Farmland environment multi-parameter intelligent monitoring system and method based on Internet of Things

Abstract

The invention discloses a farmland environment multi-parameter intelligent monitoring system and method based on Internet of Things, particularly relates to the technical field of agricultural Internet of Things monitoring, and is used for solving the problem of stress resistance evaluation distortion caused by neglecting crop physiological recovery hysteresis in the prior art. Collecting soil water content data in real time through an Internet of Things sensor, and extracting a soil water content change rate when a preset threshold value is reached; secondly, constructing a space-time gradient matrix based on different soil layer water deficit rates, and generating a stress cumulative intensity index through principal component analysis and information entropy fusion; then analyzing a coupling relationship between the soil moisture change rate and the stress accumulation effect in the historical stress event on physiological recovery delay to generate a recovery delay coefficient; the physiological state lag degree is evaluated by combining the stomatal conductance change rate ratio of the stress period and the recovery period and the recovery delay coefficient; and finally, dynamically prolonging the irrigation execution time according to the lag degree until the irrigation execution time is lower than a set tolerance value, thereby remarkably improving the reliability of crop stress resistance test data.

Description

Technical Field

[0001] The present invention relates to the field of agricultural Internet of Things monitoring technology, and more specifically, to an Internet of Things-based farmland environment multi-parameter intelligent monitoring system and method. Background Art

[0002] In the field of agricultural planting technology research and development, IoT systems are widely used for farmland environmental monitoring. Existing technologies deploy multiple sensors (such as soil moisture, temperature, and light intensity) to collect environmental parameters in real time and trigger irrigation or alarm decisions based on preset thresholds. Such systems provide data support for environmental regulation in R&D scenarios such as crop stress resistance testing and water-fertilizer efficiency research.

[0003] However, although existing monitoring methods can perceive real-time changes in environmental parameters, they ignore the nonlinear lag effect between crop physiological responses and changes in environmental parameters. When crops encounter stress (such as drought), the recovery cycle of their physiological functions lags significantly behind the improvement of environmental parameters; and the existing system terminates regulatory decisions only when environmental parameters meet the standards, resulting in misjudgment of variety resistance during the research and development process, distorted test conclusions and waste of research and development resources. Summary of the Invention

[0004] In order to overcome the above-mentioned defects of the prior art, the present invention provides a multi-parameter intelligent monitoring system and method for farmland environment based on the Internet of Things to solve the problems raised in the above-mentioned background technology.

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

[0006] The multi-parameter intelligent monitoring method of farmland environment based on the Internet of Things includes:

[0007] S1. Real-time collection of farmland environmental parameters through the IoT sensor network, including soil moisture content and soil water content data;

[0008] S2. When the soil moisture content reaches a preset threshold, obtaining the soil moisture content change rate within a set time window;

[0009] S3. Based on the soil moisture data at different soil depths within a set time window, the soil moisture deficit rate at each depth layer is calculated and a spatiotemporal gradient matrix is ​​constructed. The spatiotemporal cumulative effect of stress is quantified by performing principal component analysis on the spatiotemporal gradient matrix and integrating information entropy.

[0010] S4. Analyze the coupling relationship between the soil moisture change rate during historical stress events and the temporal and spatial cumulative effect of stress on crop physiological recovery delay, and generate a recovery delay coefficient;

[0011] S5. Obtain the ratio of the stomatal conductance decrease rate during the stress period to the stomatal conductance recovery rate during the recovery period, and evaluate the degree of hysteresis in the crop physiological state in combination with the recovery delay coefficient;

[0012] S6. Extend the irrigation decision execution time according to the lag degree of the crop physiological state until the lag degree is lower than the set tolerance value.

[0013] Furthermore, soil moisture data among farmland environmental parameters is obtained through multi-layer soil moisture sensors deployed in the crop root distribution layer. The multi-layer soil moisture sensors synchronously collect soil volumetric water content at least three typical soil layer depths.

[0014] Typical soil depths include shallow root zone, middle root zone and deep root zone;

[0015] The frequency of soil volumetric water content acquisition matches the cycle of crop transpiration rate changes, and the corresponding timestamp and sensor location coordinates are recorded synchronously for each acquisition;

[0016] The IoT sensor network adopts a star topology and transmits soil volumetric moisture content data to edge computing nodes in real time through a multi-hop routing protocol.

[0017] Furthermore, the preset threshold is a critical value of soil volumetric water content corresponding to the permanent wilting coefficient of the crop;

[0018] The start time of the time window is set to the moment when the soil volume water content first falls below the preset threshold, and the end time is the current real-time collection time;

[0019] The soil moisture change rate is calculated by linear regression fitting of all soil volumetric moisture data points within a set time window, where the independent variable is the time increment from the starting moment and the dependent variable is the measured value of the soil volumetric moisture content at the corresponding moment;

[0020] The slope value of the linear regression fitting is the rate of change of soil moisture content, and its unit is expressed in percentage of daily moisture content change.

[0021] Furthermore, soil moisture data at different soil depths refer to the soil volumetric water content in the shallow root zone, middle root zone, and deep root zone;

[0022] The soil water deficit rate is calculated by taking the first-order differential derivative of the soil volume water content series at each depth layer within the set time window to obtain the instantaneous deficit rate at each sampling moment;

[0023] The row vectors of the spatiotemporal gradient matrix correspond to the sampling moments within the set time window, and the column vectors correspond to the instantaneous deficit rate values ​​of the shallow root zone, the middle root zone, and the deep root zone.

[0024] The principal component analysis takes the spatiotemporal gradient matrix as input and extracts the first principal component eigenvalue and the corresponding eigenvector;

[0025] Information entropy fusion is achieved by calculating the Shannon entropy of the eigenvector components;

[0026] The cumulative stress intensity index was generated by multiplying the first principal component eigenvalue with the Shannon entropy of the eigenvector components.

[0027] Furthermore, historical stress events refer to crop growth cycles in which the soil volumetric water content has fallen below a preset threshold;

[0028] The soil moisture change rate was calculated by linear regression fitting during historical stress events.

[0029] The temporal and spatial cumulative effect of coercion adopts the coercion cumulative intensity index value generated from historical coercion events;

[0030] The delay in crop physiological recovery is quantified by the time required for stomatal conductance to return to normal levels after the end of the historical stress event;

[0031] The coupling relationship was established through a multiple linear regression model:

[0032] The soil moisture change rate and stress accumulation intensity index were used as independent variables, and the crop physiological recovery delay time was used as the dependent variable.

