Urban green land intelligent monitoring system and method based on Internet of Things

Abstract

The invention discloses an urban green land intelligent monitoring system and method based on the Internet of Things, and relates to the field of green land monitoring, and the method comprises the steps: dividing a plant growth cycle through setting time windows, collecting growth data of each window to generate a record table, and carrying out the comparison and extraction of difference parameters; on the basis of the difference parameters, screening positive correlation data as evaluation indexes, calculating growth evaluation scores after standardizing the evaluation indexes, determining a difference threshold by combining historical data, and dividing growth stages according to the difference threshold; acquiring data through a weather station, calculating evapotranspiration, calculating water demand by combining crop coefficients, and drawing a time-varying curve of the water demand; after curve data are preprocessed, a polynomial algorithm is used for curve fitting, and a water demand prediction model is constructed; in combination with real-time weather and growth stages, the irrigation amount is determined according to the water demand predicted by the model, the result is evaluated after irrigation, the model is corrected if the result is unqualified, and urban green land plant growth monitoring and precise irrigation management are achieved.

Description

Technical Field

[0001] This invention relates to the field of green space monitoring, specifically an intelligent monitoring system and method for urban green spaces based on the Internet of Things. Background Technology

[0002] In the field of urban green space monitoring, accurate determination of plant growth stages and on-demand irrigation are core aspects of ensuring healthy plant growth. In traditional plant growth management, the division of growth stages largely relies on manual observation of plant morphological characteristics. This method is highly subjective, lacks quantification, makes it difficult to systematically compare growth differences under different time windows, and fails to establish a standardized "growth stage-assessment indicator" correlation system. This results in significant errors in stage determination, which in turn affects the accuracy of subsequent management decisions.

[0003] In terms of water demand calculation and irrigation decision-making, existing technologies have obvious limitations: First, they do not fully incorporate the dynamic changes in crop coefficients at different growth stages, resulting in a disconnect between water demand calculations and actual plant physiological needs; second, the application of meteorological data is mostly limited to historical static data, lacking integration with real-time meteorological conditions and future meteorological trends, leading to water demand estimations that cannot match dynamic environmental changes; and third, the data preprocessing process is not standardized, lacking a unified and scientific identification and processing scheme for outliers and missing values ​​in water demand data, directly affecting subsequent data quality and model accuracy.

[0004] Furthermore, traditional irrigation volume determination only considers the water demand in the current period, without taking into account the predicted water demand for the next period and natural precipitation. This easily leads to over- or under-irrigation, making it difficult to achieve precise and efficient plant water management. To address the problems of inaccurate growth stage determination, large deviations between water demand calculation and prediction, and lack of dynamic feedback in irrigation decisions in the existing technologies, there is an urgent need for a systematic technical solution that can integrate time window division, quantitative growth assessment, dynamic water demand calculation, and iterative optimization of prediction models to achieve standardized, precise, and intelligent plant growth management. Summary of the Invention

[0005] The purpose of this invention is to provide an intelligent monitoring system and method for urban green spaces based on the Internet of Things, so as to solve the problems raised in the prior art.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a smart monitoring method for urban green spaces based on the Internet of Things, the method comprising the following steps: Step S1: Set a time window, collect growth data of the same plant at different time periods, generate growth record tables for each time period, and compare the growth record tables for different time periods. Step S1-1: Set a time window t to divide the plant's growth cycle into multiple consecutive time windows; Step S1-2: Record the plant growth data within each time window and generate a plant growth record table for the corresponding time window; Steps S1-3: Sort the plant growth record tables for each time window according to the time series, traverse and compare the plant growth record tables of two adjacent time windows, define "data names are the same but values ​​are different" as difference parameters, and extract the difference parameters.

[0007] By setting time windows to divide the growth cycle, recording growth data, and comparing and extracting difference parameters, we can obtain the changes in plant growth data in different consecutive time periods, providing data support for the determination of subsequent growth assessment indicators.

