficiency of crop photosynthetic accumulation. Attached Figure Description

[0045] Figure 1 This is a flowchart of the intelligent cotton irrigation control method based on stem flow data according to the present invention.

[0046] Figure 2 This is a flowchart of the stem flow rate calculation and dataset construction process of this invention;

[0047] Figure 3 This is a flowchart illustrating the calculation of the theoretical potential stem flow rate of this invention.

[0048] Figure 4 This is a flowchart illustrating the generation process of the maximum indentation depth and defect width in this invention.

[0049] Figure 5 This is a flowchart of the calculation and instruction construction for the pore closure resistance index of the present invention. Detailed Implementation

[0050] To make the objectives, technical solutions, and advantages of this invention clearer, the software-based technical solution is described in detail below with reference to system architecture diagrams and embodiments. It should be understood that the specific embodiments described herein are only for explaining the technical solutions of this invention and do not constitute a limitation on the scope of protection.

[0051] In the description of this invention, the system architecture relationships or data processing flows indicated by terms such as "layer," "module," "interface," "data flow," "client," and "server" are all defined based on the architecture diagram or flowchart corresponding to the embodiments. This way of describing is only used to clearly illustrate the logical relationships between the elements in the technical solution, and not to limit the physical deployment form. The term "multiple" includes two or more technical units, including but not limited to multiple data nodes, processing threads, service instances, or functional components and other scalable elements. The specific number is determined according to the actual business scenario and needs to be specifically stated.

[0052] Please see Figure 1 and Figure 2 This invention provides a technical solution: a smart control method for cotton irrigation based on stem flow data, comprising the following steps:

[0053] S1: Collect thermoelectric potential signals and ambient light intensity using a thermal diffusion stem flow sensor and a photosynthetically active radiation sensor. Calculate stem flow rate based on the thermoelectric potential signals. Construct a cotton physiological and ecological monitoring dataset by correlating stem flow rate, ambient light intensity, and collection timestamps.

[0054] The calculation of stem flow rate and the construction of the cotton physiological and ecological monitoring dataset specifically include:

[0055] The system receives the raw analog voltage transmitted by the thermal diffusion stem flow sensor, uses an analog-to-digital converter to resolve the analog voltage into a value characterizing the temperature difference between probes, combines it with the pre-calibrated maximum temperature difference reference value under zero flow conditions, calculates the liquid flow density through a nonlinear mapping relationship, and combines it with the water-conducting area of ​​the cotton stem to generate the stem flow rate.

[0056] The real-time photon flux reading uploaded by the photosynthetically active radiation sensor is acquired synchronously, and the reading is subjected to moving average filtering to eliminate instantaneous high-frequency noise caused by cloud obstruction, thereby generating ambient light intensity.

[0057] A time-series database with the collection timestamp as the unique index was established. The stem flow rate and ambient light intensity after time-series alignment were written into the corresponding fields of the database and associated with the current crop growth stage identifier to construct a cotton physiological and ecological monitoring dataset.

[0058] The calculation process for stem flow rate specifically includes:

[0059] Obtain the maximum temperature difference formed by the heat diffusion probe at night due to the cessation of evaporation, calculate the ratio of the relative difference between the current measured temperature difference and the maximum temperature difference, and use this ratio to construct a dimensionless coefficient reflecting the degree of heat dissipation.

[0060] An empirical power function model is used to perform a nonlinear transformation on the dimensionless coefficient, calculate the liquid flow velocity per unit sapwood area, and multiply the liquid flow velocity by the area of ​​the sapwood region with water transport function in the cross-section of the cotton stalk to generate the stalk flow rate.

[0061] In the data acquisition and preprocessing stage, a physical signal sensing channel is first established using a heat-diffusion stem flow sensor deployed at the junction of the phloem and xylem of the cotton stem, and a photosynthetically active radiation sensor installed above the crop canopy. For the stem flow rate calculation process, the data acquisition module receives a microvolt-level analog voltage transmitted by the heat-diffusion probe. This analog voltage is composed of the thermoelectric potential generated by the copper-constantan thermocouple at different temperatures. A high-precision analog-to-digital converter converts this analog voltage signal into a digital signal, and a pre-stored copper-constantan thermocouple calibration table is used to map the voltage value to the instantaneous temperature difference between the heating probe and the reference probe. A zero-flow reference period is set between 2:00 AM and 4:00 AM daily. During this period, a sliding window algorithm continuously monitors the stability of the temperature difference, selecting the period with the smallest temperature difference fluctuation variance, and locking the average temperature difference within this period as the maximum temperature difference benchmark value under zero-flow conditions.