[0033] The recovery delay coefficient was generated by fitting the regression coefficients of the multiple linear regression model.

[0034] Furthermore, the stress period refers to the period during which the soil volumetric water content is below a preset threshold, and the recovery period refers to the period after which the soil volumetric water content returns to the preset threshold;

[0035] The stomatal conductance decline rate during the stress period was determined by the absolute value of the slope of the linear fitting of the stomatal conductance monitoring values ​​during the stress period;

[0036] The stomatal conductance recovery rate during the recovery period was determined by the positive slope of the linear fitting of the stomatal conductance monitoring values ​​during the recovery period;

[0037] The ratio was calculated as the quotient of the absolute value of the stomatal conductance decrease rate during the stress period and the stomatal conductance recovery rate during the recovery period;

[0038] The recovery delay coefficient directly refers to the recovery delay coefficient value generated in the historical coercion event;

[0039] The hysteresis degree of crop physiological state is generated by multiplying the recovery delay coefficient value and the ratio value.

[0040] Furthermore, the tolerance value is set as a critical value where the hysteresis degree of the crop physiological state is equal to 1;

[0041] The specific ways to extend the execution time of irrigation decisions are:

[0042] Get the basic execution time of the current irrigation decision;

[0043] Calculate the absolute value of the difference between the hysteresis degree of the crop physiological state and the set tolerance value;

[0044] Multiply the absolute value of the difference by the basic execution time to obtain the irrigation extension compensation time;

[0045] Add the basic execution time and the irrigation extension compensation time to get the actual total irrigation execution time;

[0046] Continuously monitor the hysteresis of crop physiological status and terminate irrigation when the hysteresis of crop physiological status falls below the set tolerance value.

[0047] In another aspect, the present invention provides a multi-parameter intelligent monitoring system for farmland environment based on the Internet of Things, comprising:

[0048] Environmental collection module, used to collect farmland environmental parameters in real time through the IoT sensor network. Farmland environmental parameters include soil moisture content and soil water content data;

[0049] The rate analysis module is used to obtain the soil moisture content change rate within a set time window when the soil moisture content reaches a preset threshold;

[0050] The stress quantification module is used to calculate the soil moisture deficit rate at each depth layer based on soil moisture data at different soil depths within a set time window and construct a spatiotemporal gradient matrix. The spatiotemporal cumulative effect of stress is quantified by performing principal component analysis on the spatiotemporal gradient matrix and integrating information entropy.

[0051] The delay analysis module is used to analyze the coupling relationship between the soil moisture change rate in historical stress events and the temporal and spatial cumulative effect of stress on the delay of crop physiological recovery, and generate the recovery delay coefficient;

[0052] The hysteresis evaluation module is used to obtain the ratio of the stomatal conductance decline rate during the stress period to the stomatal conductance recovery rate during the recovery period, and to evaluate the degree of hysteresis in the crop physiological state in combination with the recovery delay coefficient;

[0053] The decision-making and control module is used to extend the execution time of irrigation decisions according to the lag degree of crop physiological status until the lag degree is lower than the set tolerance value.

[0054] Compared with the prior art, the present invention has the following beneficial effects:

[0055] 1. By dynamically quantifying the coupling relationship between the spatiotemporal cumulative effects of stress events and the delay in physiological recovery, the accuracy of capturing the true state of crops in stress resistance experiments is significantly improved. Based on the spatiotemporal gradient matrix constructed based on the water deficit rate of different soil layers, the stress cumulative intensity index is generated through principal component analysis and information entropy fusion, breaking through the blind spot of traditional single-point threshold monitoring in identifying continuous stress; further analyzing the impact of the synergistic effect of soil moisture change rate and stress cumulative effect on physiological recovery delay in historical stress, the generated recovery delay coefficient provides a quantitative basis for the hysteresis of crop physiological response. This mechanism of integrating real-time stress intensity with historical law analysis fundamentally avoids the problem of misjudging the physiological recovery state due to short-term improvements in environmental parameters.

[0056] 2. At the decision-making and execution level, the degree of physiological imbalance is captured in real time through the ratio of the stomatal conductance change rate between the stress period and the recovery period. The degree of lag in the crop's physiological state is evaluated in combination with the recovery delay coefficient, forming a precise mapping of environmental parameters to physiological responses. A compensation mechanism that dynamically extends the irrigation time based on the degree of lag ensures continuous water supply after the soil moisture content reaches the standard until the physiological function is fully restored. This effectively eliminates the bias in stress resistance evaluation caused by the lag in physiological recovery, and ensures that the stress response data recorded in the variety test truly reflects the intrinsic characteristics of the crop, thereby ensuring the accuracy of the R&D conclusions while reducing the resource loss caused by repeated experiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] Figure 1 This is a flow chart of the multi-parameter intelligent monitoring method for farmland environment based on the Internet of Things of the present invention;

[0058] Figure 2 This is a structural diagram of the multi-parameter intelligent monitoring system for farmland environment based on the Internet of Things of the present invention. DETAILED DESCRIPTION

[0059] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0060] Example 1: Figure 1 The present invention provides a multi-parameter intelligent monitoring method for farmland environment based on the Internet of Things, including:

[0061] S1. Real-time collection of farmland environmental parameters through the IoT sensor network, including soil moisture content and soil water content data;

[0062] S2. When the soil moisture content reaches a preset threshold, obtaining the soil moisture content change rate within a set time window;

[0063] S3. Based on the soil moisture data at different soil depths within a set time window, the soil moisture deficit rate at each depth layer is calculated and a spatiotemporal gradient matrix is ​​constructed. The spatiotemporal cumulative effect of stress is quantified by performing principal component analysis on the spatiotemporal gradient matrix and integrating information entropy.

[0064] S4. Analyze the coupling relationship between the soil moisture change rate during historical stress events and the temporal and spatial cumulative effect of stress on crop physiological recovery delay, and generate a recovery delay coefficient;

[0065] S5. Obtain the ratio of the stomatal conductance decrease rate during the stress period to the stomatal conductance recovery rate during the recovery period, and evaluate the degree of hysteresis in the crop physiological state in combination with the recovery delay coefficient;

[0066] S6. Extend the irrigation decision execution time according to the lag degree of the crop physiological state until the lag degree is lower than the set tolerance value.