[0008] Step S2: Use the difference parameters obtained from the comparison as growth assessment indicators, calculate the growth assessment score, and determine the plant growth stage based on the growth assessment score. Step S2-1: Integrate the differential parameters and construct a basic plant growth dataset; Step S2-2: Using the increase in plant height as the baseline variable Y, extract all data from the plant growth dataset as candidate data, denoted as X. Select synchronous observation data from n plants to form sample pairs (X1, Y1), (X2, Y2), ..., (X... n ,Y n ), (X1,Y1), (X2,Y2),..., (X n ,Y n ) represents the candidate data and baseline variable for the 1st, 2nd, ..., nth plants; Step S2-3: Calculate the correlation direction between the candidate data and the benchmark variable, specifically as follows: ; In the formula, Let be the Pearson correlation coefficient between the i-th candidate data and the benchmark variable Y, and n be the sample size. X is the candidate data for the k-th sample. i The specific value, where k is the sample index. The mean of the candidate data. Let Y be the baseline variable value for the k-th sample. The sample mean of the benchmark variable Y. It is the sum of discrete products of candidate index X and benchmark variable Y. Let X be the sum of squared deviations of the candidate data. The sum of squared deviations of the benchmark variable Y; Step S2-4, when When the value is greater than 0, the candidate indicator is determined to be positively correlated with the plant growth status, and positively correlated data are selected from the basic plant growth dataset as plant growth assessment indicators. Step S2-5: Standardize the plant growth assessment indicators; Step S2-6: Calculate the plant growth assessment score for each time window based on the selected plant growth assessment indicators, specifically as follows: ; In the formula, C1, C2, ..., C n As a growth assessment indicator, , ... The weights assigned to the growth assessment indicators, where S is the growth assessment score; Step S2-7: Select n plants of the same variety, collect growth evaluation indicators for the complete growth cycle time window of the n plants, and calculate the growth evaluation score for each time window. Step S2-8: Arrange the "time window-score" of each plant in chronological order, calculate the score difference between each adjacent time window, integrate the score differences of all adjacent time windows of n historical plants, construct a total difference set, calculate the arithmetic mean of the set, and use the arithmetic mean as the difference threshold. Step S2-9: For the adjacent time window fraction difference sequence of each historical plant, compare each difference with the difference threshold T. If the difference is greater than the difference threshold, determine the adjacent time window boundary corresponding to the difference as the growth stage division node. Use the division node of each plant as the boundary to divide the growth stages of the plant's complete weekly growth cycle.

[0009] Based on the differential parameters, positively correlated evaluation indicators are selected, growth evaluation scores are calculated, and growth stages are divided. Through scientific correlation analysis and threshold setting, different growth stages within the complete growth cycle of plants are accurately defined, providing a clear stage division basis for subsequent water demand management.

[0010] Step S3: Collect the water requirements of plants at different growth stages within each time window and plot the curve of water requirements changing over time. Step S3-1: Obtain meteorological data from the weather station and calculate the green space evapotranspiration (PET) using the FAO Penman-Monteith formula; Step S3-2: Collect the water requirements of the plant at each growth stage within each time window, specifically: ; In the formula, ETc is the water requirement, PET is the evapotranspiration, and Kc is the crop coefficient; Step S3-3: Sort the time windows, with time as the horizontal axis and water demand as the vertical axis, and draw the curve of water demand changing with time based on the time window sequence and water demand.

[0011] By calculating evapotranspiration from meteorological data, combining it with crop coefficients to obtain water demand, and plotting the change curves, the water demand at different growth stages in each time window can be accurately determined. The change pattern of water demand over time is presented intuitively in the form of curves, providing structured data support for water demand prediction.

[0012] Step S4: Fit the water demand change curve over time to construct a water demand prediction model, and predict the water demand of the plant in the next time window based on the water demand prediction model. Step S4-1: Preprocess the data in the water storage volume versus time curve, specifically as follows: For outlier identification and handling, box plots are used to filter out abnormal ETC values. This is achieved by calculating the interquartile range of the ETC data, identifying values ​​exceeding the range. The values ​​are marked as outliers and outliers are removed. Q1 is the first quartile, IQR is the interquartile range, and 1.5 is the coefficient for identifying outliers in the box plot method. For missing values, linear interpolation is used to fill them in, specifically: Let the missing window be t0, and the adjacent valid windows be t1 (ETc=y1) and t2 (ETc=y2). ; In the formula, The missing value for the water demand to be estimated. The water demand corresponding to the first valid time window adjacent to the missing time window. The time window corresponding to the missing water demand value. The first valid time window adjacent to the time window corresponding to the missing value. The water demand is the amount of water required for the second valid time window adjacent to the time window corresponding to the missing value. The second valid time window adjacent to the time window corresponding to the missing value; Step S4-2: Normalize the pre-processed water demand data; Step S4-3: Fit the scattered points in the water demand variation curve over time, and construct a water demand prediction model using a polynomial algorithm, specifically as follows: ; In the formula, For the water demand prediction model, a n a n-1 a1, a0 are polynomial coefficients, n is the order of the polynomial, and t is the time characteristic independent variable; Step S4-4: Predict the water demand of the plant in the next time window based on the water demand prediction model.

[0013] By first processing the water demand data and then using a polynomial algorithm to fit the curve and build a prediction model, the quality of the water demand data can be guaranteed before scientifically predicting the water demand of plants in the next time window, providing forward-looking data support for the formulation of irrigation plans.

