Soil humidity closed loop control system based on lysimeter weight measurement

Abstract

The application discloses a soil humidity closed-loop control system based on weight measurement of lysimeter and relates to the technical field of automatic control of agricultural irrigation.The system comprises a weighing type lysimeter, a data acquisition and preprocessing module, a water loss rate calculation module, a delay measurement and storage module, a dynamic threshold generation module, a comparison trigger module and an irrigation execution module.The application aims at the problem that the existing irrigation control system ignores the dynamic coupling between the inherent physical delay of the execution mechanism and the real-time water loss rate of the soil, the water loss rate of the soil is obtained in real time through the water loss rate calculation module, the inherent physical delay time of the system is calibrated on site through the delay measurement and storage module, the dynamic compensation amount is obtained by multiplying the two through the dynamic threshold generation module and is superimposed on the target minimum allowable weight to form a dynamic trigger threshold, irrigation is triggered in advance when the measured weight reaches the dynamic threshold, and the soil weight is not lower than the target threshold when the water reaches the soil column after the physical delay.The application realizes accurate offset of the under-compensation deviation.

Description

Technical Field

[0001] This invention relates to the field of automatic control technology for agricultural irrigation, specifically to a closed-loop control system for soil moisture based on lysimeter weighing. Background Technology

[0002] A gravimetric lysimeter is a standard instrument for measuring soil evaporation and plant transpiration. Its basic principle is to monitor changes in the total weight of a soil column in real time using a high-precision weighing sensor, and then infer changes in soil moisture. With the development of precision agriculture, using lysimeter weighing data to guide irrigation decisions has gradually become a research hotspot, leading to various technical solutions that combine lysimeters with irrigation control systems.

[0003] The basic control logic of these existing systems is consistent: a lower threshold for soil weight is set, irrigation is initiated when the measured weight drops below this threshold, and stops when the upper threshold is reached. However, this "threshold-triggered" control mode has a generally overlooked physical defect. From the time the controller issues the start command to the time the irrigation water actually reaches the soil surface, there is an incompressible inherent physical delay time, consisting of the solenoid valve response time, water flow transmission time, and water infiltration time. During this delay, soil evaporation and transpiration do not stop, and soil moisture continues to be lost. Therefore, the soil weight at the actual irrigation time is already lower than the weight at the time the controller issued the command; the difference is equal to the water loss rate multiplied by the delay time. The higher the water loss rate, the greater this deviation; moreover, the water loss rate fluctuates continuously throughout the day with changes in meteorological factors such as sunlight and temperature, causing the deviation itself to be dynamically changing.

[0004] In addition to the aforementioned core defects, existing systems also have other shortcomings. When soil moisture content reaches saturation, irrigation water or precipitation will seep out from the bottom of the soil column. This weight loss does not represent water loss due to evaporation and transpiration, but existing systems do not deduct it from the total weight change, potentially leading to misjudgments of soil water shortage. When operating in the field, the weight signal of a gravimetric lysimeter is inevitably affected by environmental vibrations, wind loads, and temperature drift. Existing systems lack a systematic signal processing mechanism to address these interferences, resulting in drastic fluctuations in the calculated water loss rate and frequent fluctuations in the trigger threshold, affecting system stability. Furthermore, while some technical solutions can provide irrigation forecast information, this information is only for manual reference; the system itself lacks the closed-loop capability to automatically execute irrigation control based on the forecast results.

[0005] Overall, there is currently no existing technology that can simultaneously and effectively address the dynamic coupling problem of execution delay and continuous water loss, while also taking into account multiple requirements such as leakage deduction, signal noise suppression, and adaptive control. Summary of the Invention

[0006] To overcome the above-mentioned defects of the prior art, the embodiments of the present invention provide a closed-loop control system for soil moisture based on lysimeter weighing. The technical problem to be solved is that the existing soil moisture control system based on lysimeter weighing fails to consider the dynamic coupling relationship between the inherent physical delay time of the irrigation actuator and the real-time soil moisture loss rate, resulting in the soil weight being lower than the target threshold at the actual water replenishment time, producing an under-replenishment deviation that dynamically changes with the water loss rate, and the irrigation control accuracy deteriorates significantly when the evapotranspiration rate fluctuates.

[0007] To address the aforementioned technical problems, this invention provides a closed-loop control system for soil moisture based on lysimeter weighing.

[0008] 1. Overall Concept The concept of this invention is to transform the inherent physical delay time of the irrigation actuator from an ignored implicit parameter into an explicit parameter that actively participates in control decision-making, and to multiply and couple it with the real-time water loss rate to generate a dynamic compensation quantity, which is then used to correct the irrigation trigger threshold in real time.

[0009] To understand the solution of this invention, it is essential to first recognize a key physical fact. The aforementioned mechanism for the generation of undercompensation deviation can be expressed by a simple mathematical relationship: .

[0010] in, It is the real-time rate of soil moisture loss. This refers to the inherent physical delay time of the irrigation actuator. In other words, the magnitude of the under-irrigation deviation is determined by two factors: how quickly the soil loses water and how long the execution delay is. If the system can trigger irrigation in advance... The amount of time that allows water to pass through a delayed process. If the water reaches the surface of the soil column precisely, then the weight of the soil at the moment of water replenishment will be exactly not less than the target threshold, and the deviation will be accurately offset. The difficulty lies in: It is dynamic and changing. It is not a fixed constant due to environmental factors such as water pressure and temperature, so the lead time cannot be fixed and must be calculated in real time.

[0011] Based on the above analysis, the technical solution of the present invention revolves around the following logic. To obtain accurate compensation amounts... Three things need to be done: First, the real-time water loss rate is continuously and stably extracted from the lysimeter weight signal. ; Second, the inherent physical delay time of the system under current operating conditions is accurately measured through on-site calibration. And it corrects for changes in water pressure; Third, the two are multiplied to obtain the compensation amount, which is then added to the target threshold to form a dynamic trigger threshold. Irrigation is triggered in advance when the measured weight reaches this dynamic threshold.

[0012] Based on the above logic, this invention designs a system comprising a weighing lysimeter, a data acquisition and preprocessing module, a water loss rate calculation module, a delay measurement and storage module, a dynamic threshold generation module, a comparison triggering module, and an irrigation execution module. Each component is described below.

[0013] (1) Weighing lysimeter The gravimetric lysimeter is the sensing component of the system, used to output a weight signal characterizing the weight of the soil column in real time. It includes a lysimeter cylinder, a weighing sensor located at the bottom of the cylinder, and a leakage measurement device connected to the drainage outlet at the bottom of the cylinder.

[0014] The lyostat cylinder is cylindrical or square in shape, made of stainless steel or carbon steel, with a sealed bottom, and filled with undisturbed soil or layered backfill. The cross-sectional area of ​​the cylinder is determined according to the application scenario, and is typically [area missing]. or The height of the cylinder is generally to A filter layer and a drain pipe are installed at the bottom of the cylinder. The filter layer consists of three layers, from top to bottom: a coarse sand layer (thickness...). Particle size ), fine gravel layer (thickness) Particle size ) and coarse gravel layer (thickness) Particle size Geotextile is laid between each layer, and drainage pipes are connected to leakage measurement devices.

[0015] An isolation layer is installed between the cylinder and the surrounding soil to prevent surface water from seeping into the cylinder laterally along the cylinder wall during rainfall and interfering with the measurement. This isolation layer can be implemented in two ways: One method is to fill the annular gap between the outer wall of the cylinder and the surrounding soil with bentonite (thickness...). Bentonite expands when it comes into contact with water, forming a dense, impermeable layer. Secondly, it adopts a double-layer cylinder structure, with the inner cylinder serving as the working cylinder and the outer cylinder as the protective cylinder, and a space between the inner and outer cylinders. The air gap.