[0067] Soil moisture data in farmland environmental parameters is obtained through multi-layer soil moisture sensors deployed in the crop root distribution layer. The crop root distribution layer refers to the soil depth range where the main water-absorbing roots of the crop are located. This range is determined according to the root growth characteristics of the target crop. For example, the root distribution layer of corn crops is usually from the surface to a depth of 60 cm below the surface. The multi-layer soil moisture sensor consists of at least three independent detection units. Each detection unit is fixed at a specific soil depth. It measures the soil dielectric constant through the frequency domain reflection principle and converts it into a moisture content value. The physical structure of the detection unit is a waterproof probe design. The probe length is customized according to the depth of the target soil layer. It is kept vertically inserted into the soil during installation to ensure that the surface of the detection unit is in full contact with the soil particles.

[0068] The multi-layer soil moisture sensor synchronously collects soil volumetric moisture content at at least three typical soil depths. Typical soil depths include the shallow root zone, the middle root zone, and the deep root zone. The shallow root zone corresponds to a depth from the surface to 20 cm below the surface, the middle root zone to a depth from 20 cm to 40 cm below the surface, and the deep root zone to a depth from 40 cm to 60 cm below the surface. A soil moisture detection unit is deployed at each typical soil depth, integrated into the same sensor body. Synchronous acquisition means that the three detection units trigger measurements at the same time. The measured soil volumetric moisture content is the ratio of liquid water to a unit volume of soil, measured in cubic centimeters per cubic centimeter. During the measurement process, the detection units transmit 125 MHz high-frequency electromagnetic waves and receive reflected signals. The volumetric moisture content value is output based on a pre-calibrated dielectric constant-volume moisture content relationship curve. The calibration curve is determined based on the actual soil texture using the drying and weighing method.

[0069] The frequency of soil volumetric water content collection matches the crop transpiration rate cycle. This cycle is characterized by alternating periods of increased daytime transpiration and decreased nighttime transpiration, and this cycle is positively correlated with changes in light intensity. The dynamic adjustment rule for the collection frequency is: when the daytime light intensity exceeds 20,000 lux for 10 consecutive minutes, the collection frequency is set to once every 5 minutes; when the light intensity is below 20,000 lux for 10 consecutive minutes, the collection frequency is set to once every 30 minutes. This matching mechanism is achieved by the sensor's built-in photosensor to monitor ambient light intensity in real time. For example, in corn-growing areas, high-frequency collection is typically performed every 5 minutes from 2 hours after sunrise to 2 hours before sunset.

[0070] Each acquisition simultaneously records the corresponding timestamp and sensor location coordinates. The timestamp record format includes year, month, day, hour, minute, and second information, for example, 14:25:30 on July 15, 2023. The time base is derived from the sensor's built-in real-time clock chip, which is automatically calibrated weekly using the IoT network time protocol. The sensor's location coordinates include three-dimensional data: longitude, latitude, and altitude. Longitude and latitude are obtained through the built-in Global Positioning System module, while altitude is converted from atmospheric pressure sensor measurements. A complete data record generated by a single acquisition consists of seven fields: timestamp, longitude, latitude, altitude, shallow root zone volumetric water content, mid-root zone volumetric water content, and deep root zone volumetric water content.

[0071] The IoT sensor network utilizes a star topology. In this structure, all multi-layer soil moisture sensor nodes communicate directly with a central edge computing node. The central edge computing node is deployed in the center of the farmland area, and the sensor nodes are distributed within a circular area with a radius of no more than 500 meters centered on the edge computing node. Each sensor node is assigned a unique network identification code and establishes a point-to-point communication link with the edge computing node via a 2.4 GHz wireless signal. Data transmission utilizes a physical layer frame structure, with the data packet payload carrying the seven fields previously described.

[0072] Soil volumetric moisture data is transmitted to edge computing nodes in real time via a multi-hop routing protocol. This multi-hop routing protocol addresses signal attenuation caused by physical obstacles. When the wireless signal strength between the sensor node and the edge computing node falls below -85 decibel milliwatts, a neighboring node with a signal strength greater than -85 decibel milliwatts is automatically selected as a relay. The dynamic routing path generation rules are as follows: each node periodically broadcasts the minimum number of hops required to reach the edge computing node; the minimum hop path is selected for data packet transmission, with a maximum hop limit of three. Real-time transmission requires that the total delay from data acquisition completion to delivery to the edge computing node does not exceed five seconds. Upon receiving the data, the edge computing node performs data verification by calculating the cyclic redundancy check code of the data packet and comparing it with the transmission check code. Once verified, the data is stored in a local structured database.

[0073] When the soil moisture content reaches a preset threshold, the rate of change of soil moisture content within the set time window is obtained. The preset threshold is the critical soil volumetric moisture content corresponding to the crop's permanent wilting coefficient. This threshold is determined through a potted wilting test on the target crop. Specifically, water supply is cut off during the crop's flowering period. When the leaves fail to regain their upright position under a saturated vapor pressure difference of less than 1 kPa at night, the soil volumetric moisture content in the root zone is measured. For example, the permanent wilting coefficient of corn is typically 10% to 12% (unit: cubic centimeters per cubic centimeter), and this value is preset as a fixed parameter.

[0074] The start time of the time window is set to the moment when the soil volumetric moisture content first falls below the preset threshold. This first occurrence requires that the measured value be below the threshold for two consecutive collection cycles and above the threshold in the previous cycle. The start time is the timestamp of the data collection when the value first falls below the threshold. For example, in a monitoring session, the soil volumetric moisture content in the shallow root zone collected at 2:20 PM on July 15, 2023, was 12.1% (preset threshold 12%), and the value collected at 2:25 PM was 11.9%. Therefore, the start time is determined to be 2:25:00 PM on July 15, 2023.

[0075] Set the end time of the time window to the current real-time collection time. The current real-time collection time refers to the timestamp of the most recent data packet received by the edge computing node. The time window duration is dynamically updated with each new data collection. The minimum is a single collection cycle (for example, 5 minutes) and the maximum is no more than 72 hours. For example, if the third collection after the start time completes at 14:35:00, the end time is updated to that time.