[0014] Step S5: Combine real-time meteorological data and plant growth stage, irrigate the plants according to the water demand predicted by the water demand prediction model, evaluate the irrigation results, and revise the model based on the results. Step S5-1: Use sensors to collect plant growth assessment indicators, calculate plant growth assessment scores, and determine the plant growth stage based on the plant growth assessment scores. Step S5-2: Obtain meteorological data for the current and next time window from the weather station, calculate the evapotranspiration and water demand of green space in the current time window, use the water demand prediction model to predict the water demand for the next time window, subtract the current precipitation, and determine the final actual irrigation amount, specifically: ; In the formula, I represents the actual irrigation amount. This represents the actual water requirement of the plant within the current time window. Forecast water demand for the next time window, The pre-irrigation coefficient is set according to the actual situation, and P is the current natural precipitation. Step S5-3: Collect the growth assessment scores of two adjacent time windows of the normal growth of n identical plants, calculate the difference, and set the fluctuation coefficient x according to the actual business scenario, and set the threshold range in combination with the difference. Step S5-4: After irrigation ends and the next time window begins, plant growth assessment indicators are collected again, and the growth assessment score after irrigation is calculated. Step S5-5: Calculate the difference between the growth assessment score after irrigation and the growth assessment score before irrigation. When the difference is within the threshold range, the irrigation is deemed qualified; when the difference is not within the threshold range, the irrigation is deemed unqualified. Step S5-6: When irrigation is deemed unqualified, the water demand prediction model is corrected, specifically as follows: Meteorological data was re-collected to calculate PET, then ETC was calculated, the water demand over time curve was redrawn, and a polynomial algorithm was used to refit the water demand over time curve to construct a new water demand prediction model.

[0015] By combining real-time weather and growth stage data to determine irrigation amounts, evaluate irrigation results, and revise the model, it is possible to ensure that irrigation amounts match the actual needs of plants. Through dynamic model revision, the accuracy of water demand prediction can be continuously improved, enabling dynamic and precise management of plant irrigation.

[0016] The system includes a growth data difference extraction module, a plant growth stage determination module, a plant water requirement curve drawing module, a water requirement prediction model construction module, and an irrigation execution and model correction module. The growth data difference extraction module is used to set the time window for collecting plant growth data, record data, and extract difference parameters. The plant growth stage determination module is used to calculate a score based on the difference parameter as an evaluation index, and then determine the plant growth stage. The plant water requirement curve plotting module is used to calculate the water requirement of the plant in each time window and plot the water requirement change curve over time. The water demand prediction model building module is used to preprocess water demand data and fit curves to build a model to predict water demand. The irrigation execution and model correction module is used to execute irrigation after determining the irrigation amount, evaluate the irrigation results, and correct the water demand prediction model.

[0017] The growth data difference extraction module includes a growth data recording unit and a difference parameter extraction unit; The growth data recording unit is used to divide the plant growth cycle into multiple time windows and record data to generate a growth record table. The difference parameter extraction unit is used to sort the growth record table by time, compare adjacent tables and extract difference parameters. The plant growth stage determination module includes a growth assessment score calculation unit and a growth stage division unit; The growth assessment score calculation unit is used to integrate differential parameter screening indicators and calculate the growth assessment score after standardization. The growth stage division unit is used to calculate the difference threshold based on historical plant scores and to divide the growth stage by comparing the difference.

[0018] The plant water requirement curve plotting module includes a water requirement calculation unit and a water requirement curve plotting unit; The water demand calculation unit is used to acquire meteorological data to calculate evapotranspiration and combine it with crop coefficients to calculate plant water demand. The water demand curve plotting unit is used to sort time windows by time and plot the change curve with time and water demand as axes. The water demand prediction model construction module includes a water demand data preprocessing unit and a water demand prediction model construction unit; The water demand data preprocessing unit is used to identify and remove outlier values ​​in water demand, fill in missing values, and perform normalization processing. The water demand prediction model building unit is used to fit the water demand curve with a polynomial algorithm and build a model to predict the water demand in the next time window.

[0019] The irrigation execution and model correction module includes an irrigation scheme execution unit and an irrigation result evaluation and model correction unit; The irrigation scheme execution unit is used to determine the plant growth stage and, in conjunction with meteorological data and predicted water demand, determine the actual irrigation amount. The irrigation result evaluation and model correction unit is used to evaluate the qualification of irrigation. If the qualification is not met, the water demand is recalculated and a new prediction model is constructed.

[0020] Compared with the prior art, the beneficial effects of the present invention are: 1. This invention, from growth data collection and difference extraction to water requirement calculation, prediction and irrigation decision-making, is based on standardized formulas and scientific data processing methods. It focuses on data throughout the entire process, greatly improving the accuracy of plant growth monitoring, stage division, water requirement management and irrigation decision-making.