[0016] The weighing system employs three or four high-precision sensors, evenly distributed at the support points at the bottom of the cylinder, using a direct-weighing structure to directly measure the total weight of the soil column. The sensor range is typically [missing information]. to The accuracy class is no lower than C3, and the protection class is no lower than IP68. The sensor has a temperature compensation function and can operate within a certain range. to Maintain measurement accuracy within the ambient temperature range.

[0017] The seepage measurement device is located at the drainage outlet at the bottom of the cylinder and is used to measure the amount of water seeping into the soil. It can be a tipping bucket flow meter or an electronic balance weighing method, with a measurement resolution of not less than [missing information]. Water depth. The measured leakage rate is used to correct for subsequent water loss rates.

[0018] (2) Data acquisition and preprocessing module The data acquisition and preprocessing module is electrically connected to the weighing sensor. It is used to acquire weight signals at a fixed sampling frequency and output valid weight signals after preprocessing the signals. .

[0019] The raw signal output by the load cell contains various interferences: impulse noise caused by environmental vibration and wind load impact, the sensor's own white noise, and measurement drift caused by temperature changes. If these interferences are used directly for water loss rate calculation without processing, This can cause severe jitter, leading to unstable compensation values. The role of the data acquisition and preprocessing module is to purify the raw signal and provide a reliable data foundation for subsequent calculations.

[0020] This module includes an analog-to-digital conversion unit, a median filtering unit, a moving average filtering unit, and a temperature correction unit. The analog-to-digital conversion unit operates at a fixed sampling frequency (preferably...). Convert the analog weight signal into a digital raw weight sequence. The reciprocal of the sampling frequency is denoted as The median filter unit is used for continuous... Median filtering was performed on the original weight data points. (For odd numbers between 5 and 11), output the intermediate value sequence. .

[0021] Median filtering is specifically designed to remove impulse noise—a type of noise characterized by large amplitude and short duration. Median filtering effectively identifies and eliminates this noise with minimal impact on the smooth changes in the true signal. If this step is omitted, the instantaneous weight spikes caused by wind load impacts may be misinterpreted as true changes in soil weight, leading to… An abnormal spike appeared.

[0022] Moving average filter unit for continuous The arithmetic mean of the intermediate value sequence data points is used to output the effective weight signal. . The value of is related to the inherent physical delay time. Related, specifically the relationship is as follows ,and .

[0023] and The correlation is intentionally designed: a filter window that is too large will... Phase lag exceeds The allowable range leads to a shift in the compensation timing; if the window is too small, white noise suppression is insufficient. It will still shake. By... and Hooks are used to ensure that the filtering window matches the system's response requirements under different delay times.

[0024] The temperature correction unit reads the ambient temperature sensor and calls up the pre-stored temperature drift curve coefficients. ,according to Make corrections. This is the reference temperature. This step is to eliminate residual temperature drift of the load cell. Although the sensor itself has a temperature compensation circuit, approximately [amount missing] temperature drift still exists after long-term operation. to The residual drift. If left uncorrected, this slow change will be... The calculated absorption led the system to misjudge the true trend of soil moisture changes.

[0025] (3) Water loss rate calculation module The water loss rate calculation module is electrically connected to the data acquisition and preprocessing module, and is used to calculate the water loss rate based on the effective weight signal. The rate of change over time is used to calculate the real-time water loss rate of current soil moisture. .

[0026] rate of water loss It is one of the two multiplier factors of the compensation amount, and its calculation accuracy directly determines the accuracy of the compensation amount. Too large an amount will lead to over-irrigation, while too small an amount will result in under-irrigation and the deviation will still exist.

[0027] The water loss rate calculation module uses the least squares linear regression method to calculate... Specifically, take the values ​​before the current time. One effective weight signal data point ( ), in time x-axis A linear regression was performed with the vertical axis as the ordinate, and the slope of the regression line was used as the real-time water loss rate. . The time window covered by each data point With moving average filter window Between satisfy , The range of values ​​is to . ,in for The output sampling interval.

[0028] slope The calculation formula is:

[0029] in .

[0030] The least squares linear regression method is used instead of a simple two-point difference method because the regression method utilizes... The signal contains all the information of each data point, but is not sensitive to individual outliers. Even after filtering, residual noise may still exist in the signal; regression methods can mitigate this noise. To minimize the impact.

[0031] rate of water loss A positive value indicates that soil moisture is being lost, while a negative value indicates that it is increasing. System command Since there is no risk of water shortage during the water replenishment phase, there is no need for advance compensation.

[0032] In addition, the water loss rate calculation module also obtains the leakage rate measured by the leakage measurement device. .when When it is greater than zero, the module will Revised to ,in This represents the total weight change rate of the gravimetric lysimeter. This step is necessary because when the soil reaches saturation, irrigation water or rainfall will seep out from the bottom of the soil column. This weight loss does not represent the water loss due to crop transpiration. If seepage is not deducted, the system may mistakenly believe that soil moisture is being lost rapidly, thus triggering irrigation prematurely when it is not needed.

[0033] (4) Delay measurement and storage module The delay measurement and storage module acquires the effective weight signal through the data acquisition and preprocessing module. It is used to perform a delay calibration procedure to determine and store the inherent physical delay time between the issuance of the start command and the start of irrigation water supply to the soil column. .

[0034] It is another multiplier factor of the compensation amount, and also one of the key parameters that distinguishes this scheme from existing technologies. It consists of three parts: ,in This refers to the electrical response time of the solenoid valve. The time it takes for water to travel through the pipe. This refers to the time it takes for moisture to seep from the soil surface to be effectively captured by the weighing sensor. The value cannot be accurately obtained by consulting the equipment manual because it depends not only on the solenoid valve model, but also on the actual length of the pipeline, the number of bends, the water pressure, and the infiltration characteristics of the soil. Therefore, this invention uses on-site calibration to determine the value. .

[0035] The delay measurement and storage module executes the delay calibration procedure after the system is initially installed, after changes to the irrigation pipeline, or after the solenoid valve of the irrigation execution module is replaced. The calibration procedure includes the following steps: S1: The comparison trigger module issues a test start command and simultaneously starts a timer to record the first time point. ; S2: Continuous monitoring of effective weight signals Changes; S3: When continuous The effective weight signal value of each sampling point in each sampling point is higher than that of the continuous The sampling points were adjacent to each other. The arithmetic mean of the sampling points, and the continuous When the increment of the value of the last sampling point among all sampling points relative to the arithmetic mean exceeds the preset rise threshold, it is determined that irrigation water has started to replenish the soil column, the timing is stopped, and the second time point is recorded. ,in Integers between 3 and 10 The value is an integer between 5 and 20, and the threshold for determining an increase is 3 to 10 times the resolution of the weighing sensor. S4: Calculation It is stored in non-volatile memory.

[0036] After calibration, the system repeats the above steps three times and takes the average value to eliminate random errors.

[0037] Due to water flow transmission time It is inversely proportional to the square root of the water pressure; when the water pressure changes, The actual value also changes accordingly. Therefore, the delay measurement and storage module also includes a water pressure correction unit. This unit acquires the real-time water pressure value of the irrigation water source. Read the reference water pressure value during calibration. Each time you get Then use the following formula to Make corrections:

[0038] in To calibrate the inherent physical delay time, This is the corrected current delay time. When and When the deviation exceeds the preset range (15% to 25%), it indicates that the water pressure change has exceeded the reasonable range for formula correction, and the system will automatically trigger recalibration.

[0039] The delay measurement and storage module also measures and stores the inherent physical shutdown delay time of the irrigation execution module. It is used for symmetric compensation during the stopping phase, and its measurement method is the same as... symmetry.

[0040] (5) Dynamic threshold generation module The dynamic threshold generation module is electrically connected to both the water loss rate calculation module and the delay measurement and storage module, and calculates the dynamic trigger threshold in real time according to the following formula:

[0041] in For dynamic trigger threshold, The preset target minimum allowable weight, For real-time water loss rate, This is the current delay time after water pressure correction.