[0062] The aforementioned heat diffusion stem flow sensor is an instrument that measures the stem flow rate of plants based on the principle of heat dissipation. It typically includes a heating probe and a reference probe, and the sap flow velocity within the plant is inverted by measuring the temperature difference between the two probes.

[0063] Subsequently, a nonlinear calculation of the sap flow density is performed. The measured temperature difference at the current moment is obtained, and the difference between the maximum temperature difference benchmark and the current measured temperature difference is calculated. This difference is then divided by the current measured temperature difference to obtain a dimensionless temperature difference coefficient reflecting the intensity of heat dissipation. A power function model based on the Granier empirical formula is invoked, using this dimensionless temperature difference coefficient as the base and setting the empirical exponent parameter to 1.23. The sap flow velocity per unit sapwood area is calculated through exponentiation. To obtain the stem flow rate of the entire cotton plant, the water-conducting sapwood area parameter of the cotton stem, pre-measured and entered using the growth cone drilling method, is used. The calculated sap flow velocity is multiplied by this sapwood area to generate a real-time stem flow rate in grams per hour.

[0064] During the processing of ambient light intensity, the photon flux readings output by the photosynthetically active radiation sensor were simultaneously acquired. To eliminate instantaneous high-frequency light noise caused by rapid cloud cover or leaf swaying, a 10-minute moving average filter was constructed. This filter manages the light data queue in a sequential storage manner, performs an arithmetic average operation on all readings in the queue, and outputs the smoothed ambient light intensity. Finally, a time-series database was established with millisecond-level acquisition timestamps as the unique primary key. The time-aligned stem flow rate values ​​and ambient light intensity values ​​were written into the corresponding columns of the database and associated with the current crop growth stage identifier, thus completing the construction of the cotton physiological and ecological monitoring dataset.

[0065] Table 1 shows the data examples, illustrating the collection and calculation results of key parameters in this step.

[0066]

[0067] As shown in Table 1, the stem flow rate calculated by collecting the original voltage and combining it with the maximum temperature difference benchmark value of 4.80 degrees Celsius can reflect the actual water transport in the crop. Meanwhile, the light data has been filtered.

[0068] Please see Figure 1 and Figure 3 S2: Extract stem flow rate and ambient light intensity during the morning time window from the cotton physiological and ecological monitoring dataset, calculate the average ratio to generate the morning light flow conversion coefficient, and multiply the ambient light intensity at noon by the morning light flow conversion coefficient to calculate the theoretical potential stem flow rate.

[0069] The process of generating the morning optical flow conversion coefficient and calculating the theoretical potential stem flow rate specifically includes:

[0070] By traversing the cotton physiological and ecological monitoring dataset, a specific time period from sunrise to noon was selected as the morning time window. The stem flow rate and ambient light intensity within this window were extracted point by point. Invalid data points with light intensity lower than the preset effective photosynthetic starting point were removed to generate an effective morning data sequence.

[0071] For each group of effective morning data sequences, a division operation was performed on the stem flow rate and the ambient light intensity to obtain a series of instantaneous conversion ratios. The arithmetic mean of all instantaneous conversion ratios after removing outliers was then performed to establish the morning light flow conversion coefficient that can characterize the photosynthetic driving capacity of crops under water stress.

[0072] The ambient light intensity during the midday high-radiation period is extracted and directly multiplied with the morning light flow conversion coefficient to simulate the upper limit of the transpiration rate that crops should reach under ideal water supply conditions, thus generating the theoretical potential stem flow rate.