[0076] The soil moisture change rate is calculated by linear regression fitting all soil volumetric moisture data points within a set time window. The input data includes all valid data points within the time window, each of which consists of two parameters: the independent variable is the time increment from the start time (in minutes), calculated by subtracting the start time timestamp from the data acquisition timestamp and converting it to minutes; the dependent variable is the measured soil volumetric moisture content at the corresponding time (in cubic centimeters per cubic centimeter). Data preprocessing requires removing outliers caused by sensor failures. The exclusion rule is that the moisture content of adjacent data points fluctuates by more than 5%.

[0077] Linear regression fitting is implemented using the least squares method. The specific steps are: first, the arithmetic mean of all time increment values ​​and the arithmetic mean of all measured soil volumetric moisture values ​​are calculated; second, the covariance between the time increment series and the soil volumetric moisture series, as well as the variance of the time increment series, are calculated; finally, the slope parameter is obtained by dividing the covariance by the variance. This calculation process is performed by the numerical operation library of the edge computing node, and the fitting is repeated after each new data point.

[0078] The slope of the linear regression fit is the original value of the soil moisture change rate. This value physically represents the change in soil volumetric moisture content per unit time. A positive value indicates an increase in moisture content, while a negative value indicates a decrease. For example, a slope of -0.0015 indicates a decrease in moisture content of 0.001 cubic centimeters per minute.

[0079] The units of the soil moisture change rate are converted to the daily percentage change in moisture content. This conversion is performed in two steps: first, multiply the original value by 1440 (the daily minute constant) to obtain the absolute value of the daily change. Second, divide this absolute value by the measured soil volumetric moisture content at the starting time, and then multiply by 100 to convert it to a percentage. For example, if the soil moisture content at the starting time is 12% and the original slope is -0.0015, the daily change = -0.0015 × 1440 = -2.16, and the percentage change = (-2.16 / 12) × 100 = -18.0% / day. The final output value is signed and has the unit "% / day."

[0080] Based on soil moisture data at different soil depths within a set time window, the soil moisture deficit rate at each depth layer is calculated and a spatiotemporal gradient matrix is ​​constructed. Soil moisture data at different soil depths refers to the volumetric soil moisture content of the shallow, mid, and deep root zones. This data is collected synchronously by an IoT sensor network within the set time window. The start and end times of the set time window are defined consistent with those in step S2. For example, if the start time is 14:25:00 on July 15, 2023, and the end time is 14:35:00 on the same day, the time window includes three collection times (14:25:00, 14:30:00, and 14:35:00). The measured volumetric soil moisture content of the three layers is obtained at each collection time, in cubic centimeters per cubic centimeter.

[0081] The soil water deficit rate is calculated by taking the first-order derivative of the soil volumetric water content series for each depth layer within a specified time window. The specific steps include: first, arranging the measured soil volumetric water content series for each depth layer in chronological order to form a time series dataset; second, subtracting the measured soil volumetric water content values ​​at two adjacent sampling times to obtain the change in water content; and third, dividing this change by the number of minutes between the two sampling times to obtain the instantaneous deficit rate at each sampling time. A negative instantaneous deficit rate indicates a decrease in water content. For example, if the soil volumetric water content in the shallow root zone is 11.9 cubic centimeters per cubic centimeter at 2:25:00 PM and 11.7 cubic centimeters per cubic centimeter at 2:30:00 PM, with a time interval of 5 minutes, the instantaneous deficit rate is (11.7 - 11.9) / 5 = -0.04. Boundary condition handling rules are: single-sided differencing is used at the beginning and end of the series, and central differencing is used at the intermediate time.

[0082] When constructing the spatiotemporal gradient matrix, the row vectors correspond to the sampling moments within the specified time window, and the column vectors correspond to the instantaneous deficit rates of the shallow, middle, and deep root zones. The matrix dimensions are determined by the number of valid sampling times within the time window. For example, if sampling occurs every 5 minutes within a 10-minute window, a 3-row, 3-column matrix is ​​generated. The matrix elements are arranged as follows: the first column of the i-th row (i is the row number) stores the instantaneous deficit rate of the shallow root zone at the i-th sampling moment, the second column of the i-th row stores the instantaneous deficit rate of the middle root zone, and the third column of the i-th row stores the instantaneous deficit rate of the deep root zone. For example, if the matrix row vectors at a given moment are [-0.04, -0.03, -0.01], this indicates that the deficit rates of the shallow, middle, and deep root zones decreased by 0.04, 0.03, and 0.01 cubic centimeters per minute, respectively.

[0083] Principal component analysis (PCA) takes the spatiotemporal gradient matrix as input and extracts the eigenvalue and corresponding eigenvector of the first principal component. PCA performs five core operations in sequence: First, each column of the matrix is ​​normalized using the formula (original value - column mean) / column standard deviation, so that each column has a mean of 0 and a standard deviation of 1. Second, the 3×3 covariance matrix of the normalized matrix is ​​calculated. The covariance matrix is ​​then subjected to eigenvalue decomposition using the Jacobi iterative method, with the iteration terminating when the absolute value of all off-diagonal elements is less than 0.0001. The eigenvalues ​​are then sorted in descending order. Finally, the largest eigenvalue is selected as the eigenvalue of the first principal component, and its corresponding unit eigenvector is used as the first principal component eigenvector. The eigenvector consists of three components, representing the weighted contributions of the deficit rates of the shallow, mid, and deep root zones to the principal component. For example, the eigenvector [0.62, 0.55, 0.56] calculated in one case indicates a weight of 0.62 for the shallow root zone, 0.55 for the mid root zone, and 0.56 for the deep root zone.

[0084] Information entropy fusion is achieved by calculating the Shannon entropy of the eigenvector components. Shannon entropy calculation involves three stages: First, each eigenvector component is squared to obtain a non-negative value; second, the squared values ​​are normalized so that the sum of the three squares equals 1. The normalization formula is: squared value of the component / sum of the squares of the total components; third, the normalized values ​​are substituted into the Shannon entropy formula to calculate the entropy value: the sum of the negative normalized values ​​multiplied by the natural logarithm of each value. The entropy value physically represents the uniformity of the vertical distribution of water stress. Entropy values ​​range from 0 to 1.0986 (maximum entropy for a three-dimensional system), with higher entropy values ​​indicating more uniform stress distribution. For example, for the normalized values ​​[0.38, 0.30, 0.32], the entropy value is -(0.38×ln0.38+0.30×ln0.30+0.32×ln0.32)=1.08.