[0042] This formula directly reflects the aforementioned compensation logic: This is the amount of water the soil will continue to lose within the current water loss rate and the time delay. Add this compensation amount to... Up, and you get an "early" The threshold for "triggering in seconds". When the actual weight... Down to Time -- at this moment Actually, it's even higher than The system starts irrigation. Water passes through... It reached the soil column seconds later, just as the soil weight dropped to its lowest point. .

[0043] This is set by the user based on the crop and soil type. The user sets the target lower limit for soil moisture content. Then, press Converted to weight value, where To dry the soil, The cross-sectional area of ​​the cylinder is... The height of the soil column. This is the density of water.

[0044] The dynamic threshold generation module also includes a limiting unit. When Greater than the preset irrigation stop threshold At that time, the limiting unit forced order ( (A positive constant less than the sensor resolution) to prevent the trigger threshold from exceeding the stop threshold, causing system oscillation; when Less than At that time, the limiting unit forced order This prevents the system from never triggering when the water loss rate is zero or negative.

[0045] (6) Comparison trigger module The comparison trigger module is electrically connected to the data acquisition and preprocessing module and the dynamic threshold generation module, respectively, for use in... The system sends a start command to the irrigation execution module, and in Reaching or exceeding A shutdown command is sent to the irrigation execution module.

[0046] The comparison trigger module is the execution link that converts the compensation amount into the actual control action. Its judgment logic is very straightforward: when the measured weight drops to the dynamic threshold, irrigation is started; when the weight rises back to the stop threshold, irrigation is stopped.

[0047] The comparison trigger module also includes a debounce unit. The debounce unit is used to: when the first detection... At that time, start the image stabilization timer. ( to During the timing period, the comparison results are continuously monitored; after the timing ends, the proportion of sampling points that meet the trigger conditions is counted. If the proportion is greater than the preset threshold (60% to 90%), the start command is officially output; otherwise, no output is output and the timer is reset.

[0048] This anti-vibration unit addresses a practical problem: even after filtering, weighing signals may still exhibit momentary fluctuations. If irrigation is triggered with every fluctuation, the solenoid valve will frequently open and close, affecting its lifespan and causing unnecessary irrigation. The anti-vibration unit effectively distinguishes between genuine water shortage trends and occasional signal fluctuations by requiring the triggering conditions to maintain a sufficiently high satisfaction rate for a sufficiently long period.

[0049] (7) Irrigation execution module The irrigation execution module is electrically connected to the comparison trigger module, and is used to start irrigation water replenishment after receiving an start command, and to stop irrigation water replenishment after receiving a stop command.

[0050] The irrigation execution module includes a water supply device, a solenoid valve, a flow metering device, and an irrigation network. The water supply device includes a water source and a water pump. The solenoid valve is connected to the controller via a cable. The irrigation network connects from the outlet of the solenoid valve to the irrigation nozzles or drip irrigation pipes above the soil column of the lyostat.

[0051] After the irrigation execution module starts irrigation, the comparison trigger module issues a shutdown command when the following stop conditions are met:

[0052] The stopping condition is symmetrical to the starting condition. Closing also involves a physical delay. If we wait until achieve The shutdown command has just been issued, but the actual weight will exceed [a certain value] when it stops. Therefore, the system in Still lower than But the difference is exactly equal to The circuit is closed in advance so that the weight reaches the target value after the closing delay. .

[0053] The irrigation execution module also has a maximum irrigation time protection setting. If the irrigation duration exceeds the preset limit... (default If the stop condition is still not met (adjustable), the system will be forcibly shut down and an alarm will be issued to prevent continuous flooding due to valve jamming or sensor failure.

[0054] (8) Recalibration trigger conditions The accuracy of the compensation amount directly determines its precision. The solenoid valve may drift due to aging, pipe deposits, temperature changes, and water pressure fluctuations. To address this, the present invention incorporates an automatic recalibration mechanism.

[0055] The delay measurement storage module automatically re-executes the delay calibration procedure when any of the following conditions are met: (a) The system's cumulative running time is calculated from the last calibration completion time. Every time the preset periodic calibration cycle is reached ( to ); (b) The change in ambient temperature exceeds the preset temperature change threshold. to ); (c) The deviation between the real-time water pressure value of the irrigation water source and the reference water pressure value is greater than the preset deviation threshold (15% to 25%). (d) The irrigation pipeline components of the irrigation execution module were replaced.

[0056] These four triggering conditions correspond to four typical situations: solenoid valve aging, temperature change, water pressure fluctuation, and pipeline replacement. Drift scenarios cover all possible scenarios throughout the system's entire lifecycle. The main factors of change. Covering these conditions through a combination of factors, It can always remain accurate, and the compensation amount will not be affected by... The drift causes deviation.

[0057] 3. Data Flow The following describes the data flow relationship within the system according to the direction of signal flow.

[0058] Path 1 (Main Path of Weight Signal): The weighing sensor of the gravimetric lysimeter outputs an analog voltage signal, which is transmitted via a shielded cable to the analog-to-digital converter (ADC) of the data acquisition and preprocessing module. The ADC converts the analog signal into a digital raw weight sequence. The median filtering unit is used for Perform median filtering to output the median value sequence The moving average filter unit for Perform moving average filtering to output effective weight signal Temperature correction unit Perform temperature drift correction.

[0059] Path 2 (Calculation path for water loss rate): Effective weight signal The input is sent to the water loss rate calculation module, which then retrieves the values ​​from the previous time. Least squares linear regression was performed on the data points to output the real-time water loss rate. Simultaneously obtain the leakage rate Make corrections.

[0060] Path 3 (Dynamic Threshold Generation Path): The water loss rate calculation module will... The output is sent to the dynamic threshold generation module, and the delay measurement and storage module stores the current delay time after water pressure correction. The output is sent to the dynamic threshold generation module, which generates the threshold according to the formula. Calculate the dynamic trigger threshold.

[0061] Path 4 (Trigger Decision and Execution Path): Comparison of Trigger Module Receives and Perform comparison operations The anti-shake unit confirms the comparison result and outputs an start command to the irrigation execution module after confirmation. The irrigation execution module then drives the solenoid valve to open.

[0062] Path 5 (Stop Decision Path): During irrigation, the comparison trigger module continuously receives... and and read from the latency measurement storage module Real-time determination of stopping conditions When the conditions are met, a shutdown command is sent to the irrigation execution module.

[0063] Path Six (Calibration and Correction Path): The delay measurement storage module obtains data through the data acquisition and preprocessing module. Used for calibration procedures; after calibration is completed, The water pressure is stored in non-volatile memory, and the water pressure correction unit obtains the real-time water pressure. and reference water pressure Corrected according to the formula The revised Provided to the dynamic threshold generation module.

[0064] The technical effects and advantages of this invention are as follows: First, it accurately compensates for dynamic undercompensation deviations. This invention addresses the inherent physical delay time of the system. With real-time water loss rate Multiply to obtain the compensation amount , superimposed A dynamic trigger threshold is established. When the measured weight reaches this threshold, the system is activated, and moisture content is reduced after a delay. The weight of the soil upon arrival was exactly not less than This achieves precise cancellation of undercompensation deviations.

[0065] Second, it achieves proactive and precise control. The control mode has shifted from "replenishing water only when it is already scarce" to "replenishing water in advance when it is anticipated that a water shortage is imminent." The advance amount is not an empirical estimate, but based on measured physical parameters. and real-time rate The precise calculated value.

[0066] Third, it adapts to changes in evaporation rate. Because Real-time calculation, dynamic threshold The system floats synchronously with the evapotranspiration rate, eliminating the need for manual parameter adjustments based on different seasons and weather conditions.

[0067] Fourth, bidirectional symmetrical compensation for start-up and stop. This invention addresses the start-up delay. and closing delay Simultaneously, compensation is performed to eliminate overshoot and over-adjustment during the closing process, ensuring that the stopping accuracy matches the opening accuracy.

