Dynamic adjustment system for water requirement of garden plants based on intelligent monitoring

Abstract

The application relates to the technical field of garden irrigation, and particularly discloses a garden plant water demand dynamic adjustment system based on intelligent monitoring, which aims to solve the technical problems that a quantitative correlation model cannot be established among soil moisture, meteorological factors and plant transpiration demand, and that there is a lack of dynamic optimization capability for irrigation strategies. The system comprises a data acquisition unit, a demand prediction unit, a dynamic optimization unit and a partition execution unit; the data acquisition unit acquires soil humidity, meteorological, plant physiological and irrigation equipment operation data; the demand prediction unit establishes a quantitative correlation model based on an energy balance method, comprehensively analyzes in combination with a plant type and growth stage database, and calculates water demand prediction values in a future time window of each partition. The system establishes an irrigation decision closed-loop feedback mechanism, realizes adaptive adjustment and online learning optimization of parameters, and effectively improves the scientificity of irrigation and water resource utilization efficiency through quantitative analysis.

Description

Technical Field

[0001] This invention belongs to the field of agricultural engineering, specifically to the field of garden irrigation technology, and relates to a dynamic regulation system for the water demand of garden plants based on intelligent monitoring. Background Technology

[0002] The rapid development of IoT technology has brought an opportunity for intelligent transformation in modern agriculture and landscape management. Through the integration of sensor networks, data transmission, and cloud computing, real-time perception and precise control of agricultural production environments and landscape elements have been achieved. In the field of smart landscaping, plant irrigation, as a core component of landscape maintenance, is directly related to the healthy operation of the landscape ecosystem and the efficiency of water resource utilization.

[0003] Among them, the accurate monitoring and dynamic adjustment of the water demand of garden plants is one of the key technologies for the construction of smart gardens. Traditional garden irrigation methods mainly rely on manual experience or simple time-based control, which cannot comprehensively analyze multi-dimensional factors such as soil moisture, meteorological conditions, plant species and growth stages. This results in a lack of scientific basis for irrigation decisions, and the coexistence of water waste and plant water shortage.

[0004] While existing garden irrigation systems have achieved a certain degree of automation in data acquisition, they still have significant shortcomings in data fusion and analysis, demand forecasting, and dynamic adjustment. Specifically, current technologies struggle to establish quantitative correlation models between soil moisture, meteorological factors, and plant transpiration requirements, making it impossible to achieve precise irrigation based on the actual water needs of plants. Furthermore, existing systems lack the ability to dynamically optimize irrigation strategies, making it difficult to adapt to changes in climate conditions and the differentiated needs of plant growth cycles. In addition, the differences in water requirements among different regions and plant species increase the complexity of irrigation regulation, making it difficult for traditional systems to perform refined zoning management.

[0005] Therefore, there is an urgent need for a system that can intelligently monitor and dynamically adjust the water requirements of garden plants in order to achieve the goals of scientific irrigation, water conservation, and healthy growth of garden plants. Summary of the Invention

[0006] The purpose of this invention is to provide a dynamic regulation system for the water demand of garden plants based on intelligent monitoring, in order to solve the technical problems in the prior art that are unable to establish a quantitative correlation model between soil moisture, meteorological factors and plant transpiration demand, unable to achieve precise irrigation based on the actual water demand of plants, lack the ability to dynamically optimize irrigation strategies, and are difficult to adapt to the differentiated needs of climate conditions and plant growth cycles.

[0007] The technical solution of the present invention includes: The data acquisition unit is used to collect soil moisture data, meteorological data, plant physiological data, and irrigation equipment operation status data for each zone within the garden area. Soil moisture data is collected at preset time intervals using soil moisture sensors. Meteorological data includes air temperature, air humidity, wind speed, solar radiation intensity, and precipitation. Plant physiological data is collected using plant stem flow sensors and leaf surface temperature sensors. Irrigation equipment operation status data includes the opening and closing status of solenoid valves, pipeline pressure, and flow meter readings.

[0008] The demand forecasting unit is used to establish a quantitative correlation model between soil moisture, meteorological factors, and plant transpiration demand based on multidimensional data collected by the data acquisition unit. It combines the plant species database and growth stage database for comprehensive analysis to calculate the predicted water demand values ​​for each region within a preset time window. The quantitative correlation model adopts an energy balance-based evapotranspiration calculation method, which is corrected by incorporating plant physiological characteristic parameters. The plant species database stores the root depth, optimal soil moisture range, transpiration coefficient, and drought tolerance level of different plant species. The growth stage database stores the water requirement characteristics and water sensitivity coefficient of the same plant at different growth stages.

[0009] The dynamic optimization unit compares the predicted water demand values ​​for each zone output by the demand forecasting unit with real-time soil moisture data to generate a difference value. When the difference value exceeds a preset response threshold, the optimal irrigation amount, irrigation time, and irrigation duration are calculated based on the magnitude and trend of the difference value. When the difference value does not exceed the response threshold, a signal to maintain the current irrigation strategy is generated. The dynamic optimization unit also includes a strategy evaluation module, which performs retrospective analysis on the actual effects of each irrigation decision and feeds the analysis results back to the quantitative correlation model to achieve online learning and adaptive parameter adjustment of the model.