[0085] The cumulative stress intensity index is generated by multiplying the eigenvalue of the first principal component by the Shannon entropy of the eigenvector components. The eigenvalue of the first principal component reflects the overall intensity of water deficit and has a dimension of 1; the eigenvector entropy reflects the spatial heterogeneity of water deficit and has a dimension of 1. The product of the two constitutes the dimensionless cumulative stress intensity index. A larger index value indicates more severe cumulative water stress on the crop. The index is calculated at the edge computing node, and the output result is rounded to two decimal places. For example, if the eigenvalue is 3.85 and the entropy is 0.92, the cumulative stress intensity index = 3.85 × 0.92 = 3.54. The exception handling mechanism is as follows: when the eigenvalue is less than 0.1 or the entropy is less than 0.01, the data is considered an anomaly, and the error code 99 is returned instead of the index value.

[0086] Step S3 captures the dynamic correlation between water deficit rates in different soil layers by constructing a spatiotemporal gradient matrix, extracts the core stress pattern in vertical space using principal component analysis, and then integrates information entropy to quantify the uniformity of stress distribution. Compared with traditional methods that only analyze a single soil layer or simple average, this design can simultaneously reflect the intensity and spatial heterogeneity of water stress, and more accurately characterize the cumulative stress effects of the crop root layer. Converting the spatiotemporal evolution characteristics of soil moisture dynamics into a quantifiable stress index solves the technical problem that existing technologies cannot effectively evaluate the accumulation of multi-level stress.

[0087] The coupling relationship between the rate of change in soil moisture content during historical stress events and the temporal and spatial cumulative effects of stress on crop physiological recovery delay is analyzed to generate a recovery delay coefficient. A historical stress event refers to a crop growth cycle in which the volumetric soil moisture content fell below a preset threshold. The preset threshold is the critical volumetric soil moisture value corresponding to the crop permanent wilting coefficient defined in step S2. Screening for historical stress events requires two conditions: first, three consecutive soil moisture values ​​collected over the entire growth cycle of the target crop must be below the preset threshold; second, the interval between each collection must not exceed 24 hours. For example, given a 120-day corn growth cycle and a preset threshold of 12% (unit: cubic centimeters per cubic centimeter), a valid historical stress event is identified when the values ​​collected for three consecutive days from July 15 to 17, 2023, at a 20-cm soil depth, are 11.8%, 11.6%, and 11.5%, respectively. Event data is extracted from the historical monitoring database.

[0088] The soil moisture change rate is calculated using linear regression fitting during historical stress events. This value directly references the calculation result from step S2 within the corresponding historical event time window. The time window starts at the first sampling time below the threshold and ends at the end of the event, which is defined as the time when the soil volumetric moisture content rises above the preset threshold and remains above it for two consecutive sampling cycles. The calculation process is exactly the same as step S2, and the output value is expressed as the daily percentage change in moisture content (% / day). For example, for a historical event time window of 48 hours, the linear regression fitting results in a soil moisture change rate of -18.0% / day.

[0089] The spatiotemporal cumulative stress effect uses the cumulative stress intensity index value generated during historical stress events. This value refers to the calculation result of step S3 within the same historical event time window. The calculation process is exactly the same as step S3 and includes three consecutive operations: calculating the instantaneous deficit rate based on the soil volumetric water content series of the shallow, mid-root zones, and deep root zones; constructing a spatiotemporal gradient matrix; and generating a dimensionless index by fusing information entropy through principal component analysis. For example, the cumulative stress intensity index value calculated for one event was 3.54.

[0090] Delayed crop physiological recovery is quantified by the time it takes for stomatal conductance to return to normal levels after a historical stress event. Stomatal conductance is measured using a portable photosynthesis meter at 12:00 noon daily on a fixed position on a functional leaf in the middle of the crop canopy. Three measurements are repeated and the arithmetic mean is taken. Normal levels are defined as the 90th percentile of stomatal conductance values ​​for the crop variety under non-stress conditions during the same period. This quantile is based on statistical data from healthy plants measured at 12:00 noon daily over the past three years. Recovery time, measured in days, begins the day after the stress event and ends when stomatal conductance values ​​reach or exceed normal levels for three consecutive days. For example, if stomatal conductance values ​​on the 5th, 6th, and 7th day after the event are 25, 27, and 29 mmol / m2 / s, respectively (normal level: 30 mmol / m2 / s), the recovery time is determined to be 7 days.

[0091] The coupling relationship was established using a multivariate linear regression model. Model construction involved four sequential steps: The first step was to prepare the dataset, collecting complete data from at least 30 valid historical stress events. Each event included the soil moisture change rate (independent variable X1, unit: % / day), the cumulative stress intensity index (independent variable X2, dimensionless), and the recovery time (dependent variable Y, unit: days). The second step was to establish a mathematical model, using the equation Y = constant term + X1 × coefficient β1 + X2 × coefficient β2 + error term. The third step was to solve the parameters, using the least squares method to estimate the coefficients. The optimal solution for β1 and β2 was obtained by solving a linear equation system involving the sample mean and the deviation matrix. The fourth step was to validate the model, requiring a goodness-of-fit R² ≥ 0.6 and a significant p-value < 0.05 for all independent variables. If this was not achieved, the historical event sample size was expanded to 50 and the model was refitted.

[0092] The recovery delay coefficient is generated by fitting the regression coefficients of a multivariate linear regression model. The regression coefficients β1 and β2 form the recovery delay coefficient vector [β1, β2]. Their physical meaning is: β1 represents the number of days of recovery delay contributed by a 1% / day increase in the rate of change of soil moisture content, and β2 represents the number of days of recovery delay contributed by a 1-unit increase in the cumulative intensity index of stress. Coefficient values ​​are rounded to three decimal places. For example, the fitting results are β1 = -0.152 days per percentage per day, and β2 = 1.237 days per index unit. The recovery delay coefficient is stored in the parameter database of the irrigation decision-making system and used to predict recovery delay for real-time stress events.