[0068] Fifth, multiple anti-interference mechanisms ensure stability. Median filtering eliminates impulse noise, moving average filtering suppresses white noise, temperature correction eliminates temperature drift, leakage deduction eliminates gravity water interference, anti-vibration unit distinguishes between real water shortage and occasional fluctuations, and amplitude limiting unit prevents system oscillation under extreme operating conditions.

[0069] Sixth, full lifecycle adaptive maintenance. Through four mechanisms—regular calibration, temperature change-triggered calibration, water pressure deviation-triggered calibration, and pipeline replacement-triggered calibration—it ensures... It remains effective throughout the entire system lifecycle.

[0070] Seventh, it has low implementation cost and good scalability. This invention extends the control strategy based on the existing weighing lysimeter hardware platform without changing the basic structure and measurement principle, and is mainly implemented through software programming.

[0071] Eighth, it has wide applicability. This invention is not only applicable to agricultural irrigation, but also to vegetation water management in ecological restoration projects, precision irrigation in greenhouses, and various scientific research and engineering scenarios for water control based on lysimeter weighing. Attached Figure Description

[0072] Figure 1 This is a schematic diagram of the system module composition of the present invention; Figure 2 This is a flowchart of the delay calibration procedure of the present invention; Figure 3 This is a flowchart illustrating the anti-shake unit operation of the present invention. Figure 4 This is a flowchart illustrating the operation of the water pressure correction unit of the present invention. Detailed Implementation

[0073] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0074] Example 1 As attached Figures 1 to 4 The specific implementation details of the closed-loop soil moisture control system based on lysimeter weighing, as shown below, are as follows: I. Overall System Structure A weighing lysimeter includes a lysimeter cylinder, a weighing sensor located at the bottom of the cylinder, and a leakage measurement device located at the drainage outlet at the bottom of the cylinder, for outputting a weight signal characterizing the weight of the soil column in real time. The data acquisition and preprocessing module is electrically connected to the weighing sensor and is used to acquire the weight signal at a fixed sampling frequency, preprocess the weight signal, and output a valid weight signal. ; The water loss rate calculation module, electrically connected to the data acquisition and preprocessing module, is used to calculate the water loss rate based on the effective weight signal. The rate of change over time is used to calculate the real-time water loss rate of current soil moisture. ; The delay measurement and storage module acquires the effective weight signal through the data acquisition and preprocessing module. This is used to determine the effective weight signal after the comparison trigger module issues a test start command. The changing trend is used to perform a delay calibration procedure to determine and store the inherent physical delay time between the issuance of the start command and the start of irrigation water supply to the soil column. And the inherent physical shutdown delay time between issuing the shutdown command and the complete cessation of irrigation water supply to the soil column. ; The dynamic threshold generation module is electrically connected to both the water loss rate calculation module and the delay measurement and storage module, and is used to generate thresholds according to the formula. Real-time calculation of dynamic trigger thresholds, where The preset minimum allowable weight; The comparison trigger module is electrically connected to both the data acquisition and preprocessing module and the dynamic threshold generation module, and is used to trigger the comparison of the effective weight signal. Less than or equal to the dynamic trigger threshold The system sends an activation command to the irrigation execution module, and the effective weight signal is received. Reaching or exceeding the preset irrigation stop threshold A shutdown command is sent to the irrigation execution module at that time; The system also includes an irrigation execution module, which is electrically connected to the comparison trigger module. This module is used to initiate irrigation water supply to the soil column upon receiving the start command, and to stop irrigation water supply upon receiving the stop command.

[0075] The specific structure and function of each module will be described in detail below through several embodiments.

[0076] II. Specific Implementation Methods of Weighing-Type Delphimeter A weighing lysimeter consists of a lysimeter cylinder, a weighing sensor, and a leakage measurement device.

[0077] The lysostat body is cylindrical, and in this embodiment, the inner diameter is [missing information]. ,high Wall thickness The material is 304 stainless steel. Bottom of the cylinder. This is a reverse filtration layer, which consists of three layers: The first layer from top to bottom is a coarse sand layer, with a thickness of... Particle size to ; The second layer is a fine gravel layer, with a thickness of... Particle size to ; The third layer is a coarse gravel layer, with a thickness of... Particle size 10 to .

[0078] Geotextile is laid between each layer to prevent the mixing and loss of materials between layers. Upper part of the cylinder. The original soil was filled in, and the backfilling method was used. The loose thickness of each layer did not exceed [a certain value]. After compaction to the original soil density, the thickness is controlled at [a certain value]. The cylinder is buried in the field, with its opening above the ground. To prevent surface runoff from entering the cylinder, an isolation layer is installed between the cylinder and the surrounding soil. In this embodiment, this is achieved by filling the annular gap between the outer wall of the cylinder and the surrounding soil with bentonite, with a filling thickness of [missing information]. When bentonite comes into contact with water, it expands to form a dense impermeable layer, which can effectively prevent surface water from seeping into the cylinder wall laterally during rainfall.

[0079] A concentric windproof cover is installed around the perimeter of the lysimeter cylinder. Its specific structure is as follows: [The text abruptly ends here, so the translation stops as well.] A perforated aluminum alloy baffle is installed at the location, with a hole diameter of... Open area ratio The baffle height is flush with the cylinder opening. This wind shield can reduce the horizontal wind load on the cylinder by approximately [missing information]. This effectively reduces the instantaneous impact of wind on the weighing sensor.

[0080] The weighing sensor uses three cylindrical compressive load cells, each with a range of [missing information]. The three sensors are evenly distributed under the load-bearing plate at the bottom of the cylinder, in a... The signals from three evenly distributed sensors are connected in parallel to a junction box and then combined into a single analog signal output. The sensors have built-in temperature-sensitive elements. to Within the ambient temperature range, the built-in circuitry can automatically perform zero-point temperature compensation and sensitivity temperature compensation, and the temperature drift coefficient after compensation does not exceed [a certain value]. If the ambient temperature exceeds the above range, the system will issue an "over-limit alarm" and pause the automatic control mode, switching to manual mode. It will automatically switch back to automatic mode once the temperature returns to the allowable range.

[0081] The leakage measurement device is located at the drain outlet at the bottom of the cylinder; in this embodiment, a tipping bucket flow meter is used. The capacity of each bucket is... The corresponding water depth is The measurement resolution is Water depth. Each time the bucket flips, it generates a pulse signal. The main control chip uses pulse counting to calculate the cumulative leakage amount and leakage rate. .

[0082] The specific calculation method is as follows: the main control chip uses... To read the cumulative value of the pulse counter at intervals and calculate the number of pulses per second. Multiply by the capacity per bucket Then multiply by the density of water Converted to The leakage rate is expressed in units, i.e.:

[0083] When the tipping bucket has not tipped over Zero, It is also zero.

[0084] III. Detailed Implementation of the Data Acquisition and Preprocessing Module The data acquisition and preprocessing module includes an analog-to-digital conversion unit, a median filtering unit, a moving average filtering unit, and a temperature correction unit.

[0085] The analog-to-digital conversion unit operates as follows: The three load cell signals are connected in parallel within the junction box and then transmitted via shielded cable to a high-precision transmitter in the data acquisition cabinet. The length of the shielded cable does not exceed [a certain value]. A twisted-pair shielding structure is used to reduce electromagnetic interference. The transmitter's internal hardware includes a first-order low-pass RC filter with a cutoff frequency of [frequency missing]. Used for filtering Power frequency interference.

[0086] The transmitter integrates 24-bit... - The ADC chip is set to a sampling rate of 20 times per second. .Should This is what is referred to in this invention as the "fixed sampling frequency," and its reciprocal equal The digital signal converted by the ADC is sent to the ARM Cortex-M4 main control chip (model STM32F407, operating frequency...). ), each main control chip Read the ADC register value once, and convert it to a scale value. Original weight value in units .