[0021] 2. This invention combines the division of plant growth stages with the calculation, prediction, and depth of water requirements and irrigation execution. First, the stages are defined based on growth data, then the water requirements are calculated for different stages, and finally, an irrigation plan is formulated based on the stage characteristics and water requirements prediction. This achieves coordinated linkage between growth management and irrigation management, and avoids the disconnect between irrigation and plant growth needs.

[0022] 3. This invention sets up irrigation result evaluation and model correction links. After irrigation, the irrigation qualification is judged by the score difference. If the qualification is not qualified, the water demand prediction model is reconstructed, forming a closed loop of "data collection - model construction - decision execution - result evaluation - model optimization". This ensures that the water demand prediction model and irrigation management plan can continuously adapt to changes in plant growth and improve long-term management results. Attached Figure Description

[0023] Figure 1 This is a flowchart illustrating an intelligent monitoring method for urban green spaces based on the Internet of Things according to the present invention. Figure 2 This is a schematic diagram of the structure of an intelligent monitoring system for urban green spaces based on the Internet of Things according to the present invention. Detailed Implementation

[0024] 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.

[0025] Example 1: As Figures 1-2 As shown, the present invention provides a technical solution, a smart monitoring method for urban green spaces based on the Internet of Things, the method comprising the following steps: Step S1: Set a time window, collect growth data of the same plant at different time periods, generate growth record tables for each time period, and compare the growth record tables for different time periods. Step S1-1: Set a time window t to divide the plant's growth cycle into multiple consecutive time windows; Step S1-2: Record the plant growth data within each time window and generate a plant growth record table for the corresponding time window; Steps S1-3: Sort the plant growth record tables for each time window according to the time series, traverse and compare the plant growth record tables of two adjacent time windows, define "data names are the same but values ​​are different" as difference parameters, and extract the difference parameters.

[0026] By setting time windows to divide the growth cycle, recording growth data, and comparing and extracting difference parameters, we can obtain the changes in plant growth data in different consecutive time periods, providing data support for the determination of subsequent growth assessment indicators.

[0027] Step S2: Use the difference parameters obtained from the comparison as growth assessment indicators, calculate the growth assessment score, and determine the plant growth stage based on the growth assessment score. Step S2-1: Integrate the differential parameters and construct a basic plant growth dataset; Step S2-2: Using the increase in plant height as the baseline variable Y, extract all data from the plant growth dataset as candidate data, denoted as X. Select synchronous observation data from n plants to form sample pairs (X1, Y1), (X2, Y2), ..., (X... n ,Y n ), (X1,Y1), (X2,Y2),..., (X n ,Y n ) represents the candidate data and baseline variable for the 1st, 2nd, ..., nth plants; Step S2-3: Calculate the correlation direction between the candidate data and the benchmark variable, specifically as follows: ; In the formula, Let be the Pearson correlation coefficient between the i-th candidate data and the benchmark variable Y, and n be the sample size. X is the candidate data for the k-th sample. i The specific value, where k is the sample index. The mean of the candidate data. Let Y be the baseline variable value for the k-th sample. The sample mean of the benchmark variable Y. It is the sum of discrete products of candidate index X and benchmark variable Y. Let X be the sum of squared deviations of the candidate data. The sum of squared deviations of the benchmark variable Y; Step S2-4, when When the value is greater than 0, the candidate indicator is determined to be positively correlated with the plant growth status, and positively correlated data are selected from the basic plant growth dataset as plant growth assessment indicators. Step S2-5: Standardize the plant growth assessment indicators; Step S2-6: Calculate the plant growth assessment score for each time window based on the selected plant growth assessment indicators, specifically as follows: ; In the formula, C1, C2, ..., C n As a growth assessment indicator, , ... The weights assigned to the growth assessment indicators, where S is the growth assessment score; Step S2-7: Select n plants of the same variety, collect growth evaluation indicators for the complete growth cycle time window of the n plants, and calculate the growth evaluation score for each time window. Step S2-8: Arrange the "time window-score" of each plant in chronological order, calculate the score difference between each adjacent time window, integrate the score differences of all adjacent time windows of n historical plants, construct a total difference set, calculate the arithmetic mean of the set, and use the arithmetic mean as the difference threshold. Step S2-9: For the adjacent time window fraction difference sequence of each historical plant, compare each difference with the difference threshold T. If the difference is greater than the difference threshold, determine the adjacent time window boundary corresponding to the difference as the growth stage division node. Use the division node of each plant as the boundary to divide the growth stages of the plant's complete weekly growth cycle.