[0093] The ratio of the stomatal conductance decline rate during the stress period to the stomatal conductance recovery rate during the recovery period is calculated and combined with the recovery delay coefficient to assess the degree of lag in crop physiological status. The stress period refers to the period of time during which the soil volumetric moisture content remains below a preset threshold. The preset threshold is the critical soil volumetric moisture content value corresponding to the crop permanent wilting coefficient defined in step S2. The stress period starts at the first sampling time when the soil volumetric moisture content falls below the preset threshold, and ends at the last sampling time when it falls below the threshold. The recovery period starts at the first sampling time when the soil volumetric moisture content rises above the preset threshold, and ends at the current monitoring time, no earlier than 24 hours after the start of the recovery period. For example, if the preset threshold is 12%, when the soil volume moisture content is 11.9% at 14:25:00 on July 15 (lower than the preset threshold) and 11.8% at 09:00:00 on July 17 (still lower than the preset threshold), and the moisture content is 12.2% at 14:00:00 on July 17 (rising above the preset threshold), the stress period is from 14:25:00 on July 15 to 09:00:00 on July 17, and the recovery period starts at 14:00:00 on July 17.

[0094] The rate of decline in stomatal conductance during the stress period was determined by the absolute value of the slope of a linear fit of the monitored stomatal conductance values ​​during the stress period. Three steps were performed: First, stomatal conductance of functional leaves in the middle of the crop canopy was measured using a portable photosynthesis meter at 12:00:00 noon daily during the stress period. Each leaf was measured three times and the arithmetic mean was calculated, with units of millimolar square meters seconds. Second, each acquisition time was converted to days relative to the start of the stress period (accurate to 0.01 day) to form a time-stomatal conductance series. Third, a linear regression fit was performed using the least squares method, and the absolute value of the slope of the fitted line was taken as the rate of decline, in millimolar square meters seconds per day. For example, if the stomatal conductance series measured over five days of stress is [300, 280, 250, 230, 210] millimolar square meters seconds, and the fitted slope is -22.5 millimolar square meters seconds per day, then the rate of decline is 22.5 millimolar square meters seconds per day.

[0095] The recovery rate of stomatal conductance during the recovery period was determined by the positive slope of a linear fit of the stomatal conductance measurements during the recovery period. The following procedure was used: stomatal conductance was measured daily at 12:00:00 PM using the same protocol; the acquisition time was converted to days relative to the start of the recovery period; and after performing a linear regression fit, the positive slope was directly taken as the recovery rate, expressed in millimolar square meters seconds per day. For example, if the stomatal conductance values ​​measured on day 3 of the recovery period were [210, 230, 250] millimolar square meters seconds, and the fitted slope was +20.0 millimolar square meters seconds per day, the recovery rate would be 20.0 millimolar square meters seconds per day. The boundary condition was that if the fitted slope ≤ 0, the data was considered invalid and remeasured.

[0096] The ratio is calculated as the quotient of the absolute value of the stomatal conductance decline rate during the stress period and the stomatal conductance recovery rate during the recovery period. The calculation formula is: Ratio = Decline Rate / Recovery Rate. Before calculation, verify that the recovery rate is > 0.1 mmol / m2 / s / day; otherwise, the calculation terminates and returns an error code -999. The result is dimensionless and rounded to three decimal places. For example, if the decline rate is 22.5 mmol / m2 / s / day and the recovery rate is 20.0 mmol / m2 / s / day, the ratio is 22.5 / 20.0 = 1.125.

[0097] The recovery delay coefficient directly references the recovery delay coefficient value generated during historical stress events. This value is retrieved from the parameter database of the irrigation decision system. The retrieval rule is to match the current crop variety name with the historical recovery delay coefficient vector generated in step S4. This coefficient consists of two components: β1, which represents the number of recovery delay days contributed by a 1% / day increase in the soil moisture change rate (unit: days per percentage per day), and β2, which represents the number of recovery delay days contributed by a 1% / day increase in the cumulative stress intensity index (unit: days per index unit). For example, the recovery delay coefficient vector for a corn variety is [β1 = -0.152, β2 = 1.237].

[0098] The hysteresis of crop physiological status is generated by multiplying the recovery delay coefficient by the ratio. The generation rule is: calculate two hysteresis components: the first hysteresis component = β1 × ratio, and the second hysteresis component = β2 × ratio. The calculation results are expressed in days, with two decimal places. Positive values ​​indicate a hysteresis effect, while negative values ​​indicate accelerated recovery. For example, if the ratio is 1.125, β1 = -0.152, and β2 = 1.237, then the first hysteresis component = -0.152 × 1.125 ≈ -0.17 days, and the second hysteresis component = 1.237 × 1.125 ≈ 1.39 days. The final output hysteresis vector [-0.17, 1.39] days is stored in the crop physiological status database for irrigation decision-making.

[0099] Based on the hysteresis of the crop's physiological state, the duration of irrigation decisions is extended until the hysteresis falls below a set tolerance. This tolerance is determined by controlling the soil moisture gradient in physiological experiments on target crop varieties and measuring the relationship between stomatal conductance recovery rate and biomass accumulation. When the hysteresis corresponding to the recovery efficiency inflection point stabilizes at 1, this value is set as the tolerance. For example, in experiments with the corn variety Zhengdan 958, the leaf photosynthetic rate recovered to 95% of its normal level when the hysteresis reached 1.

[0100] Obtain the basic execution time for the current irrigation decision. The basic execution time is derived from the initial calculation results of the irrigation decision system, which are calculated using the target crop's daily water requirement data, the irrigation system's designed flow parameters, and the irrigation efficiency coefficient. The specific calculation rule is: the basic execution time (minutes) equals the crop's daily water requirement (mm) divided by the irrigation system flow rate (mm / min), then divided by the irrigation efficiency coefficient. The crop's daily water requirement is calculated by multiplying the reference crop evapotranspiration (measured by the meteorological station) by the crop coefficient. The irrigation system flow parameters are obtained from the irrigation equipment's technical manual. The irrigation efficiency coefficient ranges from 0.75 to 0.90 and is determined based on the dripper type and pipe network pressure. For example, if corn's daily water requirement is 6 mm, the irrigation system flow rate is 0.1 mm / min, the dripper type is pressure-compensated, and the pipe network pressure is 0.3 MPa, then an efficiency coefficient of 0.85 is selected. The basic execution time is 6 / (0.1 × 0.85) ≈ 70.6 minutes.