[0087] The scaling coefficients are obtained through weight calibration: standard weights are placed on the lysimeter cylinder during system installation. (Accuracy M1 level) Record changes in ADC readings Calculate the slope .in This represents the change in ADC readings before and after the weights were placed. The unit is Reading. This slope is stored in non-volatile memory. Weight calibration is repeated annually and immediately after sensor replacement.

[0088] The median filtering unit works as follows: A circular array buffer is allocated to store the latest consecutive... indivual Data point, corresponding to the most recent The data was used to sort the 11 data points by numerical value, and the 6th value, i.e., the median, was taken as the median value at that time. Eleven points were chosen as the filtering window because 11 is an odd number, ensuring a unique median exists in all cases and avoiding the need to average two median values. The 11 points correspond to... The time window is sufficient to cover the duration of a typical wind load pulse (weight spikes caused by field gusts typically last 200 to 400 ms), thus ensuring that pulse noise is effectively eliminated.

[0089] The moving average filter unit works as follows: it takes the latest continuous... indivual The values ​​are averaged to obtain the effective weight signal. . The value of is related to the inherent physical delay time. Related, specifically:

[0090] in ,and The range of values ​​is limited to Up to 10.

[0091] For example, when hour, ,but The minimum value is 3, so the actual value is 3. ;when hour, ;when hour, ;when hour, .

[0092] The lower limit of the value is determined by the white noise suppression requirements. When the value is less than 3, there are too few data points participating in the averaging, which cannot effectively suppress sensor white noise; the upper limit is determined by the phase lag constraint. When the value is greater than 10, the phase lag of the moving average exceeds one sampling period. This will cause the output of the water loss rate calculation module to be... It cannot reflect the current true transient evaporation rate, thus reducing the accuracy of compensation.

[0093] The temperature correction unit operates as follows: Each Read the ambient temperature sensor (model DS18B20, accuracy) once. (Place the lysimeter in a shaded area to avoid direct sunlight) and call the pre-stored temperature drift curve coefficient. ,right Make corrections. The correction formula is:

[0094] in This is the reference temperature. The measurement process is as follows: The lysimeter cylinder equipped with a weighing sensor is placed in a programmable constant temperature chamber. Under stable conditions of no wind and no vibration, the temperature of the constant temperature chamber is set sequentially to [temperature value missing]. , , , , , , , , , , , , There are 13 temperature points in total, and each temperature point is kept at a constant temperature. Then read the sensor output value. The output value at that time is used as a reference to calculate the output deviation at each temperature point. The temperature drift curve was obtained by fitting. Temperature points at each point The values ​​are stored in the EEPROM in the form of a lookup table, where the interval between adjacent temperature points is [value missing]. Temperature correction is based on the current temperature. Obtain the corresponding value through linear interpolation. value.

[0095] For the sake of simplicity, the following terms will be used throughout the text. Both refer to the effective weight signal after temperature correction.

[0096] IV. Detailed Implementation of the Water Loss Rate Calculation Module The water loss rate calculation module uses the least squares linear regression method to calculate the real-time water loss rate. .

[0097] The data points are selected as follows: Data points from times before the current time are taken. One effective weight signal data point ( ),in For the first The time corresponding to each data point, and the time interval between adjacent data points are... In this embodiment Output interval . The value is determined by the time window. Confirmed, specifically:

[0098] With moving average filter window Related:

[0099] Let's take a few typical cases as examples: When hour, , at this time ; when hour, , ; when hour, , .

[0100] The value of is in to Between, falling within the scope defined in the claims. to Within the range. The lower limit of the value is determined by the minimum number of data points required for linear regression—less than two points are insufficient for effective regression; the upper limit is determined by the required response speed—if the time window is too long... The response lags behind the actual change in evapotranspiration rate, which can change significantly within 1 to 2 minutes under strong sunlight conditions.

[0101] The calculation method for linear regression is as follows: using time... For the horizontal axis and effective weight signal Using the ordinate as the vertical axis, a straight line is fitted using the least squares method. Slope The calculation formula is:

[0102] in The slope This is the current net water loss rate. The unit is The least squares linear regression method is used instead of a simple two-point difference method because the regression method utilizes... All information from each data point, calculated using regression under the same noise level. The standard deviation is approximately equal to that of the finite difference method. The more data points there are, the more stable the rate estimation will be.

[0103] The subsequent processing of the rate value shall be carried out according to the following rules: like If the weight is less than zero (i.e., the cylinder weight increases, indicating rainfall or irrigation), the system will... This judgment is based on three consecutive Execution is performed only when all calculated values ​​are less than zero, rather than a single judgment, to prevent errors caused by instantaneous signal noise. Abnormal return to zero. When The system will automatically recover when the value is positive and three consecutive calculated values ​​are greater than zero. Real-time calculation.

[0104] like Greater than zero and leakage rate If it is greater than zero, then follow the rules. Make corrections. If the corrections are... If it is less than zero, then let This is to prevent the system from failing to trigger if the dynamic trigger threshold is lower than the target minimum allowable weight. When the leakage rate... The duration of consecutive positive values ​​exceeding zero exceeds At that time, the comparison trigger module pauses issuing the start command until... Zeroing and remaining at zero for more than [time period missing] It can be restored later.

[0105] V. Specific Implementation Method of the Delay Measurement and Storage Module The delay measurement and storage module executes the delay calibration program after the system is initially installed, after the irrigation pipeline is changed, or after the solenoid valve of the irrigation execution module is replaced.

[0106] Before starting the calibration procedure, the following prerequisites must be met: First, the soil column of the lyostat is in an unsaturated state, that is, the current soil moisture content is lower than the field capacity, ensuring that irrigation water can be absorbed by the soil and immediately reflected in the weight change. Second, the ambient wind force does not exceed level 3. According to the Beaufort wind scale, level 3 wind corresponds to a wind speed of... to If the wind speed exceeds this value, the calibration will be suspended, because the weight fluctuations caused by strong winds may interfere with the accurate determination of the rising inflection point. Third, the system is in a stable state, that is, in the past... Internal effective weight signal The fluctuation range is less than 10 times the effective resolution of the weighing sensor, ensuring that the weight of the soil column is in a stable natural descent process.

[0107] The calibration sequence is automatically executed by the main control chip, and the specific steps are as follows: The first step is for the main control chip to send an opening calibration pulse to the solenoid valve drive circuit through the I / O port, with a pulse width of... This I / O port operation simultaneously performs two actions at the software level: "outputting a high level" and "starting a timer." Both actions are completed within the same instruction cycle, with an interval of less than [missing information]. ,ensure It precisely corresponds to the start time of the physical action. The main control chip simultaneously activates its internal high-precision timer, with a resolution of [missing information]. Recording moments .

[0108] The second step involves the solenoid valve opening upon receiving a pulse, allowing the water source to operate at its rated pressure of 0.2 to... Water is supplied downwards to the soil column. The outlet of the solenoid valve in the water supply pipeline is connected to the drip irrigation tape above the soil column, and the total length of the water flow path does not exceed [a certain value]. .

[0109] The third step is for the main control chip to continuously... The effective weight signal output by the frequency monitoring data acquisition and preprocessing module .

[0110] The fourth step involves using each new sampling point as a potential starting point for a sliding window judgment. The specific judgment method is as follows: when consecutive... Each sampling point in the sampling points The values ​​are all higher than the values ​​of the five consecutive sampling points immediately preceding them. The arithmetic mean of five consecutive sampling points, and the last sampling point among these five consecutive sampling points. When the increment of the value relative to the arithmetic mean of the 10 consecutive sampling points is greater than the rise threshold, it is determined that irrigation water has begun to replenish the soil column and cause an increase in weight, and the time is recorded. .

[0111] In this embodiment, the threshold for determining the rise is set to 5 times the effective resolution of the weighing sensor. The effective resolution of the sensor is determined by continuously collecting 100 data points under no-load conditions after the sensor is installed, and then calculating the standard deviation of these data points. ,by As the effective resolution, the nominal resolution of the sensor in this embodiment is [missing information]. The actual effective resolution is approximately Therefore, the threshold for determining the rise is .