[0010] The zoning execution unit is used to generate corresponding zoning irrigation control commands in response to the optimal irrigation volume, irrigation time and irrigation duration generated by the dynamic optimization unit, and send them to the corresponding solenoid valves and irrigation equipment to realize differentiated irrigation control for each zoning. The zoning execution unit is also used to divide the garden area into several irrigation zones according to the topography and plant distribution, and configure an independent irrigation control loop for each zone.

[0011] The quantitative correlation model in the demand forecasting unit is constructed using the following method: A historical data sample set is obtained, where each sample contains actual measurements of soil moisture, meteorological factors, and plant physiological parameters; a multiple regression model is established with plant evapotranspiration as the target variable, using meteorological factors as the main input variable, soil moisture as the constraint condition, and plant physiological parameters as correction coefficients; the multiple regression model is trained using the historical data sample set to obtain model parameters; and the trained multiple regression model is deployed in the demand forecasting unit for real-time calculation of plant evapotranspiration.

[0012] The strategy evaluation module in the dynamic optimization unit adopts the following evaluation method: After each irrigation is completed, soil moisture recovery data, plant stem flow change data, and leaf temperature change data are collected within a preset time node after irrigation; an evaluation function is constructed with water use efficiency and plant health index as comprehensive evaluation indicators; the evaluation results are compared with the preset optimization target, and the comparison deviation is fed back to the parameter adjustment module of the quantitative correlation model through an adaptive algorithm to achieve iterative optimization of model parameters.

[0013] The zonal execution unit is also equipped with a priority scheduling module, which is used to determine the irrigation execution order and time allocation ratio of each zonal based on the drought tolerance level of plants, the soil moisture content warning level and the preset irrigation priority rules when the irrigation demand of multiple zonals is triggered at the same time. The drought tolerance level comes from the drought tolerance level parameters in the plant species database. The soil moisture content warning level is determined based on the deviation of real-time soil moisture data from the optimal soil moisture range of each plant.

[0014] The data acquisition unit is also equipped with a data quality verification module, which is used to detect outliers and handle missing values ​​in the acquired raw data; when continuous anomalies are detected in the sensor data, the sensor self-test program is automatically triggered and alarm information is generated; when missing data is detected, a time series-based interpolation method is used to complete the data.

[0015] The system also includes a user interaction terminal, which is used to visually display soil moisture distribution maps, meteorological change curves, water demand forecasts, and irrigation execution records for each zone; the user interaction terminal also supports users to modify preset parameters and manually intervene in irrigation strategies.

[0016] Compared with the prior art, the advantages and positive effects of the present invention are as follows: This invention establishes a quantitative correlation model between soil moisture, meteorological factors, and plant transpiration requirements, upgrading the calculation of actual plant water requirements from empirical judgment to quantitative analysis based on energy balance and plant physiological characteristics, thus achieving a technological leap from qualitative irrigation to precision irrigation. This quantitative correlation model fully considers the root characteristics, transpiration characteristics, and drought tolerance of different plant species, and dynamically corrects them based on the water requirements of different growth stages, making water requirement predictions more consistent with the actual physiological state of plants. This effectively avoids the problems of over-irrigation or under-irrigation caused by empirical estimation in traditional irrigation methods.

[0017] This invention establishes a closed-loop feedback mechanism for irrigation decision-making through the strategy evaluation module of the dynamic optimization unit. The actual effect data after each irrigation is fed back to the quantitative correlation model for adaptive parameter adjustment, enabling the system to have the ability to learn online and continuously optimize. As the system runs for a long time, the prediction accuracy of the quantitative correlation model will continue to improve, and the irrigation strategy will continuously approach the optimal state, realizing the intelligent evolution of the irrigation system.

[0018] This invention achieves differentiated and precise irrigation for different plant species within a garden area through the collaborative work of a partitioned execution unit and a priority scheduling module. The system can automatically adjust irrigation priorities based on the drought tolerance level of plants and the soil moisture content warning level of each partition, ensuring that high water demand areas and sensitive plants are given priority protection, while moderately controlling water for low water demand areas and drought-resistant plants, effectively improving the scientific nature of overall irrigation and water resource utilization efficiency.