[0073] The extraction process of stem flow rate and ambient light intensity during the morning time window specifically includes:

[0074] Obtain local sunrise time data for the day, set the time period between one and three hours after sunrise as the filter interval, monitor the rate of change of light intensity within this interval, and identify and eliminate unstable periods of intense light fluctuations caused by rapid cloud movement.

[0075] Continuous data segments within this interval that show a synchronous increase in stem flow rate with ambient light intensity and a correlation coefficient higher than a preset linear threshold are selected to ensure that the extracted data reflects the physiological state of fully open stomata without inhibition by midday high temperatures. The values ​​within this data segment are then extracted as the stem flow rate and ambient light intensity for the morning time window.

[0076] This stage aims to establish a water response model for crops under conditions without water stress. First, the cotton physiological and ecological monitoring dataset is traversed, and the sunrise time of the day is calculated based on local latitude and longitude, with 1 to 3 hours after sunrise designated as the morning time window. Within this window, data filtering logic is executed: the first derivative of light intensity is calculated, and when the absolute value of the rate of change in light intensity exceeds 50 micromoles per square meter per second (µmol / m² / s), it is identified as an unstable period due to cloud interference and is removed. Simultaneously, an effective starting threshold of 200 µmol / m² / s for photosynthetically active radiation is set, and low-light data below this threshold are removed, generating a valid morning data sequence.

[0077] For each selected set of synchronous data, the stem flow rate was used as the numerator and the ambient light intensity as the denominator, and a division operation was performed to obtain the instantaneous conversion ratio. To eliminate random errors, the interquartile range method was used to identify and remove outliers from the instantaneous conversion ratios. Subsequently, the arithmetic mean of the remaining ratio sequence was performed to generate the morning light flow conversion coefficient. This coefficient characterizes the ability of a unit of light intensity to drive crop transpiration under conditions of low morning temperature, low vapor pressure difference, and relatively sufficient soil moisture.

[0078] The interquartile range method mentioned above is a statistical method that determines the range of outliers by calculating the difference between the upper quartile and the lower quartile of the data. Any data points that fall outside this range are considered outliers and are removed.

[0079] Subsequently, ambient light intensity data during the midday high-radiation period were extracted. Based on crop physiology principles, under the condition of fully adequate water supply, the stem flow rate during midday should maintain the same linear response relationship with light intensity as in the morning. Therefore, each ambient light intensity value during midday was directly multiplied by the morning light flow conversion coefficient to simulate and calculate the theoretically maximum transpiration rate that should be reached at that time, i.e., the theoretical potential stem flow rate.

[0080] Specific example: The stem flow rate at 08:00 AM is 4.25 g / h, corresponding to an ambient light intensity of 445 μmol / m² / s. A division operation is performed to calculate the instantaneous conversion ratio at this time: 4.25 divided by 445, approximately 0.00955. Similarly, the ratio at 08:30 AM is calculated as 6.97 divided by 605, approximately 0.01152. After calculating and removing outliers from 50 sets of data within the morning window, the arithmetic mean is obtained, determining the morning light flow conversion coefficient to be 0.0110. Moving to the noon period, the ambient light intensity at 13:00 is read as 2100 μmol / m² / s. This light intensity is multiplied by the morning light flow conversion coefficient of 0.0110 (2100 multiplied by 0.0110), yielding a theoretical potential stem flow rate of 23.1 g / h.

[0081] Please see Figure 1 and Figure 4 S3: Extract the stem flow rate at noon from the cotton physiological and ecological monitoring dataset as the actual stem flow rate, subtract the actual stem flow rate from the theoretical potential stem flow rate to calculate the instantaneous deficit depth, calculate the deficit width based on the duration during which the actual stem flow rate is continuously lower than the theoretical potential stem flow rate, and select the maximum instantaneous deficit depth to generate the maximum depression depth.

[0082] The generation process of instantaneous defect depth, defect width, and maximum depression depth specifically includes:

[0083] The theoretical potential stem flow rate and the actual stem flow rate at the same time point are differentially calculated. When the difference is positive, the difference is retained as the effective water deficit. When the difference is negative or zero, the deficit is set to zero, thereby generating an instantaneous deficit depth that changes continuously with time.