[0112] Step 5, Calculation .

[0113] After calibration, the system will automatically perform verification: repeat the above steps three times, and take the average of the three measurements as the final value. Value. If the standard deviation of the three measurements exceeds the mean. If an abnormality occurs during calibration, a calibration alarm will be issued, prompting manual inspection of the pipeline and valve status—possible causes include unstable solenoid valve operation, excessive water pressure fluctuations, and air bubbles in the pipeline. If any abnormality occurs during any calibration process (such as no response from the solenoid valve, or the weight signal failing to rise), the system will immediately terminate the calibration program and issue an alarm. If the calibrated... or If the calibration fails, a "calibration error" alarm will be issued, and the system will be prohibited from entering automatic mode. This indicates that the response speed of the solenoid valve exceeds the physical possibility range. The possible reason is that the rise judgment threshold is set too low, resulting in false noise judgment. This indicates that the pipeline is too long or the water pressure is too low, exceeding the system design range.

[0114] Close delay time The determination method and The measurement method is symmetrical. Specifically, the main control chip sends a shutdown calibration pulse (setting the solenoid valve drive I / O port to low level) and simultaneously starts a timer to record the time. Continue with Frequency monitoring .when When the upward trend changes to a flat trend -- specifically determined by 5 consecutive sampling points The increment of the value relative to the arithmetic mean of its previous 10 sampling points is consistently lower than (Approximately 2.5 times the effective resolution of the sensor, slightly higher than the noise level to ensure the reliability of the judgment) -- Recording time .calculate It is stored in non-volatile memory.

[0115] The water pressure correction unit works as follows: A pressure transmitter is installed on the water source pipeline, with a range of... to Accuracy class 0.5, output to Current signal. The system every [time period]. Read the water pressure value once Pressure value The conversion method is as follows: the main control chip reads the voltage value corresponding to the current signal output by the pressure transmitter through the ADC, and calculates the current value. ( For sampling resistors, typical value Then convert it to a pressure value using the following formula:

[0116] Units are The calibrated reference water pressure is... This refers to the pressure value recorded and stored by the pressure transmitter at the calibration time. Each time the real-time water pressure value is acquired... Then, correct it according to the following formula. and :

[0117]

[0118] This corrected formula is derived from Bernoulli's equation: In steady flow of an incompressible fluid, the flow velocity in the pipe is directly proportional to the square root of the water pressure, and the water transport time is inversely proportional to the flow velocity. Therefore, the water transport time is inversely proportional to the square root of the pressure. If the water pressure change exceeds... (Right now or The system automatically triggers recalibration and issues a prompt: "Water pressure is abnormal; it is recommended to check the pipeline."

[0119] VI. Detailed Implementation of the Dynamic Threshold Generation Module The dynamic threshold generation module follows the formula Calculate dynamic trigger thresholds in real time.

[0120] The setting method will be explained using a specific application example. Taking corn planting as an example, during the jointing stage, the target soil moisture content lower limit is... Set as (Volume water content). Cross-sectional area of ​​the soil column from the lysimeter. Right now ,high Right now Soil bulk density The density of water Values Dry soil weight:

[0121]

[0122] Data is input via a human-computer interaction interface (7-inch touchscreen) and stored in ferroelectric memory (FRAM), and can be retained for more than 10 years after power failure.

[0123] The calculation process for the dynamic threshold is as follows: the dynamic threshold generation module reads from the water loss rate calculation module. Read from the delay measurement storage module Perform multiplication operations , and then with Add them together to get .

[0124] The limiting unit operates as follows: When Greater than At that time, a mandatory order ,in . The value is less than the resolution of the weighing sensor. ,ensure Always lower At least one minimum discernible increment is required to prevent the trigger threshold from exceeding the stop threshold, which could cause the system to immediately meet the stop condition after irrigation is started, resulting in high-frequency oscillations. Less than At that time, a mandatory order To prevent water loss rate When the value is zero or negative, the dynamic trigger threshold is lower than the target minimum allowable weight, causing the system to never trigger irrigation.

[0125] VII. Specific Implementation Method of the Comparison Trigger Module The baseline comparison logic of the comparison trigger module is as follows: The output cycle (in this embodiment, it is) The comparison cycle is defined as , and a comparison operation is performed once per comparison cycle. The main control chip reads the current valid weight signal from the data acquisition and preprocessing module. Read the current dynamic trigger threshold from the dynamic threshold generation module. Compare them. If If so, the triggering condition is confirmed to be met.

[0126] The anti-shake unit works as follows: The first step, when it is first detected At that time, start the image stabilization timer. In this embodiment And lock the comparison status as "suspected trigger".

[0127] The second step, in During the time period ( The system continues to use Frequency (i.e., every) Once, monitored synchronously with the ADC sampling period. and The comparison results.

[0128] The third step, when Statistics after timeout During the time period, meet The number of sampling points accounts for The proportion of all sampling points during the timing period. At sampling frequency, A total of 10 sampling points were collected. If the number of sampling points meeting the criteria reaches 8 or more (i.e., the proportion reaches...), then... If the number of sampling points that meet the conditions is less than 8, then the activation command Trigger=1 will be output. If the number of sampling points that meet the conditions is less than 8, then the activation command will not be output and the debounce timer will be reset, returning to standby mode.

[0129] Pick The basis for this is that field measurement data shows that under level 3 wind conditions, the pulse width of the lysimeter weight signal is typically between 200 and 400 ms. The image stabilization time can cover The above wind load pulses. The proportional threshold is taken as... The basis is: Sampling frequency There are a total of 10 sampling points. At least 8 points are required to meet the triggering conditions. This will eliminate false triggering caused by 1 to 2 random noise points, and will not reject real triggering due to individual abnormal points. Under real water shortage trend, all 10 sampling points usually meet the triggering conditions.

[0130] VIII. Specific Implementation Method of Irrigation Execution Module The irrigation execution module is electrically connected to the comparison trigger module, and its start-up process is as follows: When Trigger=1, the main control chip's I / O port outputs a high level. The solid-state relay is activated, closing the circuit and energizing the solenoid valve to open. The solenoid valve is a pilot-operated diaphragm valve, DN20, normally closed, with an opening response time not exceeding [time value missing]. The shutdown response time does not exceed The operating voltage is AC. The above response time is for the solenoid valve under rated operating conditions (water pressure 0.2 to...). Ambient temperature The electrical response time was measured. A drip irrigation tape was connected downstream of the valve and laid out in concentric rings on the surface of the soil column, with dripper spacing... Flow rate per orifice Work pressure .

[0131] The stop determination process is as follows: The comparison trigger module issues a shutdown command when the stop condition is met. The stop condition is:

[0132] in This is the current shut-off delay time after water pressure correction; the method for determining it is described in Part Five, Shut-off Delay Time. The determination. When the above inequality holds, the system is ahead of schedule. A shutdown command is issued (the main control chip's I / O port outputs a low level, the solid-state relay disconnects, and the solenoid valve is de-energized and closed). At this time, the weight of the soil column is [distance missing]. Still need The difference, but due to the closing delay, the weight just reaches the value when the solenoid valve is fully closed. .

[0133] The longest irrigation time protection works as follows: The system sets a longest irrigation time timer. The default value is It is stored in FRAM and can be accessed by the user through a human-machine interface. to Adjustments can be made within the specified range. Once the activation time exceeds [a certain limit]... The stopping condition has not yet been met (i.e.) Not yet reached The system forcibly closes the valve and issues an "irrigation timeout" alarm. The alarm information is displayed on the human-machine interface, and an alarm signal is simultaneously sent to the remote monitoring platform via the RS485 interface, awaiting manual inspection.

[0134] The basis for this protection is: under normal irrigation conditions, the lysimeter soil column... Replenish water to The required time is usually between 2 and 8 minutes. Approximately twice the maximum normal operating time, sufficient to cover all normal irrigation scenarios; exceeding... Not yet reached This indicates that the system may have abnormal conditions such as leakage, blockage, or insufficient water pressure.