[0019] This invention ensures the reliability and integrity of input data through the outlier detection and missing value handling functions of the data quality verification module, providing a data foundation for the accurate calculation of quantitative correlation models and the scientific formulation of irrigation decisions. The introduction of sensor self-checking and alarm mechanisms enables the system to detect equipment failures in a timely manner and take countermeasures, effectively reducing the risk of irrigation errors caused by equipment problems. Attached Figure Description

[0020] Figure 1 This is a schematic diagram of the overall technical solution architecture proposed in this invention; Figure 2 This is a schematic diagram of the core principle framework of the quantitative correlation model between soil moisture, meteorological factors and plant transpiration demand in this invention; Figure 3 This is a logical flow diagram of the data acquisition and demand forecasting phase in this invention; Figure 4 This is a schematic diagram of the multi-level interaction relationship and data flow between dynamic optimization and partitioned execution in this invention; Figure 5 This is a schematic diagram of the closed-loop feedback mechanism and parameter adaptive adjustment principle of the strategy evaluation module in this invention. Detailed Implementation

[0021] Example 1 Please refer to the attached document. Figure 1 The overall technical architecture of the intelligent monitoring-based dynamic regulation system for water demand of garden plants includes four core modules: data acquisition unit, demand prediction unit, dynamic optimization unit, and zone execution unit. The modules interact and coordinate control through standardized data interfaces.

[0022] The data acquisition unit, serving as the perception layer of the entire system, is responsible for acquiring multi-dimensional environmental data and equipment status data from various zones within the garden area. This unit consists of soil moisture sensors, meteorological monitoring equipment, plant physiological sensors, and irrigation equipment status acquisition modules. Each sensor and acquisition module performs data acquisition tasks at preset time intervals.

[0023] Soil moisture sensors are deployed in the root zone of each irrigation area, with the sampling depth set according to the root distribution characteristics of different plant species. For shallow-rooted plants such as turfgrass, the sensor sampling depth is set to 15 to 20 cm below the soil surface; for shrubs with medium root depth, the sampling depth is set to 30 to 40 cm below the soil surface; and for deep-rooted trees, the sampling depth is set to 60 to 80 cm below the soil surface. The soil moisture sensors use frequency domain reflectometry, with a measurement accuracy of ±2% volumetric water content, and a sampling frequency of once every 15 minutes. The raw data collected by the sensors include volumetric water content percentage, soil temperature, and soil electrical conductivity.

[0024] Meteorological monitoring equipment was deployed at representative locations within the park area, collecting meteorological data including air temperature, air humidity, wind speed, solar radiation intensity, and precipitation. The air temperature sensor measures from -40°C to +80°C with an accuracy of ±0.3°C; the air humidity sensor measures from 0% to 100% relative humidity with an accuracy of ±2% relative humidity; the wind speed sensor measures from 0 to 60 meters per second with an accuracy of ±0.3 meters per second; the solar radiation sensor measures from 0 to 2000 watts per square meter with an accuracy of ±5 watts per square meter; and the precipitation sensor has an accuracy of ±0.2 millimeters. Meteorological data was collected every 15 minutes, synchronized with soil moisture data.

[0025] The plant physiological sensors include a stem flow sensor and a leaf surface temperature sensor. The stem flow sensor uses the principle of heat diffusion to measure the sap flow rate in the plant stem, with a measurement range of 0 to 1000 grams per hour and an accuracy of ±5%. The leaf surface temperature sensor uses infrared thermometry, with a measurement range of -10 degrees Celsius to +70 degrees Celsius and an accuracy of ±0.2 degrees Celsius. The plant physiological sensors collect data every 30 minutes, with the data collection timing aligned with the hourly times of the meteorological data collection window.

[0026] The irrigation equipment status acquisition module collects data including the opening and closing status of solenoid valves, pipeline pressure, and flow meter readings. Solenoid valve status is obtained through valve position feedback signals, with the opening and closing status represented by binary values. Pipeline pressure is measured using pressure sensors, with a measurement range of 0 to 1.0 MPa and an accuracy of ±0.01 MPa. Electromagnetic flow meters are used, with a measurement range of 0 to 100 cubic meters per hour and an accuracy of ±0.5% of the measured value. Irrigation equipment status data is collected every 60 seconds, reflecting the equipment's operating status in real time.

[0027] The data acquisition unit is also equipped with a data quality verification module, which performs outlier detection and missing value handling on the acquired raw data. Outlier detection uses the three-standard-deviation criterion from statistical methods. When the values ​​collected by a single sensor within three consecutive acquisition cycles deviate from the sensor's historical data mean by more than three standard deviations, it is considered outlier data. Upon the generation of outlier data, the system automatically triggers a sensor self-test program, which includes sensor power supply voltage detection, signal line integrity detection, and sensor zero-point calibration verification. The self-test results generate alarm information and are recorded in the system log. Missing value handling uses a time-series-based linear interpolation method. When a sensor detects a missing value at a certain acquisition time, the system performs linear interpolation calculations based on the acquisition values ​​from two time points before and after the missing value. The interpolation formula is:

[0028] in This is the interpolation result for the missing time step. This is the value collected at the second time point before the missing time. This is the value collected at the second time point after the missing time. For the missing timestamp, and These are the timestamps for the two points in time, one before and one after.

[0029] Valid data processed by the data quality verification module is encapsulated in a unified data format, which includes five fields: partition number, sensor type code, acquisition timestamp, data value, and verification flag. The encapsulated data is then transmitted to the demand forecasting unit via the internal communication network.