[0084] Identify continuous time intervals where the instantaneous deficit depth is consistently greater than zero, count the total number of time steps contained within this interval, and multiply this total number by the time interval of a single sampling period to quantify the duration of continuous water stress experienced by the crop and generate the deficit width.

[0085] Within the aforementioned continuous time interval, an extreme value search is performed on the instantaneous deficit depth at all times to screen out the peak difference in the water deficit process, which characterizes the strongest water supply and demand imbalance experienced by the crop during that period and generates the maximum depression depth.

[0086] The process of generating the maximum indentation depth specifically includes:

[0087] A sliding window with a dynamic length is constructed, and this window is used to perform a traversal scan on the time series of instantaneous deficit depth. The fluctuation variance of the data within each window is calculated to eliminate abrupt spikes caused by random noise from the sensor.

[0088] In the defect curve after noise interference is eliminated, the amplitude value corresponding to the inflection point when the slope of the positioning curve changes from positive to negative is extracted. In the case of multiple inflection points, the vertex with the largest absolute value of amplitude is selected by comparison and determined as the maximum depression depth representing the extreme state of this stress event.

[0089] This step focuses on quantifying the gap between actual crop transpiration and theoretical demand. First, the actual stem flow rate at noon is extracted from the dataset and aligned with the theoretical potential stem flow rate generated in step two on the same time axis. For each time step, a difference operation is performed: the actual stem flow rate is subtracted from the theoretical potential stem flow rate. A logical judgment module then intervenes; if the difference is positive, the value is retained as the instantaneous deficit depth, representing the specific magnitude of water shortage; if the difference is negative or zero, the instantaneous deficit depth is forcibly set to zero, thus generating a continuous curve reflecting the degree of water deficit over time.

[0090] To calculate the deficit width, the instantaneous deficit depth sequence is scanned to identify continuous time intervals where the values ​​are consistently greater than zero. The number of data points contained within this interval is counted, and this number is multiplied by the sampling period to calculate the total duration of the crop's water deficit state, thus generating the deficit width.

[0091] To determine the maximum depression depth, data smoothing and extreme value search algorithms are introduced. A sliding window of 5 data points is constructed and slides across the instantaneous deficit depth sequence, calculating the variance of the data within the window. If the variance is less than a preset noise threshold of 0.5, the data is considered stationary; if the variance is too large, the mean of the window is used to replace the center point value to eliminate spurious peaks caused by sensor random noise. In the smoothed curve, the slope change of adjacent data points is calculated to locate the inflection point where the slope changes from positive to negative. These inflection points correspond to local peaks of water deficit. The instantaneous deficit depth values ​​corresponding to all inflection points are compared, and the value with the largest absolute value is defined as the maximum depression depth.

[0092] The sliding window mentioned above is a data processing technique commonly used in time series analysis. By defining a subsequence window of a fixed length and gradually moving the window over the original data, statistics or calculations are performed on the data within the window to achieve data smoothing or feature extraction.

[0093] Specific example: Assume that at 13:00, the theoretical potential stem flow rate is 23.1 g / h, while the actual stem flow rate measured by the sensor is 15.5 g / h. Subtracting 15.5 from 23.1 yields an instantaneous deficit depth of 7.6 g / h at that moment. Monitoring showed that from 12:30 to 14:30, the instantaneous deficit depth was consistently greater than zero. This time period included 13 sampling points with a sampling interval of 10 minutes. Multiplying the number of sampling points (13) by the 10-minute interval, the deficit width was calculated to be 130 minutes. Within this 130-minute period, the instantaneous deficit depths at each sampling point were 2.1 g / h, 3.5 g / h, 5.2 g / h, 7.6 g / h, 8.9 g / h, and 8.4 g / h, respectively. After noise reduction, the value 8.9 was identified as the maximum value in the sequence. Therefore, 8.9 g / h was determined as the maximum depression depth for this water stress event.

[0094] Please see Figure 1 and Figure 5 S4: Calculate the pore closure resistance index by multiplying the maximum indentation depth by the defect width, compare the pore closure resistance index with the benchmark threshold, and construct an opening command when the pore closure resistance index is greater than the benchmark threshold.