[0135] IX. Complete Flow of Control Methods The control method of the present invention includes the following steps: Step S1: System initialization and parameter setting. After the system is powered on, the main control chip executes a self-test program to check whether the communication of each sensor is normal, whether the non-volatile memory is readable and writable, and whether the solenoid valve responds to the test pulse.

[0136] After the self-test passes, the user inputs the following parameters through the human-computer interaction interface: target minimum allowable weight (In this embodiment) Irrigation stop threshold (In this embodiment) , , corresponding to Water depth, crop type, soil type, etc.

[0137] The system automatically reads the inherent delay time stored in non-volatile memory. and If the calibration validity flag stored in the FRAM is "invalid" (either during initial installation or after a previous calibration failure), the system will automatically proceed to the delayed calibration step S0.

[0138] Step S0: Delay calibration (independent subroutine). This subroutine is executed in the following order: S0.1, the system enters calibration mode. The main control chip displays "Performing delay calibration, please do not operate" through the human-machine interface, while simultaneously checking that the soil column is in an unsaturated state (currently). Below and ), environmental wind force class.

[0139] S0.2, the comparison trigger module issues a test start command, and the main control chip simultaneously starts its internal high-precision timer TIM2 (resolution). ), record time .

[0140] S0.3, continuing with Effective weight signal output by the frequency monitoring data acquisition and preprocessing module .

[0141] S0.4, using each new sampling point as a potential starting point, performs a sliding window judgment -- when continuous Each sampling point in the sampling points The values ​​are all higher than the values ​​of the five consecutive sampling points immediately preceding them. The arithmetic mean of five consecutive sampling points, and the increment of the last sampling point relative to the arithmetic mean is greater than the rise threshold. When the calibration is complete, the time is recorded. .

[0142] S0.5, Calculation ,Will Store in FRAM and set the calibration validity flag to "valid". S0.6, Repeat the above steps three times, and take the average of the three measurements as the final value. Value storage. If the standard deviation of the three measurements exceeds the mean... If the calibration fails, the human-machine interface will display "Calibration error, please check the pipeline and valves", and the system will be prohibited from entering automatic mode.

[0143] After step S0 is completed, if the calibration is successful (the calibration validity flag is set to "valid"), the system will automatically proceed to step S1; if the calibration fails (the calibration validity flag is still "invalid"), the system will remain in calibration mode after issuing a "calibration abnormality" alarm, and will be prohibited from entering the automatic control mode until manual intervention is performed to troubleshoot the fault and re-execute the calibration procedure.

[0144] Step S2: Real-time data acquisition. The weighing sensor of the gravimetric lysimeter continuously outputs an analog weight signal, which is then converted into a digital signal by the analog-to-digital converter of the data acquisition and preprocessing module. The sampling frequency converts the analog signal into a raw weight sequence of digital quantities. .

[0145] Step S3: Signal preprocessing. The main control chip performs the following sub-steps: (1) Median filtering: for continuous Median filtering is performed on the original weight data points to output the median value sequence. ; (2) Moving average filtering: for continuous The arithmetic mean of the intermediate value sequence data points is calculated. The value is determined by Confirmed, in this embodiment hour ; (3) Temperature correction: each Read the ambient temperature sensor once, based on the temperature drift curve coefficient. right Make corrections: (For the method of obtaining the correction factor, please refer to the temperature correction unit in Part III).

[0146] Step S4: Water loss rate calculation. The water loss rate calculation module calculates the effective weight signal. Perform least squares linear regression: take the values ​​from the previous time step. Data points, , , In this embodiment hour , With time as the horizontal axis, A linear regression was performed with the vertical axis as the ordinate, and the slope of the regression line was used as the real-time water loss rate. Simultaneously obtain the leakage rate. ,like Then make corrections: If corrected Then let .

[0147] Step S5: Dynamic Threshold Generation. The dynamic threshold generation module reads the current delay time, corrected for water pressure, from the delay measurement and storage module. According to the formula Calculate the dynamic trigger threshold and perform amplitude limiting.

[0148] Step S6: Comparison and Trigger Judgment. The comparison trigger module compares... and .like Activate the image stabilization unit: Image stabilization timer ,exist Inner Frequency is continuously monitored and compared. If At least The sampling points (i.e., at least 8 out of 10 sampling points) satisfy If the output is positive, then the command Trigger=1 is output; otherwise, the debounce timer is reset and the process returns to step S2 to continue monitoring.

[0149] After confirmation by the anti-jitter unit, the start instruction Trigger=1 is stored in the register of the main control chip in the form of a Boolean signal. At the same time, a high-priority interrupt service routine is triggered, which is responsible for executing the irrigation start operation in step S7.

[0150] Step S7: Irrigation Execution and Stop Control. Upon detecting Trigger=1, the irrigation execution module initiates the irrigation start operation: opening the solenoid valve and starting irrigation water replenishment. During irrigation, the system continuously executes steps S2 to S5. The comparison trigger module determines the stop condition in real time.

[0151] If successful, a shutdown command is issued, and the irrigation execution module stops irrigation upon receiving the shutdown command. Simultaneously, if the irrigation duration exceeds [a certain limit], [further action will be taken]. If the stopping conditions are not met, the irrigation will be forcibly stopped and an "irrigation timeout" alarm will be issued.

[0152] Step S8: Return to step S2 to form a continuous closed-loop control. This return path is a loop return after completing one irrigation cycle. At this point, the system completes one full control cycle and enters the monitoring and decision-making phase of the next cycle.

[0153] 10. Triggering of recalibration conditions The delay measurement storage module automatically re-executes the delay calibration procedure when any of the following recalibration trigger conditions are met: (a) The cumulative system runtime is calculated from the last calibration completion time. A recalibration is triggered. The main control chip's internal RTC (model DS3231, accuracy...) Continuously record system runtime; when the cumulative runtime reaches [a certain value], [the system will continue to run]. Automatic recalibration is triggered at certain times. This cycle setting is based on fatigue characteristic testing of the solenoid valve spring, under typical operating conditions of approximately 20 actuations per day. After approximately 600 cumulative actions, the change in response time caused by the fatigue deformation of the spring is approximately equal to the initial value. Exceeding the allowable range of system compensation accuracy ( ).

[0154] (b) The range of change in ambient temperature is greater than The main control chip continuously monitors the ambient temperature via a temperature sensor. When the difference between the current temperature and the temperature at the last calibration exceeds [a certain threshold], [the chip will detect the temperature]. Recalibration is automatically triggered at certain times.

[0155] (c) Real-time water pressure value and reference water pressure value of irrigation water source The deviation between them is greater than The main control chip continuously monitors the water pressure via a pressure transmitter. Recalibration is automatically triggered at certain times.

[0156] (d) The irrigation pipeline components of the irrigation execution module have been replaced. When the user inputs the command "pipeline components have been replaced" through the human-machine interface, or when the system detects a solenoid valve replacement signal (achieved through the insertion and removal detection circuit of the solenoid valve connector), recalibration is automatically triggered.

[0157] In conclusion, 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 within the protection scope of the present invention.