[0030] The demand forecasting unit is used to establish a quantitative correlation model between soil moisture, meteorological factors and plant transpiration demand, and to perform comprehensive analysis by combining plant species database and growth stage database to calculate the predicted water demand value of each region in the future preset time window.

[0031] Please refer to the attached document. Figure 2 The core framework of the quantitative correlation model between soil moisture, meteorological factors, and plant transpiration demand is based on energy balance theory. Plant evapotranspiration consists of two parts: plant transpiration and soil surface evaporation. The model uses a modified form of the Penman formula to calculate evapotranspiration. The construction method of the quantitative correlation model includes the following steps: The first step is to obtain a historical data sample set, which contains measured data on soil moisture, meteorological factors, and plant physiological parameters accumulated over the past year. Each historical data sample includes eight parameters: soil volumetric water content, air temperature, air humidity, wind speed, solar radiation intensity, precipitation, plant stem flow rate, and leaf surface temperature. The total number of samples in the historical data sample set is no less than 8,000, and the sample time span covers all four seasons and various weather conditions such as sunny, rainy, and foggy weather.

[0032] The second step is to establish a multiple regression model with plant evapotranspiration as the target variable. The model input variables are divided into three categories: the first category consists of the main meteorological input variables, including net radiation, air temperature difference, air humidity difference, and wind speed difference; the second category consists of soil moisture constraint variables, including soil volumetric water content and available soil water content; and the third category consists of plant physiological correction coefficients, including plant stem flow correction coefficient and leaf surface temperature correction coefficient. The target variable is plant evapotranspiration, measured in millimeters per day.

[0033] The third step is the model training process, which uses the least squares method to calibrate the parameters of the multiple regression model. During training, principal component analysis is performed on the meteorological factor input variables to extract principal components and eliminate multicollinearity among variables. After the model training is completed, the regression coefficients are stored in the model parameter library, which uses a hash table structure. The key is a combination of plant species number and growth stage number, and the value is the corresponding regression coefficient matrix.

[0034] The fourth step is model deployment, where the trained multivariate regression model is deployed in the inference engine of the demand forecasting unit. After receiving the real-time collected multidimensional data, the inference engine first queries the corresponding plant species number and current growth stage number based on the partition number, then loads the corresponding regression coefficient matrix from the model parameter library, and finally substitutes the real-time data into the regression equation to calculate the predicted plant evapotranspiration value.

[0035] The plant species database stores physiological characteristic parameters of different plant species. The database table structure includes six fields: plant species ID, plant name, root depth, optimal soil moisture range, transpiration coefficient, and drought tolerance level. The root depth field stores numerical data in centimeters; the optimal soil moisture range stores two values, a lower limit and an upper limit, in volume percentage; the transpiration coefficient stores a dimensionless value between 0 and 1; and the drought tolerance level stores integer values ​​from 1 to 5, where 1 represents extremely drought-tolerant and 5 represents extremely intolerant of drought. The plant species database pre-loads parameter data for 156 common garden plants, and users can add new plant species through the user interface terminal.

[0036] The growth stage database stores the water requirement characteristics and water sensitivity coefficients of the same plant at different growth stages. The database table structure includes five fields: plant species ID, growth stage ID, growth stage name, daily water requirement, and water sensitivity coefficient. The growth stages are divided into five phases: budding, vegetative growth, flowering, fruiting, and dormancy. Daily water requirement is expressed in millimeters per day, and the water sensitivity coefficient is a dimensionless value between 0 and 2; a higher value indicates greater sensitivity to water deficiency at that stage. The initial data for the growth stage database comes from agricultural research literature, and it is dynamically corrected during system operation based on feedback data from the strategy evaluation module.

[0037] The water demand forecast output by the demand forecasting unit includes forecasts for four time windows: the next 3 hours, the next 6 hours, the next 12 hours, and the next 24 hours. Each time window's forecast includes four fields: zone number, start and end times of the forecast window, daily water demand forecast, and forecast confidence level. The forecast confidence level is calculated using the root mean square error (RMSE), and the confidence level ranges from 0 to 1; a value closer to 1 indicates a more reliable forecast.

[0038] The dynamic optimization unit is used to compare the predicted water demand values ​​of each zone with the real-time soil moisture data, generate difference values, and calculate the optimal irrigation amount, irrigation time, and irrigation duration based on the difference values.

[0039] Please refer to the attached document. Figure 3The logical flow of the data acquisition and demand forecasting phase is as follows: The data acquisition unit continuously acquires multidimensional data according to the preset acquisition cycle, and the data is stored in the time series database after quality verification; the demand forecasting unit extracts the continuous data of the most recent 24 hours from the time series database every 60 minutes as a calculation cycle, and inputs it into the quantitative correlation model to perform evapotranspiration forecasting calculation; the calculated forecast value, together with the partition number and timestamp, is packaged and sent to the dynamic optimization unit.