[0095] The calculation of the pore closure resistance index and the construction of the opening command specifically include:

[0096] The maximum depression depth and deficit width are obtained, and multiplication is performed to construct a two-dimensional rectangular area model that can reflect the cumulative effect of water deficit intensity and duration. The value of this area model is defined as a quantitative index characterizing the degree of crop stomatal regulation fatigue, and a stomatal closure resistance index is generated.

[0097] The preset irrigation decision benchmark parameter is called. This parameter is set based on the drought resistance characteristics of the cotton at the current growth stage. The stomatal closure resistance index is compared with the benchmark parameter to determine whether the current crop is in an irreversible stomatal closure critical state.

[0098] When the stomatal closure resistance index exceeds the benchmark parameter, the crop is determined to be in a state of water stress that urgently needs water replenishment. A control signal containing the target irrigation amount and execution priority is immediately generated to construct the opening command.

[0099] The process of constructing the start instruction specifically includes:

[0100] Before generating the control signal, the current reading of the soil moisture sensor is read as an auxiliary verification parameter to determine whether the soil moisture content is unsaturated, so as to prevent accidental irrigation caused by the lag in stem flow response when rainfall occurs.

[0101] After confirming that the soil moisture content has not reached the saturation threshold, the pulse duration of the required replenishment water is dynamically calculated based on the proportion of the stomatal closure resistance index exceeding the benchmark threshold. This pulse duration is then encoded into a solenoid valve drive level sequence to generate an opening command.

[0102] During the decision-making and execution phase, irrigation instructions are constructed by comprehensively considering the intensity and duration of water deficit. The maximum depression depth and deficit width generated in step three are obtained, and a multiplication operation is performed to construct the stomatal closure resistance index. This index, in a physical sense, approximates the area under the water deficit curve, comprehensively reflecting the cumulative effect of water stress on crops. Simultaneously, the irrigation decision-making baseline parameters stored in the database are retrieved. These parameters are empirical thresholds set based on cotton drought resistance experiments during the flowering and boll-forming stage. The specific setting process involves: creating different gradients of water stress in the experimental field through artificial water control; recording the resistance index when cotton leaf water potential and stomatal conductance decrease to 50%; and taking 80% of this index as the safety warning line.

[0103] The calculated stomatal closure resistance index is compared with a baseline threshold. If the index is less than the threshold, the system remains in standby mode; if the index is greater than the threshold, the crop is determined to be in urgent need of irrigation. At this time, the soil moisture verification logic is activated, and the volumetric water content of the root zone soil moisture sensor is read. If the water content is higher than 90% of the saturated field capacity, the irrigation command is locked; if the water content is lower than 90%, the lockout is released.

[0104] After the lockout is released, the opening duration of the solenoid valve is calculated using proportional-integral control logic based on the percentage difference between the pore closure resistance index and the benchmark threshold. The excess difference is multiplied by a preset time conversion coefficient to obtain the required irrigation pulse duration, which is then encoded into a high-level signal sequence to generate an opening command that is sent to the field controller.

[0105] The aforementioned proportional-integral control logic refers to a control algorithm that calculates the error between the setpoint and the actual output value, and then linearly combines the proportional and integral terms of this error to adjust the control quantity to eliminate steady-state error.

[0106] Specific example: The maximum depression depth calculated above is 8.9 grams per hour, and the deficit width is 130 minutes, equivalent to approximately 2.17 hours. Performing a multiplication operation, 8.9 multiplied by 2.17, the stomatal closure resistance index is calculated to be approximately 19.31 grams. The preset baseline threshold is set to 15.0. This value comes from a control experiment during the flowering and boll-forming stage; when the cumulative deficit exceeds 15.0 grams, the yield per cotton plant will significantly decrease. Comparing the current index of 19.31 with the threshold of 15.0, it is confirmed that 19.31 is greater than 15.0, triggering an irrigation request. At this time, the soil moisture sensor reading is 65% of field capacity, not reaching the saturation threshold of 90%, allowing irrigation. Calculating the excess: 19.31 minus 15.0 equals 4.31. The time conversion factor is set to 5 minutes of irrigation time per unit of index. Performing a multiplication operation, 4.31 multiplied by 5, the required supplementary irrigation time is 21.55 minutes. A drive signal that remains high for 21.55 minutes is then generated to control the solenoid valve to open.