[0028] Based on the differential parameters, positively correlated evaluation indicators are selected, growth evaluation scores are calculated, and growth stages are divided. Through scientific correlation analysis and threshold setting, different growth stages within the complete growth cycle of plants are accurately defined, providing a clear stage division basis for subsequent water demand management.

[0029] Step S3: Collect the water requirements of plants at different growth stages within each time window and plot the curve of water requirements changing over time. Step S3-1: Obtain meteorological data from the weather station and calculate the green space evapotranspiration (PET) using the FAO Penman-Monteith formula; Step S3-2: Collect the water requirements of the plant at each growth stage within each time window, specifically: ; In the formula, ETc is the water requirement, PET is the evapotranspiration, and Kc is the crop coefficient; Step S3-3: Sort the time windows, with time as the horizontal axis and water demand as the vertical axis, and draw the curve of water demand changing with time based on the time window sequence and water demand.

[0030] By calculating evapotranspiration from meteorological data, combining it with crop coefficients to obtain water demand, and plotting the change curves, the water demand at different growth stages in each time window can be accurately determined. The change pattern of water demand over time is presented intuitively in the form of curves, providing structured data support for water demand prediction.

[0031] Step S4: Fit the water demand change curve over time to construct a water demand prediction model, and predict the water demand of the plant in the next time window based on the water demand prediction model. Step S4-1: Preprocess the data in the water storage volume versus time curve, specifically as follows: For outlier identification and handling, box plots are used to filter out abnormal ETC values. This is achieved by calculating the interquartile range of the ETC data, identifying values ​​exceeding the range. The values ​​are marked as outliers and outliers are removed. Q1 is the first quartile, IQR is the interquartile range, and 1.5 is the coefficient for identifying outliers in the box plot method. For missing values, linear interpolation is used to fill them in, specifically: Let the missing window be t0, and the adjacent valid windows be t1 (ETc=y1) and t2 (ETc=y2). ; In the formula, The missing value for the water demand to be estimated. The water demand corresponding to the first valid time window adjacent to the missing time window. The time window corresponding to the missing water demand value. The first valid time window adjacent to the time window corresponding to the missing value. The water demand is the amount of water required for the second valid time window adjacent to the time window corresponding to the missing value. The second valid time window adjacent to the time window corresponding to the missing value; Step S4-2: Normalize the pre-processed water demand data; Step S4-3: Fit the scattered points in the water demand variation curve over time, and construct a water demand prediction model using a polynomial algorithm, specifically as follows: ; In the formula, For the water demand prediction model, a n a n-1a1, a0 are polynomial coefficients, n is the order of the polynomial, and t is the time characteristic independent variable; Step S4-4: Predict the water demand of the plant in the next time window based on the water demand prediction model.

[0032] By first processing the water demand data and then using a polynomial algorithm to fit the curve and build a prediction model, the quality of the water demand data can be guaranteed before scientifically predicting the water demand of plants in the next time window, providing forward-looking data support for the formulation of irrigation plans.

[0033] Step S5: Combine real-time meteorological data and plant growth stage, irrigate the plants according to the water demand predicted by the water demand prediction model, evaluate the irrigation results, and revise the model based on the results. Step S5-1: Use sensors to collect plant growth assessment indicators, calculate plant growth assessment scores, and determine the plant growth stage based on the plant growth assessment scores. Step S5-2: Obtain meteorological data for the current and next time window from the weather station, calculate the evapotranspiration and water demand of green space in the current time window, use the water demand prediction model to predict the water demand for the next time window, subtract the current precipitation, and determine the final actual irrigation amount, specifically: ; In the formula, I represents the actual irrigation amount. This represents the actual water requirement of the plant within the current time window. Forecast water demand for the next time window, The pre-irrigation coefficient is set according to the actual situation, and P is the current natural precipitation. Step S5-3: Collect the growth assessment scores of two adjacent time windows of the normal growth of n identical plants, calculate the difference, and set the fluctuation coefficient x according to the actual business scenario, and set the threshold range in combination with the difference. Step S5-4: After irrigation ends and the next time window begins, plant growth assessment indicators are collected again, and the growth assessment score after irrigation is calculated. Step S5-5: Calculate the difference between the growth assessment score after irrigation and the growth assessment score before irrigation. When the difference is within the threshold range, the irrigation is deemed qualified; when the difference is not within the threshold range, the irrigation is deemed unqualified. Step S5-6: When irrigation is deemed unqualified, the water demand prediction model is corrected, specifically as follows: Meteorological data was re-collected to calculate PET, then ETC was calculated, the water demand over time curve was redrawn, and a polynomial algorithm was used to refit the water demand over time curve to construct a new water demand prediction model.