[0040] After receiving the water demand forecast from the demand forecasting unit, the dynamic optimization unit executes the difference calculation process. The difference calculation formula is:

[0041] in For the difference value, This refers to the daily water demand forecast output by the demand forecasting unit. This represents the current soil moisture content. It refers to the soil field water holding capacity. This represents the current percentage of soil volumetric water content. The physical meaning of the difference value is the difference between the plant's daily water requirement and the soil's current available water.

[0042] The dynamic optimization unit is equipped with a response threshold determination module, and the response threshold is configured differently according to plant species and growth stage. The default value of the response threshold is set to 2 mm. For highly sensitive plants with a water sensitivity coefficient greater than 1.5, the response threshold is lowered to 1 mm; for low-sensitive plants with a water sensitivity coefficient less than 0.8, the response threshold is raised to 3 mm. When the difference value exceeds the response threshold, the irrigation decision generation process is triggered; when the difference value does not exceed the response threshold, a signal to maintain the current irrigation strategy is generated. This signal includes a partition number and a maintenance identifier.

[0043] The irrigation decision generation process consists of three sub-steps. The first step is calculating the optimal irrigation volume. The theoretical irrigation volume is determined based on the magnitude of the difference in water usage. This theoretical volume equals the difference multiplied by the area of ​​the zone, then divided by the effective utilization rate of the irrigation system. The effective utilization rate is determined based on the pipe network pressure and sprinkler type, with a default value of 0.85. The theoretical irrigation volume needs to be verified by soil saturation. If the calculated irrigation volume would cause the soil moisture content to exceed field capacity, it is automatically corrected to the amount of water needed to reach field capacity. The second step is determining the optimal irrigation time. Irrigation time selection follows the principle of avoiding high-temperature periods and periods of high wind speed, prioritizing the period from 4:00 AM to 6:00 AM. If this period is unavailable for other reasons, it is postponed to 6:00 PM to 8:00 PM. The third step is calculating the optimal irrigation duration. The irrigation duration equals the optimal irrigation volume divided by the pipe network design flow velocity, then divided by the area of ​​the zone. When the pipe network flow velocity is affected by water usage in other zones, the system automatically calls the pipe network hydraulic model for correction calculations.

[0044] The dynamic optimization unit also includes a strategy evaluation module, which is used to retrospectively analyze the actual effects of each irrigation decision.

[0045] Please refer to the attached document. Figure 5 The closed-loop feedback mechanism and parameter adaptive adjustment principle of the strategy evaluation module are as follows: After each irrigation is completed, the system starts an effect evaluation timer, which is set to three detection nodes: 2 hours, 6 hours, and 12 hours after irrigation. At each detection node, the system collects soil moisture recovery data, plant stem flow change data, and leaf temperature change data. Soil moisture recovery data reflects the recovery of soil moisture content after irrigation, plant stem flow change data reflects the recovery status of water transport within the plant after irrigation, and leaf temperature change data reflects the degree to which the plant leaf temperature has returned to normal levels after irrigation.

[0046] The strategy evaluation module constructs a comprehensive evaluation function, which uses water use efficiency and plant health index as evaluation indicators. The formula for calculating water use efficiency is:

[0047] in For water use efficiency This refers to the increase in plant biomass or ornamental value per unit area. This represents the actual evapotranspiration. The plant health index is calculated by weighting the plant stem flow recovery rate and leaf surface temperature recovery rate, with a weighting ratio of 0.6 for stem flow recovery rate and 0.4 for leaf surface temperature recovery rate.

[0048] The evaluation results are compared with the preset optimization objective, and the comparison deviation is fed back to the parameter adjustment module of the quantitative correlation model through an adaptive algorithm. The adaptive algorithm uses an improved recursive least squares method. The algorithm fine-tunes the model regression coefficients based on the deviation value of each evaluation, and the fine-tuning step size is determined according to the deviation magnitude. When the deviation magnitude exceeds a preset large deviation threshold, a larger parameter adjustment is performed; when the deviation magnitude is within the preset small deviation threshold range, a smaller parameter adjustment is performed. After the parameter adjustment is completed, the new parameters replace the original parameters and are stored in the model parameter library, realizing online learning and continuous optimization of the model.

[0049] The zone execution unit is used to generate corresponding zone irrigation control commands in response to the optimal irrigation amount, irrigation time and irrigation duration generated by the dynamic optimization unit, and to send them to the corresponding solenoid valves and irrigation equipment.

[0050] Please refer to the attached document. Figure 4The multi-level interaction and data flow between dynamic optimization and partitioned execution are as follows: The irrigation decision generated by the dynamic optimization unit includes four fields: partition number, irrigation amount, irrigation start time, and irrigation duration. This decision is packaged in JSON format and sent to the partitioned execution unit. After receiving the irrigation decision, the partitioned execution unit first performs a readiness check on the irrigation equipment. The check items include the power supply status of the solenoid valve, whether the pipeline pressure is within the normal range, and whether the flow meter is in a measurable state.