[0107] Table 2 Results of Irrigation Decision-Making Logical Operations

[0108]

[0109] As shown in Table 2, by quantifying the accumulated physiological deficit into a stomatal closure resistance index and comparing it with the experimentally calibrated threshold, irrigation instructions accurate to the minute were generated.

[0110] A cotton irrigation intelligent control system based on stem flow data, the cotton irrigation intelligent control system based on stem flow data is used to execute the above-mentioned cotton irrigation intelligent control method based on stem flow data, the system includes:

[0111] The data acquisition and processing module is used to acquire thermoelectric potential signals and ambient light intensity through thermal diffusion stem flow sensors and photosynthetically active radiation sensors, calculate stem flow rate based on thermoelectric potential signals, and construct cotton physiological and ecological monitoring dataset by associating stem flow rate, ambient light intensity and acquisition timestamp.

[0112] The potential transpiration prediction module is used to extract stem flow rate and ambient light intensity in the morning time window from cotton physiological and ecological monitoring datasets, calculate the ratio mean to generate the morning light flow conversion coefficient, and multiply the ambient light intensity at noon by the morning light flow conversion coefficient to calculate the theoretical potential stem flow rate.

[0113] The water deficit analysis module is used to extract the stem flow rate at noon from the cotton physiological and ecological monitoring dataset as the actual stem flow rate, calculate the instantaneous deficit depth by subtracting the actual stem flow rate from the theoretical potential stem flow rate, calculate the deficit width based on the duration during which the actual stem flow rate is continuously lower than the theoretical potential stem flow rate, filter the maximum instantaneous deficit depth, and generate the maximum depression depth.

[0114] The irrigation decision control module is used to calculate the pore closure resistance index by multiplying the maximum depression depth by the deficit width, compare the pore closure resistance index with the benchmark threshold, and generate an opening command when the pore closure resistance index is greater than the benchmark threshold.

[0115] The above embodiments illustrate preferred embodiments of the present invention. Any equivalent adjustments to the technical solution based on software engineering methods are within the scope of protection, including but not limited to: implementing algorithm logic using different programming languages, refactoring functional modules into services, adjusting data interaction protocols, and optimizing resource scheduling strategies. Any implementation scheme derived from reasonable modifications to the data processing flow, service call chain, or system architecture layer without departing from the core technology of the present invention should be considered within the protection scope defined by the technical solution of the present invention.

Claims

1. A smart control method for cotton irrigation based on stem flow data, characterized in that, Includes the following steps: S1: Collect thermoelectric potential signals and ambient light intensity using a thermal diffusion stem flow sensor and a photosynthetically active radiation sensor. Calculate stem flow rate based on the thermoelectric potential signals. Construct a cotton physiological and ecological monitoring dataset by associating stem flow rate, ambient light intensity, and collection timestamp. S2: Extract the stem flow rate and the ambient light intensity of the morning time window from the cotton physiological and ecological monitoring dataset, calculate the average ratio to generate the morning light flow conversion coefficient, and multiply the ambient light intensity at noon by the morning light flow conversion coefficient to calculate the theoretical potential stem flow rate. S3: Extract the stem flow rate at noon from the cotton physiological and ecological monitoring dataset as the actual stem flow rate, subtract the actual stem flow rate from the theoretical potential stem flow rate to calculate the instantaneous deficit depth, calculate the deficit width based on the duration during which the actual stem flow rate is continuously lower than the theoretical potential stem flow rate, and filter the maximum instantaneous deficit depth to generate the maximum depression depth. S4: Calculate the pore closure resistance index by multiplying the maximum indentation depth by the defect width, compare the pore closure resistance index with the benchmark threshold, and construct an opening command when the pore closure resistance index is greater than the benchmark threshold.