[0034] By combining real-time weather and growth stage data to determine irrigation amounts, evaluate irrigation results, and revise the model, it is possible to ensure that irrigation amounts match the actual needs of plants. Through dynamic model revision, the accuracy of water demand prediction can be continuously improved, enabling dynamic and precise management of plant irrigation.

[0035] The system includes a growth data difference extraction module, a plant growth stage determination module, a plant water requirement curve drawing module, a water requirement prediction model construction module, and an irrigation execution and model correction module. The growth data difference extraction module is used to set the time window for collecting plant growth data, record data, and extract difference parameters. The plant growth stage determination module is used to calculate a score based on the difference parameter as an evaluation index, and then determine the plant growth stage. The plant water requirement curve plotting module is used to calculate the water requirement of the plant in each time window and plot the water requirement change curve over time. The water demand prediction model building module is used to preprocess water demand data and fit curves to build a model to predict water demand. The irrigation execution and model correction module is used to execute irrigation after determining the irrigation amount, evaluate the irrigation results, and correct the water demand prediction model.

[0036] The growth data difference extraction module includes a growth data recording unit and a difference parameter extraction unit; The growth data recording unit is used to divide the plant growth cycle into multiple time windows and record data to generate a growth record table. The difference parameter extraction unit is used to sort the growth record table by time, compare adjacent tables and extract difference parameters. The plant growth stage determination module includes a growth assessment score calculation unit and a growth stage division unit; The growth assessment score calculation unit is used to integrate differential parameter screening indicators and calculate the growth assessment score after standardization. The growth stage division unit is used to calculate the difference threshold based on historical plant scores and to divide the growth stage by comparing the difference.

[0037] The plant water requirement curve plotting module includes a water requirement calculation unit and a water requirement curve plotting unit; The water demand calculation unit is used to acquire meteorological data to calculate evapotranspiration and combine it with crop coefficients to calculate plant water demand. The water demand curve plotting unit is used to sort time windows by time and plot the change curve with time and water demand as axes. The water demand prediction model construction module includes a water demand data preprocessing unit and a water demand prediction model construction unit; The water demand data preprocessing unit is used to identify and remove outlier values ​​in water demand, fill in missing values, and perform normalization processing. The water demand prediction model building unit is used to fit the water demand curve with a polynomial algorithm and build a model to predict the water demand in the next time window.

[0038] The irrigation execution and model correction module includes an irrigation scheme execution unit and an irrigation result evaluation and model correction unit; The irrigation scheme execution unit is used to determine the plant growth stage and, in conjunction with meteorological data and predicted water demand, determine the actual irrigation amount. The irrigation result evaluation and model correction unit is used to evaluate the qualification of irrigation. If the qualification is not met, the water demand is recalculated and a new prediction model is constructed.

[0039] Example 2: Set a time window t=7 days, divide the 140-day growth cycle of grass into 20 consecutive time windows, record the plant height, number of leaves, and chlorophyll content of grass, generate 20 growth record tables, sort the growth record tables by time, compare adjacent time windows, and extract the difference parameters of plant height, number of leaves, and chlorophyll value. A basic dataset was constructed by integrating differential parameters. With the increase in plant height as the baseline variable Y, 20 grass plants were selected for synchronous observation data to form 30 sample pairs. The Pearson correlation coefficient was calculated, and the number of leaves, the increase in plant height, and the chlorophyll value were selected as evaluation indicators. The indicators were standardized to the interval [0,1]. The weights for plant height increment are assigned as 0.4, leaf number as 0.3, and chlorophyll value as 0.4. After standardization in the second time window, the plant height increment is 0.8, the leaf number is 0.7, and the chlorophyll value is 0.6. The score for the second time window is calculated as S = 0.71. Calculate the growth assessment scores of 30 grass plants for each time window, and the arithmetic mean is 0.05. Set the difference threshold to 0.05. The difference between the scores of the 5th and 6th time windows is calculated to be 0.07, which is greater than the difference threshold. Therefore, the boundary is determined to be a dividing node. Meteorological data for the 6th time window was obtained from the weather station, and PET was calculated to be 4 mm / d; the crop coefficient of grass, kc, was 0.8, and ETc was calculated to be 3.2 mm / d. Similarly, the ETc values ​​for 20 time windows were calculated and curves were plotted. Using a polynomial algorithm to fit the model, ETc = 0.002t was obtained. 2 -0.01t+3.1; It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

Claims

1. A smart monitoring method for urban green spaces based on the Internet of Things, characterized in that: The method includes the following steps: Step S1: Set a time window, collect growth data of the same plant at different time periods, generate growth record tables for each time period, and compare the growth record tables for different time periods. Step S2: Use the difference parameters obtained from the comparison as growth assessment indicators, calculate the growth assessment score, and determine the plant growth stage based on the growth assessment score. Step S3: Collect the water requirements of plants at different growth stages within each time window and plot the curve of water requirements changing over time. Step S4: Fit the water demand change curve over time to construct a water demand prediction model, and predict the water demand of the plant in the next time window based on the water demand prediction model. Step S5: Combine real-time meteorological data and plant growth stage, irrigate the plants according to the water demand predicted by the water demand prediction model, evaluate the irrigation results, and revise the model based on the results.