[0051] The zonal execution unit is also equipped with a priority scheduling module, which functions when irrigation demands from multiple zonals are triggered simultaneously. When irrigation decisions for two or more zonals are generated within the same time window, the priority scheduling module determines the irrigation execution order and time allocation ratio for each zonal based on the drought tolerance level of plants, the soil moisture content warning level, and preset irrigation priority rules. The drought tolerance level is derived from drought tolerance parameters in the plant species database; a higher value indicates less drought tolerance and a higher priority. The soil moisture content warning level is determined based on the deviation between real-time soil moisture data and the optimal soil moisture range for each plant. A deviation exceeding 20% ​​is a high-level warning, a deviation between 10% and 20% is a medium-level warning, and a deviation below 10% is a low-level warning. The irrigation priority rules are configured as a three-level rule: the first priority is for zonals with high warning levels, the second priority is for zonals with high drought tolerance levels, and the third priority is sorted by predicted water demand from largest to smallest.

[0052] The zoning unit divides the garden area into several irrigation zones based on topography and plant distribution. Each zone is equipped with an independent irrigation control loop, including solenoid valves, pressure regulators, flow meters, and sprinklers or drip irrigation devices. The zoning principle is to group areas with similar topography and plant species into the same zone. The zone area is determined based on the water supply capacity of the pipeline network and the coverage capacity of the sprinklers, generally controlled between 500 square meters and 2000 square meters.

[0053] The zonal irrigation control command generated by the zonal execution unit includes six fields: zonal number, solenoid valve number, irrigation volume, irrigation start time, irrigation duration, and irrigation mode. Irrigation modes are divided into three types: micro-sprinkler, drip irrigation, and atomization, with different modes corresponding to different plant species and growth stages. The control command is sent to the solenoid valves and irrigation equipment in the corresponding zonal area through the irrigation control system. After receiving the valve opening command, the solenoid valve starts the motor to drive the valve core, with the opening time controlled within 3 seconds. The flow meter monitors the actual irrigation flow in real time and sends the data back to the system for irrigation volume verification.

[0054] The user interface terminal is the system's human-computer interface, used to visually display soil moisture distribution maps, meteorological change curves, water demand forecasts, and irrigation execution records for each zone. The soil moisture distribution map is displayed as a heatmap, with the legend color ranging from blue to red to indicate increasing soil moisture content. The meteorological change curve is displayed as a line graph, with the horizontal axis representing time and the vertical axes showing four curves: air temperature, air humidity, wind speed, and solar radiation intensity. The water demand forecast is displayed as a bar chart, with the horizontal axis representing the time window and the vertical axis representing the predicted water demand. The irrigation execution records are displayed in a table format, with fields including execution time, zone name, irrigation amount, irrigation duration, and operator.

[0055] The user interface terminal allows users to modify preset system parameters, including response thresholds, irrigation priority rules, irrigation schedules for each zone, and physiological parameters for each plant species. The terminal also supports manual intervention in irrigation strategies, allowing users to manually trigger irrigation for a specific zone or forcibly interrupt ongoing irrigation at any time. Manual intervention requires secondary confirmation; the confirmation interface displays the expected results of the intervention, and the operation takes effect upon user confirmation.

[0056] The overall system operation flow is as follows: The data acquisition unit continuously collects multidimensional data and stores it in the time series database after quality verification; the demand forecasting unit starts a forecast calculation every 60 minutes, extracts data from the time series database, inputs it into the quantitative correlation model, and outputs the predicted water demand values ​​for each future time window; the dynamic optimization unit compares the predicted demand values ​​with real-time soil moisture, generates a difference value, compares it with the response threshold, and generates irrigation decisions or maintenance signals based on the comparison results; the partition execution unit receives the irrigation decision, performs equipment readiness checks and priority scheduling, generates control commands, and issues them for execution; the strategy evaluation module collects effect data at each detection node after irrigation is completed, calculates the comprehensive evaluation results, and feeds back the deviation to the demand forecasting unit for adaptive adjustment of model parameters; the user interaction terminal displays the system's operating status in real time and accepts user parameter modifications and manual intervention.

[0057] The system's communication network adopts a hybrid architecture combining industrial Ethernet and wireless sensor networks. Various sensors in the data acquisition unit aggregate data to the area gateway via the wireless sensor network, and the area gateway then uploads the data to the system server via industrial Ethernet. The solenoid valves and irrigation equipment in the zone execution unit are connected to the system server via a wired control network, and control commands are issued from the server to the devices. The system server employs a dual-machine hot standby architecture to ensure high system reliability.

[0058] The system's power supply configuration employs a complementary approach of municipal power and solar power. The sensors and area gateways in the data acquisition unit utilize solar power in conjunction with battery banks, ensuring continuous data acquisition for over 72 hours even in the event of a municipal power outage. The solenoid valves and irrigation equipment in the zone execution unit utilize municipal power to ensure reliable control operations.

[0059] Example 2 Based on Example 1, this example provides a system enhancement scheme for extreme climate conditions. This scheme adds a climate adaptation module and an emergency irrigation module to the original system architecture.