2. The intelligent control method for cotton irrigation based on stem flow data according to claim 1, characterized in that, The calculation of the stem flow rate and the construction of the cotton physiological and ecological monitoring dataset specifically include: The system receives the raw analog voltage transmitted by the thermal diffusion stem flow sensor, uses an analog-to-digital converter to resolve the analog voltage into a value characterizing the temperature difference between probes, combines it with the pre-calibrated maximum temperature difference reference value under zero flow conditions, calculates the liquid flow density through a nonlinear mapping relationship, and combines it with the water-conducting area of ​​the cotton stem to generate the stem flow rate. The real-time photon flux reading uploaded by the photosynthetically active radiation sensor is acquired synchronously, and the reading is subjected to moving average filtering to eliminate instantaneous high-frequency noise caused by cloud cover, thereby generating the ambient light intensity. A time-series database with the collection timestamp as the unique index is established. The stem flow rate after time-series alignment and the ambient light intensity are written into the corresponding fields of the database and associated with the current crop growth stage identifier to construct the cotton physiological and ecological monitoring dataset.

3. The intelligent control method for cotton irrigation based on stem flow data according to claim 1, characterized in that, The process of generating the morning optical flow conversion coefficient and calculating the theoretical potential stem flow rate specifically includes: The cotton physiological and ecological monitoring dataset is traversed, and a specific time period from sunrise to noon is locked as the morning time window. The stem flow rate and the ambient light intensity within the window are extracted point by point. Invalid data points with light intensity lower than the preset effective photosynthetic starting point are removed to generate an effective morning data sequence. For each group of effective morning data sequences, a division operation is performed between the stem flow rate and the ambient light intensity to obtain a series of instantaneous conversion ratios. An arithmetic mean operation is performed on all instantaneous conversion ratios after removing outliers to establish the morning light flow conversion coefficient that can characterize the photosynthetic driving capacity of crops under water stress. The ambient light intensity during the midday high-radiation period is extracted and directly multiplied by the morning light flow conversion coefficient to simulate the upper limit of the transpiration rate that crops should reach under ideal water supply conditions, thus generating the theoretical potential stem flow rate.

4. The intelligent control method for cotton irrigation based on stem flow data according to claim 1, characterized in that, The process of generating the instantaneous defect depth, defect width, and maximum indentation depth specifically includes: The theoretical potential stem flow rate and the actual stem flow rate at the same timestamp are differentially calculated. When the difference is positive, the difference is retained as the effective water deficit. When the difference is negative or zero, the deficit is set to zero, thereby generating the instantaneous deficit depth that changes continuously with time. Identify the continuous time interval where the instantaneous deficit depth is continuously greater than zero, count the total number of time steps contained in the interval, and multiply the total number by the time interval of a single sampling period to quantify the duration of continuous water stress experienced by the crop and generate the deficit width. Within the aforementioned continuous time interval, an extreme value search is performed on the instantaneous deficit depth at all times to screen out the peak difference in the water deficit process, which characterizes the strongest water supply and demand imbalance experienced by the crop during that period, and generates the maximum depression depth.

5. The intelligent control method for cotton irrigation based on stem flow data according to claim 1, characterized in that, The process of calculating the pore closure resistance index and constructing the opening command specifically includes: The maximum indentation depth and deficit width are obtained, and a multiplication operation is performed to construct a two-dimensional rectangular area model that can reflect the cumulative effect of water deficit intensity and duration. The value of the area model is defined as a quantitative index characterizing the degree of crop stomatal regulation fatigue, and the stomatal closure resistance index is generated. The preset irrigation decision benchmark parameter is invoked. This parameter is set based on the drought resistance characteristics of the cotton at the current growth stage. The stomatal closure resistance index is compared with the benchmark parameter to determine whether the current crop is in an irreversible stomatal closure critical state. When the stomatal closure resistance index exceeds the benchmark parameter, the crop is determined to be in a state of water stress that urgently needs water replenishment. A control signal containing the target irrigation amount and execution priority is immediately generated to construct an opening command.