[0101] Calculate the absolute value of the difference between the hysteresis of the crop physiological state and the set tolerance. The hysteresis of the crop physiological state is derived from the hysteresis vector output by step S5. This vector consists of two components: the first hysteresis component represents the number of hysteresis days due to the rate of change of water content, and the second hysteresis component represents the number of hysteresis days due to the cumulative effects of stress. The arithmetic mean of these two components is taken as the composite hysteresis value. The absolute value of the difference is calculated as: |comprehensive hysteresis value - 1|. For example, if the hysteresis vector output by step S5 is negative 0.17 days and 1.39 days (e.g., [-0.17, 1.39] days), the composite hysteresis value = (-0.17 + 1.39) / 2 = 0.61, and the absolute value of the difference = |0.61 - 1| = 0.39.

[0102] Multiply the absolute value of the difference by the base duration to calculate the extended irrigation compensation time. The calculation rule is: Extended irrigation compensation time (minutes) equals the absolute value of the difference multiplied by the base duration (minutes). Round the result up to the nearest whole minute. For example, if the base duration is 70.6 minutes and the absolute value of the difference is 0.39, the extended irrigation compensation time = 70.6 × 0.39 ≈ 27.5 minutes, rounded up to 28 minutes.

[0103] Add the base irrigation duration and the extended irrigation compensation duration to get the actual total irrigation duration. The actual total irrigation duration (minutes) equals the base irrigation duration (minutes) plus the extended irrigation compensation duration (minutes). For example, if the base irrigation duration is 70.6 minutes and the extended irrigation compensation duration is 28 minutes, the actual total irrigation duration = 70.6 + 28 = 98.6 minutes.

[0104] The hysteresis of crop physiological conditions is continuously monitored, and irrigation is terminated when it falls below a set tolerance. The monitoring operation consists of three sequential steps: First, the calculation function in step S5 is called every 30 minutes after irrigation begins to obtain an updated crop physiological condition hysteresis vector; second, a comprehensive hysteresis value is calculated according to the aforementioned rules; and third, a termination command is immediately sent to the irrigation controller when the comprehensive hysteresis value falls below 1. For example, after 65 minutes of actual irrigation, the latest hysteresis vector [-0.05, 0.82] days is obtained. If the comprehensive value = (-0.05 + 0.82) / 2 = 0.385 < 1, the irrigation valve is closed. Since 0.385 is less than 1, the irrigation controller closes the valve and stops water supply. The exception handling mechanism is as follows: if the crop physiological condition hysteresis data obtained three times in a row is abnormal, an error code 300 is returned and manual verification is initiated.

[0105] Example 2: Figure 2 The present invention provides a structural diagram of a multi-parameter intelligent monitoring system for farmland environment based on the Internet of Things. The multi-parameter intelligent monitoring system for farmland environment based on the Internet of Things includes:

[0106] Environmental collection module, used to collect farmland environmental parameters in real time through the IoT sensor network. Farmland environmental parameters include soil moisture content and soil water content data;

[0107] The rate analysis module is used to obtain the soil moisture content change rate within a set time window when the soil moisture content reaches a preset threshold;

[0108] The stress quantification module is used to calculate the soil moisture deficit rate at each depth layer based on soil moisture data at different soil depths within a set time window and construct a spatiotemporal gradient matrix. The spatiotemporal cumulative effect of stress is quantified by performing principal component analysis on the spatiotemporal gradient matrix and integrating information entropy.

[0109] The delay analysis module is used to analyze the coupling relationship between the soil moisture change rate in historical stress events and the temporal and spatial cumulative effect of stress on the delay of crop physiological recovery, and generate the recovery delay coefficient;

[0110] The hysteresis evaluation module is used to obtain the ratio of the stomatal conductance decline rate during the stress period to the stomatal conductance recovery rate during the recovery period, and to evaluate the degree of hysteresis in the crop physiological state in combination with the recovery delay coefficient;

[0111] The decision-making and control module is used to extend the execution time of irrigation decisions according to the lag degree of crop physiological status until the lag degree is lower than the set tolerance value.

[0112] The calculations involved in the embodiments are all dimensionless numerical calculations, and the preset parameters and thresholds in the calculations are set by those skilled in the art according to actual conditions.

[0113] It should be noted that the present invention can be deployed on the device itself to implement embedded applications, and can also be run on a PC or other terminal with a user interface, thereby meeting various hardware environments and usage requirements.

[0114] The above embodiments can be implemented in whole or in part via software, hardware, firmware, or any other combination. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product comprises one or more computer instructions or computer programs. When loaded or executed on a computer, the processes or functions described in the embodiments of this application are fully or partially performed. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wireless or wired transmission. Wired transmission methods include optical fiber, twisted pair, coaxial cable, etc.; wireless transmission methods include infrared, microwave, etc. The computer-readable storage medium can be any available medium accessible by a computer, or a data storage device such as a server or data center that contains a collection of one or more available media. The available medium can be magnetic media (e.g., floppy disks, hard disks, tapes), optical media (e.g., DVDs), or semiconductor media. Semiconductor media can be solid-state drives.

[0115] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and modules described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0116] In the several embodiments provided in this application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are merely schematic. For example, the division of the modules is only a logical function division. In actual implementation, there may be other division methods, such as multiple modules or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or modules, which can be electrical, mechanical or other forms.

[0117] The modules described as separate components may or may not be physically separate, and the components shown as modules may or may not be physical modules, and may be located in one place or distributed across multiple network modules. Some or all of the modules may be selected to achieve the purpose of this embodiment according to actual needs.

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

[0119] If the functions are implemented in the form of software function modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

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

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

Claims

1. A multi-parameter intelligent monitoring method for farmland environment based on the Internet of Things, characterized in that: include: S1. Real-time collection of farmland environmental parameters through the IoT sensor network, including soil moisture content and soil water content data; S2. When the soil moisture content reaches a preset threshold, obtaining the soil moisture content change rate within a set time window; S3. Based on the soil moisture data at different soil depths within a set time window, the soil moisture deficit rate at each depth layer is calculated and a spatiotemporal gradient matrix is ​​constructed. The spatiotemporal cumulative effect of stress is quantified by performing principal component analysis on the spatiotemporal gradient matrix and integrating information entropy. S4. Analyze the coupling relationship between the soil moisture change rate during historical stress events and the temporal and spatial cumulative effect of stress on crop physiological recovery delay, and generate a recovery delay coefficient; S5. Obtain the ratio of the stomatal conductance decrease rate during the stress period to the stomatal conductance recovery rate during the recovery period, and evaluate the degree of hysteresis in the crop physiological state in combination with the recovery delay coefficient; S6. Extend the irrigation decision execution time according to the lag degree of the crop physiological state until the lag degree is lower than the set tolerance value.