[0060] The climate adaptation module is used to perform climate-adaptive corrections to the quantitative correlation model under abnormal climate conditions such as extreme high temperatures, extreme low temperatures, persistent strong winds, and torrential rain. The module incorporates a climate type identification algorithm, which compares real-time meteorological data with historical averages to identify the current climate state. The criteria for extreme high temperatures are: temperatures exceeding 38 degrees Celsius for three consecutive hours; extreme low temperatures are: temperatures below -5 degrees Celsius for three consecutive hours; persistent strong winds are: wind speeds exceeding 15 meters per second for two consecutive hours; and torrential rain is: precipitation exceeding 50 millimeters per hour.

[0061] When the climate adaptation module identifies extreme climate conditions, it automatically calls the corresponding correction parameter set to correct the parameters of the quantitative correlation model. The correction parameter set for extreme high temperature conditions includes a transpiration coefficient correction factor of 1.2, adjustment of soil moisture stress coefficient, and an increase of 5 degrees Celsius in the leaf temperature threshold; the correction parameter set for extreme low temperature conditions includes a transpiration coefficient correction factor of 0.3, a decrease of 5 degrees Celsius in the transpiration cessation calculation temperature threshold, and the introduction of a freezing coefficient for irrigation water demand calculation; the correction parameter set for continuous strong wind conditions includes a wind speed correction factor of 1.5, an evaporation correction factor of 2.0, and an irrigation water compensation factor of 1.3; the correction parameter set for heavy rainfall conditions includes rainfall interception calculation, soil infiltration rate correction, and irrigation delay time calculation.

[0062] The emergency irrigation module maintains basic irrigation functions in the event of a main system controller failure or communication interruption. It consists of an emergency controller, an emergency power supply, and emergency solenoid valves, independent of the system server. The emergency controller has a built-in simplified irrigation control program that controls the opening and closing of solenoid valves in the corresponding zones based on direct signals from soil moisture sensors. The soil moisture threshold of the emergency controller is set to a fixed value; irrigation automatically starts when the soil moisture content of a zone falls below 20% by volume, and automatically shuts off when the soil moisture content reaches 35% by volume. The emergency power supply is a battery pack with a capacity sufficient for 48 hours of continuous operation.

[0063] The introduction of climate adaptation and emergency irrigation modules further enhances the system's adaptability to complex environmental conditions, ensuring that basic water supply for garden plants can still be guaranteed even in extreme weather conditions and equipment failures.

[0064] Example 3 Based on Example 1, this example provides a differentiated configuration scheme for different garden functional areas. This scheme adds a garden functional area division subsystem and a multi-mode irrigation strategy library to the system architecture.

[0065] The garden functional zoning subsystem divides the garden area into five functional types based on its usage and landscape characteristics: viewing area, sports area, leisure area, management area, and ecological conservation area. The viewing area, primarily composed of flowers and exquisite landscape plants, has the highest water requirement precision; the sports area, mainly composed of lawns, has a high water frequency but relatively stable individual water demand; the leisure area, mainly composed of a combination of trees and shrubs, has a moderate water requirement; the management area, mainly for maintenance work, requires irrigation to be coordinated with maintenance time windows; and the ecological conservation area, mainly composed of native and drought-tolerant plants, has the lowest water requirement.

[0066] The multi-mode irrigation strategy library stores irrigation strategy templates for different functional zones and plant species. Each strategy template includes the selection of irrigation mode, configuration of irrigation time period, calculation coefficient of irrigation volume, and setting of response threshold. The strategy template for the viewing area is configured as micro-sprinkler mode, daily morning irrigation, irrigation volume calculation coefficient of 1.2, and response threshold of 1 mm; the strategy template for the sports area is configured as misting mode, daily evening irrigation, irrigation volume calculation coefficient of 1.0, and response threshold of 2 mm; the strategy template for the leisure area is configured as drip irrigation mode, irrigation every other night, irrigation volume calculation coefficient of 0.9, and response threshold of 3 mm; the strategy template for the management area is configured as micro-sprinkler mode, irrigation after maintenance work, irrigation volume calculation coefficient of 1.1, and response threshold of 2 mm; and the strategy template for the ecological conservation area is configured as drip irrigation mode, weekly irrigation, irrigation volume calculation coefficient of 0.7, and response threshold of 5 mm.

[0067] The introduction of a multi-mode irrigation strategy library enables the system to provide targeted irrigation strategies based on the different functional characteristics of the garden area, further enhancing the system's intelligence level and applicability.