2. The method for intelligent monitoring of urban green spaces based on the Internet of Things according to claim 1, characterized in that: The specific steps of step S1 are as follows: Step S1-1: Set a time window t to divide the plant's growth cycle into multiple consecutive time windows; Step S1-2: Record the plant growth data within each time window and generate a plant growth record table for the corresponding time window; Steps S1-3: Sort the plant growth record tables for each time window according to the time series, traverse and compare the plant growth record tables of two adjacent time windows, define "data names are the same but values ​​are different" as the difference parameter, and extract the difference parameter.

3. The method for intelligent monitoring of urban green spaces based on the Internet of Things according to claim 2, characterized in that: The specific steps of step S2 are as follows: Step S2-1: Integrate the differential parameters and construct a basic plant growth dataset; Step S2-2: Using the increase in plant height as the baseline variable Y, extract all data from the plant growth dataset as candidate data, denoted as X. Select synchronous observation data from n plants to form sample pairs (X1, Y1), (X2, Y2), ..., (X... n ,Y n ), (X1,Y1), (X2,Y2),..., (X n ,Y n ) represents the candidate data and baseline variable for the 1st, 2nd, ..., nth plants; Step S2-3: Calculate the correlation direction between the candidate data and the benchmark variable, specifically as follows: ； In the formula, Let be the Pearson correlation coefficient between the i-th candidate data and the benchmark variable Y, and n be the sample size. X is the candidate data for the k-th sample. i The specific value, where k is the sample index. The mean of the candidate data. Let Y be the baseline variable value for the k-th sample. The sample mean of the benchmark variable Y. It is the sum of discrete products of candidate index X and benchmark variable Y. Let X be the sum of squared deviations of the candidate data. The sum of squared deviations of the benchmark variable Y; Step S2-4, when When the value is greater than 0, the candidate indicator is determined to be positively correlated with the plant growth status, and positively correlated data are selected from the basic plant growth dataset as plant growth assessment indicators. Step S2-5: Standardize the plant growth assessment indicators; Step S2-6: Calculate the plant growth assessment score for each time window based on the selected plant growth assessment indicators, specifically as follows: ； In the formula, C1, C2, ..., C n As a growth assessment indicator, , ... The weights assigned to the growth assessment indicators, where S is the growth assessment score; Step S2-7: Select n plants of the same variety, collect growth evaluation indicators for the complete growth cycle time window of the n plants, and calculate the growth evaluation score for each time window. Step S2-8: Arrange the "time window-score" of each plant in chronological order, calculate the score difference between each adjacent time window, integrate the score differences of all adjacent time windows of n historical plants, construct a total difference set, calculate the arithmetic mean of the set, and use the arithmetic mean as the difference threshold. Step S2-9: For the adjacent time window fraction difference sequence of each historical plant, compare each difference with the difference threshold T. If the difference is greater than the difference threshold, determine the adjacent time window boundary corresponding to the difference as the growth stage division node. Use the division node of each plant as the boundary to divide the growth stages of the plant's complete weekly growth cycle.

4. The method for intelligent monitoring of urban green spaces based on the Internet of Things according to claim 3, characterized in that: The specific steps of step S3 are as follows: Step S3-1: Obtain meteorological data from the weather station and calculate the green space evapotranspiration (PET) using the FAO Penman-Monteith formula; Step S3-2: Collect the water requirements of the plant at each growth stage within each time window, specifically: ； In the formula, ETc is the water requirement, PET is the evapotranspiration, and Kc is the crop coefficient; Step S3-3: Sort the time windows, with time as the horizontal axis and water demand as the vertical axis, and draw the curve of water demand changing with time based on the time window sequence and water demand.