2. The method for multi-parameter intelligent monitoring of farmland environment based on the Internet of Things according to claim 1, characterized in that: Soil moisture data among farmland environmental parameters is obtained through multi-layer soil moisture sensors deployed at the crop root distribution layer. The multi-layer soil moisture sensors simultaneously collect soil volumetric moisture content at at least three typical soil layer depths. Typical soil depths include shallow root zone, middle root zone and deep root zone; The frequency of soil volumetric water content acquisition matches the cycle of crop transpiration rate changes, and the corresponding timestamp and sensor location coordinates are recorded synchronously for each acquisition; The IoT sensor network adopts a star topology and transmits soil volumetric moisture content data to edge computing nodes in real time through a multi-hop routing protocol.

3. The method for multi-parameter intelligent monitoring of farmland environment based on the Internet of Things according to claim 2, characterized in that: The preset threshold is the critical value of soil volumetric water content corresponding to the permanent wilting coefficient of the crop; The start time of the time window is set to the moment when the soil volume water content first falls below the preset threshold, and the end time is the current real-time collection time; The soil moisture change rate is calculated by linear regression fitting of all soil volumetric moisture data points within a set time window, where the independent variable is the time increment from the starting moment and the dependent variable is the measured value of the soil volumetric moisture content at the corresponding moment; The slope value of the linear regression fitting is the rate of change of soil moisture content, and its unit is expressed in percentage of daily moisture content change.

4. The method for multi-parameter intelligent monitoring of farmland environment based on the Internet of Things according to claim 3, characterized in that: Soil moisture data at different soil depths refer to the soil volumetric water content in the shallow root zone, middle root zone, and deep root zone; The soil water deficit rate is calculated by taking the first-order differential derivative of the soil volume water content series at each depth layer within the set time window to obtain the instantaneous deficit rate at each sampling moment; The row vectors of the spatiotemporal gradient matrix correspond to the sampling moments within the set time window, and the column vectors correspond to the instantaneous deficit rate values ​​of the shallow root zone, the middle root zone, and the deep root zone. The principal component analysis takes the spatiotemporal gradient matrix as input and extracts the first principal component eigenvalue and the corresponding eigenvector; Information entropy fusion is achieved by calculating the Shannon entropy of the eigenvector components; The cumulative stress intensity index was generated by multiplying the first principal component eigenvalue with the Shannon entropy of the eigenvector components.

5. The method for multi-parameter intelligent monitoring of farmland environment based on the Internet of Things according to claim 4 is characterized in that: Historical stress events refer to crop growth cycles in which the soil volumetric water content has fallen below a preset threshold; The soil moisture change rate was calculated by linear regression fitting during historical stress events. The temporal and spatial cumulative effect of coercion adopts the coercion cumulative intensity index value generated from historical coercion events; The delay in crop physiological recovery is quantified by the time required for stomatal conductance to return to normal levels after the end of the historical stress event; The coupling relationship was established through a multiple linear regression model: The soil moisture change rate and stress accumulation intensity index were used as independent variables, and the crop physiological recovery delay time was used as the dependent variable. The recovery delay coefficient was generated by fitting the regression coefficients of the multiple linear regression model.

6. The method for multi-parameter intelligent monitoring of farmland environment based on the Internet of Things according to claim 5, characterized in that: The stress period refers to the period when the soil volumetric water content is lower than the preset threshold, and the recovery period refers to the period after the soil volumetric water content returns to the preset threshold; The stomatal conductance decline rate during the stress period was determined by the absolute value of the slope of the linear fitting of the stomatal conductance monitoring values ​​during the stress period; The stomatal conductance recovery rate during the recovery period was determined by the positive slope of the linear fitting of the stomatal conductance monitoring values ​​during the recovery period; The ratio was calculated as the quotient of the absolute value of the stomatal conductance decrease rate during the stress period and the stomatal conductance recovery rate during the recovery period; The recovery delay coefficient directly refers to the recovery delay coefficient value generated in the historical coercion event; The hysteresis degree of crop physiological state is generated by multiplying the recovery delay coefficient value and the ratio value.

7. The method for multi-parameter intelligent monitoring of farmland environment based on the Internet of Things according to claim 6, characterized in that: The tolerance value is set as the critical value of the hysteresis degree of the crop physiological state equal to 1; The specific ways to extend the execution time of irrigation decisions are: Get the basic execution time of the current irrigation decision; Calculate the absolute value of the difference between the hysteresis degree of the crop physiological state and the set tolerance value; Multiply the absolute value of the difference by the basic execution time to obtain the irrigation extension compensation time; Add the basic execution time and the irrigation extension compensation time to get the actual total irrigation execution time; Continuously monitor the hysteresis of crop physiological status and terminate irrigation when the hysteresis of crop physiological status falls below the set tolerance value.

8. A multi-parameter intelligent monitoring system for farmland environment based on the Internet of Things, used to implement the multi-parameter intelligent monitoring method for farmland environment based on the Internet of Things according to any one of claims 1 to 7, characterized in that: include: Environmental collection module, used to collect farmland environmental parameters in real time through the IoT sensor network. Farmland environmental parameters include soil moisture content and soil water content data; The rate analysis module is used to obtain the soil moisture content change rate within a set time window when the soil moisture content reaches a preset threshold; The stress quantification module is used to calculate the soil moisture deficit rate at each depth layer based on soil moisture data at different soil depths within a set time window and construct a spatiotemporal gradient matrix. The spatiotemporal cumulative effect of stress is quantified by performing principal component analysis on the spatiotemporal gradient matrix and integrating information entropy. The delay analysis module is used to analyze the coupling relationship between the soil moisture change rate in historical stress events and the temporal and spatial cumulative effect of stress on the delay of crop physiological recovery, and generate the recovery delay coefficient; The hysteresis evaluation module is used to obtain the ratio of the stomatal conductance decline rate during the stress period to the stomatal conductance recovery rate during the recovery period, and to evaluate the degree of hysteresis in the crop physiological state in combination with the recovery delay coefficient; The decision-making and control module is used to extend the execution time of irrigation decisions according to the lag degree of crop physiological status until the lag degree is lower than the set tolerance value.

Images