Claims

1. A dynamic regulation system for the water requirements of garden plants based on intelligent monitoring, characterized in that, include: The data acquisition unit is used to collect soil moisture data, meteorological data, plant physiological data, and irrigation equipment operation status data for each zone within the garden area. The demand forecasting unit is used to establish a quantitative correlation model between soil moisture, meteorological factors and plant transpiration demand based on the multidimensional data, and to perform comprehensive analysis in combination with plant species database and growth stage database to calculate the predicted water demand value of each zone in the future preset time window. The dynamic optimization unit is used to compare the predicted water demand with real-time soil moisture data to generate a difference value, and calculate the optimal irrigation amount, irrigation time and irrigation duration based on whether the difference value exceeds a preset response threshold, or generate a signal to maintain the current irrigation strategy. The dynamic optimization unit also includes a strategy evaluation module, which is used to perform retrospective analysis on the actual effect after each irrigation and feed the results back to the quantitative correlation model to achieve adaptive parameter adjustment. The zoning execution unit is used to generate zoning irrigation control instructions based on the optimal irrigation volume, irrigation time, and irrigation duration, and send them to the corresponding solenoid valves and irrigation equipment to achieve differentiated irrigation control for each zoning zone. The zoning execution unit is also used to divide the garden area into several irrigation zones according to the topography and plant distribution, and to configure an independent irrigation control loop for each zone.

2. The dynamic regulation system for water demand of garden plants based on intelligent monitoring according to claim 1, characterized in that, The demand forecasting unit includes: The historical data sample acquisition module is used to acquire a historical data sample set containing soil moisture, meteorological factors, and plant physiological parameters; The multiple regression modeling module is used to establish a multiple regression model with plant evapotranspiration as the target variable, taking meteorological factors as the main input variable, soil moisture as the constraint condition, and plant physiological parameters as correction coefficients. The model training and deployment module is used to train the multivariate regression model using the historical data sample set to obtain model parameters, and to deploy the trained model in the demand forecasting unit for real-time calculation of plant evapotranspiration.

3. The dynamic regulation system for water demand of garden plants based on intelligent monitoring according to claim 2, characterized in that, The multivariate regression modeling module includes: The meteorological feature extraction submodule is used to extract net radiation, air temperature difference, air humidity difference, and wind speed difference from meteorological data as input variables. The soil moisture constraint submodule is used to use soil volumetric water content and soil available water content as model constraints. The physiological correction factor submodule is used to generate plant physiological correction coefficients based on plant stem flow rate and leaf surface temperature.

4. The dynamic regulation system for water demand of garden plants based on intelligent monitoring according to claim 1, characterized in that, The strategy evaluation module in the dynamic optimization unit includes: The effect data acquisition submodule is used to collect soil moisture recovery data, plant stem flow change data, and leaf temperature change data at multiple preset time nodes after irrigation is completed. The comprehensive evaluation function construction submodule is used to construct evaluation functions based on water use efficiency and plant health index. The parameter feedback adjustment submodule is used to feed back the deviation between the evaluation results and the preset optimization target to the parameter adjustment module of the quantitative correlation model through an adaptive algorithm, so as to achieve iterative optimization of the model parameters.

5. The dynamic regulation system for water demand of garden plants based on intelligent monitoring according to claim 1, characterized in that, The partition execution unit includes: The irrigation zone division module is used to divide the garden area into multiple irrigation zones based on similar topography and plant species. An independent control loop configuration module is used to configure an independent control loop for each irrigation zone, including a solenoid valve, pressure regulator, flow meter, and irrigation terminal. The control instruction generation module is used to generate zonal irrigation control instructions that include zonal number, solenoid valve number, irrigation volume, irrigation start time, irrigation duration, and irrigation mode.

6. The dynamic regulation system for water demand of garden plants based on intelligent monitoring according to claim 1, characterized in that, The partition execution unit also includes a priority scheduling module, which is used to determine the irrigation execution order and time allocation ratio of each partition according to the drought tolerance level of plants in each partition, the soil moisture content warning level and the preset irrigation priority rules when the irrigation demand of multiple partitions is triggered at the same time.

7. The dynamic regulation system for water demand of garden plants based on intelligent monitoring according to claim 6, characterized in that, The priority scheduling module includes: The drought tolerance level query submodule is used to obtain drought tolerance level parameters of plants in each region from the plant species database; The soil early warning level determination submodule is used to determine the soil moisture content early warning level based on the degree of deviation between real-time soil moisture data and the optimal soil moisture range for each plant. The priority sorting submodule is used to determine the irrigation execution order based on a three-level rule: priority for high early warning level, priority for high drought sensitivity, and priority for water demand forecast value.

8. The dynamic regulation system for water demand of garden plants based on intelligent monitoring according to claim 1, characterized in that, The data acquisition unit also includes a data quality verification module, which is used to detect outliers and handle missing values ​​in the acquired raw data.

9. The dynamic regulation system for water demand of garden plants based on intelligent monitoring according to claim 8, characterized in that, The data quality verification module includes: The outlier detection submodule is used to identify continuous abnormal data using statistical methods and trigger sensor self-test procedures. The missing value imputation submodule is used to complete the data using a time series-based interpolation method when missing data is detected. The alarm information generation submodule is used to generate alarm information when the sensor self-test fails or the data is continuously abnormal.

10. The dynamic regulation system for water demand of garden plants based on intelligent monitoring according to claim 1, characterized in that, It also includes a user interaction terminal, which is used to display soil moisture distribution maps, meteorological change curves, water demand forecasts and irrigation execution records for each zone in a visual manner, and supports users to modify preset parameters and manually intervene in irrigation strategies.

Images