6. The intelligent control method for cotton irrigation based on stem flow data according to claim 2, characterized in that, The calculation process for the stem flow rate specifically includes: Obtain the maximum temperature difference formed by the heat diffusion probe at night due to the cessation of evaporation, calculate the ratio of the relative difference between the current measured temperature difference and the maximum temperature difference, and use this ratio to construct a dimensionless coefficient reflecting the degree of heat dissipation. An empirical power function model is used to perform a nonlinear transformation on the dimensionless coefficient, calculate the liquid flow velocity per unit sapwood area, and multiply the liquid flow velocity by the area of ​​the sapwood region with water transport function in the cross-section of the cotton stalk to generate the stalk flow rate.

7. The intelligent control method for cotton irrigation based on stem flow data according to claim 3, characterized in that, The extraction process of the stem flow rate and the ambient light intensity during the morning time window specifically includes: Obtain local sunrise time data for the day, set the time period between one and three hours after sunrise as the filter interval, monitor the rate of change of light intensity within this interval, and identify and eliminate unstable periods of intense light fluctuations caused by rapid cloud movement. Filter the continuous data segment within the interval where the stem flow rate increases synchronously with the ambient light intensity and the correlation coefficient is higher than the preset linear threshold. Ensure that the extracted data reflects the physiological state where the stomata are fully open and not inhibited by midday high temperature. Extract the values ​​within this data segment as the stem flow rate and ambient light intensity in the morning time window.

8. The intelligent control method for cotton irrigation based on stem flow data according to claim 4, characterized in that, The process of generating the maximum indentation depth specifically includes: A sliding window of dynamic length is constructed, and the window is used to perform a traversal scan on the time series of the instantaneous deficit depth. The fluctuation variance of the data within each window is calculated to eliminate abrupt spikes caused by sensor random noise. In the defect curve after noise interference is eliminated, the inflection point when the slope of the positioning curve changes from positive to negative is located. The amplitude value corresponding to the inflection point is extracted. In the case of multiple inflection points, the vertex with the largest absolute amplitude value is selected by comparison and determined as the maximum depression depth that characterizes the extreme state of this stress event.

9. The intelligent control method for cotton irrigation based on stem flow data according to claim 5, characterized in that, The process of constructing the start instruction specifically includes: Before generating the control signal, the current reading of the soil moisture sensor is read as an auxiliary verification parameter to determine whether the soil moisture content is unsaturated, so as to prevent accidental irrigation caused by the lag in stem flow response when rainfall occurs. After confirming that the soil moisture content has not reached the saturation threshold, the pulse duration of the required replenishment water is dynamically calculated based on the proportion of the stomatal closure resistance index exceeding the benchmark threshold. This pulse duration is then encoded into a solenoid valve drive level sequence to generate an opening command.

10. A smart control system for cotton irrigation based on stem flow data, characterized in that, The system is used to implement the intelligent control method for cotton irrigation based on stem flow data as described in any one of claims 1-9, the system comprising: The data acquisition and processing module is used to acquire thermoelectric potential signals and ambient light intensity through a thermal diffusion stem flow sensor and a photosynthetically active radiation sensor, calculate stem flow rate based on the thermoelectric potential signal, and construct the cotton physiological and ecological monitoring dataset by associating stem flow rate, ambient light intensity and acquisition timestamp. The potential transpiration prediction module is used to extract the stem flow rate and the ambient light intensity of the morning time window from the cotton physiological and ecological monitoring dataset, calculate the ratio mean to generate the morning light flow conversion coefficient, and multiply the ambient light intensity at noon by the morning light flow conversion coefficient to calculate the theoretical potential stem flow rate. The water deficit analysis module is used to extract the stem flow rate at noon from the cotton physiological and ecological monitoring dataset as the actual stem flow rate, subtract the actual stem flow rate from the theoretical potential stem flow rate to calculate the instantaneous deficit depth, calculate the deficit width based on the duration during which the actual stem flow rate is continuously lower than the theoretical potential stem flow rate, filter the maximum instantaneous deficit depth, and generate the maximum depression depth. The irrigation decision control module is used to calculate the pore closure resistance index by multiplying the maximum depression depth by the deficit width, compare the pore closure resistance index with a benchmark threshold, and generate an opening command when the pore closure resistance index is greater than the benchmark threshold.

Images