Claims

1. A closed-loop control system for soil moisture based on lysimeter weighing, characterized in that, include: A weighing lysimeter includes a lysimeter cylinder, a weighing sensor disposed at the bottom of the cylinder, and a leakage measurement device connected to the drainage outlet at the bottom of the cylinder. The weighing lysimeter is used to output a weight signal that characterizes the weight of the soil column inside the cylinder in real time. The data acquisition and preprocessing module is electrically connected to the weighing sensor and is used to acquire the weight signal at a fixed sampling frequency, preprocess the weight signal, and output a valid weight signal. ; The water loss rate calculation module, electrically connected to the data acquisition and preprocessing module, is used to calculate the water loss rate based on the effective weight signal. The rate of change over time is used to calculate the real-time water loss rate of current soil moisture. ; The delay measurement and storage module acquires the effective weight signal through the data acquisition and preprocessing module. This is used to determine the effective weight signal after the comparison trigger module issues a test start command. The change trend is used to perform a delay calibration procedure to determine and store the inherent physical delay time between the issuance of the start command and the start of irrigation water supply to the soil column. ; The dynamic threshold generation module is electrically connected to both the water loss rate calculation module and the delay measurement and storage module, and is used to calculate the dynamic trigger threshold in real time according to the following formula: in, The dynamic trigger threshold, The preset target minimum allowable weight, The real-time water loss rate is... This refers to the inherent physical delay time; The comparison trigger module is electrically connected to the data acquisition and preprocessing module and the dynamic threshold generation module, respectively, and is used to trigger the comparison of the effective weight signal. Less than or equal to the dynamic trigger threshold The system sends an activation command to the irrigation execution module, and the effective weight signal is received. Reaching or exceeding the preset irrigation stop threshold A shutdown command is sent to the irrigation execution module at that time; The irrigation execution module is electrically connected to the comparison trigger module and is used to start irrigation water supply to the soil column after receiving the start command, and to stop irrigation water supply after receiving the stop command.

2. The system according to claim 1, characterized in that, The water loss rate calculation module uses the least squares linear regression method to calculate the real-time water loss rate. The specific calculation method is as follows: Take the time before the current moment For each effective weight signal data point, a linear regression is performed with the time corresponding to each data point as the x-axis and the effective weight signal value corresponding to each data point as the y-axis. The slope of the resulting straight line is taken as the real-time water loss rate. ; Among them, the The time window length covered by each data point is The time window The moving average filter window of the data acquisition and preprocessing module Between , It is the reciprocal of the fixed sampling frequency, and The value range is from 0.5s to 3.0s; The moving average filter window For continuous The window size for the arithmetic mean of the data points, and .

3. The system according to claim 1, characterized in that, The delay measurement and storage module executes the delay calibration program after the system is initially installed, after the irrigation pipeline is changed, or after the solenoid valve of the irrigation execution module is replaced. The delay calibration program includes the following steps executed in sequence: S1: The comparison trigger module issues a test start command, simultaneously starts timing and records the first time point. ; S2: Continuously monitor the effective weight signal output by the data acquisition and preprocessing module. ; S3: When continuous The effective weight signal value of each sampling point in each sampling point is higher than that of the continuous The sampling points were adjacent to each other. The arithmetic mean of the sampling points, and the continuous The effective weight signal value of the last sampling point in the series of sampling points is relative to the continuous The sampling points were adjacent to each other. When the increment of the arithmetic mean of the sampling points exceeds the preset rise threshold, it is determined that irrigation water has begun to replenish the soil column and cause the weight to rise. Timing is then stopped and a second time point is recorded. ; S4: Calculation The calculated result The value is stored in non-volatile memory; in, Integers between 3 and 10 The threshold value is an integer from 5 to 20, and the threshold value for determining the increase is 3 to 10 times the resolution of the weighing sensor.

4. The system according to claim 1, characterized in that, The dynamic threshold generation module further includes a limiting unit, which is used to limit the dynamic trigger threshold. The following amplitude limiting measures will be applied: When the dynamic trigger threshold Greater than the preset irrigation stop threshold At that time, the dynamic trigger threshold is forcibly set. Revised to ,in It is a preset positive constant and is smaller than the resolution of the weighing sensor; When the dynamic trigger threshold Less than the preset target minimum allowable weight At that time, the dynamic trigger threshold is forcibly set. Revised to .

5. The system according to claim 1, characterized in that, The comparison trigger module further includes a dejitter unit, which is used to perform the following dejitter confirmation operation when the comparison trigger module detects a trigger condition: S1: When the comparison trigger module first detects the valid weight signal Less than or equal to the dynamic trigger threshold At that time, the preset image stabilization timer is activated. The anti-shake timer The timing duration ranges from 0.3s to 1.0s; S2: In the anti-shake timer During the timing period, the effective weight signal is continuously monitored. With the dynamic trigger threshold The comparison results; S3: When the anti-shake timer After the timeout, the anti-shake timer is counted. During the time period, meet The number of sampling points accounts for the percentage of the anti-shake timer The proportion of all sampling points during the timing period. If the proportion is greater than the preset proportion threshold, the start command is officially output; if the proportion is less than or equal to the proportion threshold, the start command is not output and the anti-shake timer is reset. The ratio threshold is 60% to 90%.

6. The system according to claim 1, characterized in that, Upon receiving the start command, the irrigation execution module initiates irrigation and maintains the irrigation supply until it receives the stop command from the comparison trigger module, at which point it stops the irrigation supply. The comparison trigger module issues the stop command when the following stop conditions are met: in, The effective weight signal at the current moment. The real-time water loss rate at the current moment, This refers to the inherent physical shutdown delay time between the irrigation execution module receiving the shutdown command and the complete cessation of irrigation water supply to the soil column. The preset irrigation stop threshold; the inherent physical shutdown delay time. The delay is measured and stored by the delay measurement and storage module.

7. The system according to claim 2, characterized in that, The data acquisition and preprocessing module includes: An analog-to-digital conversion unit converts the analog weight signal output by the weighing lysimeter into a raw weight sequence in digital form at a sampling frequency of at least 1 Hz. The median filter unit, electrically connected to the analog-to-digital converter unit, is used to filter continuous... The original weight data points are processed by median filtering to remove impulse noise, and the intermediate value sequence is output. Odd numbers between 5 and 11; The moving average filter unit, electrically connected to the median filter unit, is used for continuous... The intermediate value sequence data points are subjected to an arithmetic mean to suppress white noise, and the effective weight signal is output. ; The temperature correction unit, electrically connected to the moving average filter unit, is used to acquire the ambient temperature and adjust the effective weight signal according to the ambient temperature. Perform temperature drift correction.

8. The system according to claim 1, characterized in that, The delay measurement and storage module also includes a water pressure correction unit, which is used to: acquire the real-time water pressure value of the irrigation water source. ; Read the calibrated inherent physical delay time stored in the delay measurement storage module. Reference water pressure value at that time Each time the real-time water pressure value is acquired... Then, the inherent physical delay time is calculated according to the following formula. Perform real-time corrections: in, To calibrate the inherent physical delay time, The current delay time is the corrected value; the dynamic threshold generation module uses the corrected current delay time. As the inherent physical delay time To calculate the dynamic trigger threshold ; When the real-time water pressure value Compared with the reference water pressure value When the deviation between the two values ​​exceeds a preset deviation threshold, the delay measurement and storage module is triggered to re-execute the delay calibration program; the deviation threshold is 15% to 25%.

9. The system according to claim 1, characterized in that, The water loss rate calculation module calculates the real-time water loss rate. At the same time, the leakage rate measured by the leakage measurement device is also acquired. ; When the leakage rate When the value is greater than zero, the water loss rate calculation module will calculate the real-time water loss rate. Revised to ,in The total weight change rate of the weighing lysimeter; When the leakage rate When the duration of a continuous positive value exceeding zero exceeds a preset time threshold, the comparison trigger module pauses issuing the start command until the leakage rate is reached. After the time threshold is reset to zero, the comparison trigger module resumes issuing the start command; the time threshold is 30 seconds to 5 minutes.

10. The system according to claim 1, characterized in that, When any of the following recalibration trigger conditions are met, the delay measurement storage module automatically re-executes the delay calibration program: The cumulative running time of the system is calculated from the time of the last calibration completion. Each time a preset periodic calibration cycle is reached, the periodic calibration cycle is 15 days to 60 days. The change in ambient temperature is greater than a preset temperature change threshold, which is 10°C to 20°C. The real-time water pressure value of the irrigation water source and the reference water pressure value The deviation between them is greater than a preset deviation threshold, which is 15% to 25%; The irrigation pipeline components of the irrigation execution module were replaced.

Images