5. The method for intelligent monitoring of urban green spaces based on the Internet of Things according to claim 4, characterized in that: The specific steps of step S4 are as follows: Step S4-1: Preprocess the data in the water storage volume versus time curve, specifically as follows: For outlier identification and handling, box plots are used to filter out abnormal ETC values. This is achieved by calculating the interquartile range of the ETC data, identifying values ​​exceeding the range. The values ​​are marked as outliers and outliers are removed. Q1 is the first quartile, IQR is the interquartile range, and 1.5 is the coefficient for identifying outliers in the box plot method. For missing values, linear interpolation is used to fill them in, specifically: Let the missing window be t0, and the adjacent valid windows be t1 (ETc=y1) and t2 (ETc=y2). ; In the formula, The missing value for the water demand to be estimated. The water demand corresponding to the first valid time window adjacent to the missing time window. The time window corresponding to the missing water demand value. The first valid time window adjacent to the time window corresponding to the missing value. The water demand is the amount of water required for the second valid time window adjacent to the time window corresponding to the missing value. The second valid time window adjacent to the time window corresponding to the missing value; Step S4-2: Normalize the pre-processed water demand data; Step S4-3: Fit the scattered points in the water demand variation curve over time, and construct a water demand prediction model using a polynomial algorithm, specifically as follows: ； In the formula, For the water demand prediction model, a n a n-1 a1, a0 are polynomial coefficients, n is the order of the polynomial, and t is the time characteristic independent variable; Step S4-4: Predict the water demand of the plant in the next time window based on the water demand prediction model.

6. The method for intelligent monitoring of urban green spaces based on the Internet of Things according to claim 5, characterized in that: The specific steps of step S5 are as follows: Step S5-1: Use sensors to collect plant growth assessment indicators, calculate plant growth assessment scores, and determine the plant growth stage based on the plant growth assessment scores. Step S5-2: Obtain meteorological data for the current and next time window from the weather station, calculate the evapotranspiration and water demand of green space in the current time window, use the water demand prediction model to predict the water demand for the next time window, subtract the current precipitation, and determine the final actual irrigation amount, specifically: ； In the formula, I represents the actual irrigation amount. This represents the actual water requirement of the plant within the current time window. Forecast water demand for the next time window, The pre-irrigation coefficient is set according to the actual situation, and P is the current natural precipitation. Step S5-3: Collect the growth assessment scores of two adjacent time windows of the normal growth of n identical plants, calculate the difference, and set the fluctuation coefficient x according to the actual business scenario, and set the threshold range in combination with the difference. Step S5-4: After irrigation ends and the next time window begins, plant growth assessment indicators are collected again, and the growth assessment score after irrigation is calculated. Step S5-5: Calculate the difference between the growth assessment score after irrigation and the growth assessment score before irrigation. When the difference is within the threshold range, the irrigation is deemed qualified; when the difference is not within the threshold range, the irrigation is deemed unqualified. Step S5-6: When irrigation is deemed unqualified, the water demand prediction model is corrected, specifically as follows: Meteorological data was re-collected to calculate PET, then ETC was calculated, the water demand over time curve was redrawn, and a polynomial algorithm was used to refit the water demand over time curve to construct a new water demand prediction model.

7. An intelligent monitoring system for urban green spaces based on the Internet of Things, characterized in that: The system includes a growth data difference extraction module, a plant growth stage determination module, a plant water requirement curve drawing module, a water requirement prediction model construction module, and an irrigation execution and model correction module. The growth data difference extraction module is used to set the time window for collecting plant growth data, record data, and extract difference parameters. The plant growth stage determination module is used to calculate a score based on the difference parameter as an evaluation index, and then determine the plant growth stage. The plant water requirement curve plotting module is used to calculate the water requirement of the plant in each time window and plot the water requirement change curve over time. The water demand prediction model building module is used to preprocess water demand data and fit curves to build a model to predict water demand. The irrigation execution and model correction module is used to execute irrigation after determining the irrigation amount, evaluate the irrigation results, and correct the water demand prediction model.

8. The intelligent monitoring system for urban green spaces based on the Internet of Things according to claim 7, characterized in that: The growth data difference extraction module includes a growth data recording unit and a difference parameter extraction unit; The growth data recording unit is used to divide the plant growth cycle into multiple time windows and record data to generate a growth record table. The difference parameter extraction unit is used to sort the growth record table by time, compare adjacent tables and extract difference parameters. The plant growth stage determination module includes a growth assessment score calculation unit and a growth stage division unit; The growth assessment score calculation unit is used to integrate differential parameter screening indicators and calculate the growth assessment score after standardization. The growth stage division unit is used to calculate the difference threshold based on historical plant scores and to divide the growth stage by comparing the difference.

9. The intelligent monitoring system for urban green spaces based on the Internet of Things according to claim 7, characterized in that: The plant water requirement curve plotting module includes a water requirement calculation unit and a water requirement curve plotting unit; The water demand calculation unit is used to acquire meteorological data to calculate evapotranspiration and combine it with crop coefficients to calculate plant water demand. The water demand curve plotting unit is used to sort time windows by time and plot the change curve with time and water demand as axes. The water demand prediction model construction module includes a water demand data preprocessing unit and a water demand prediction model construction unit; The water demand data preprocessing unit is used to identify and remove outlier values ​​in water demand, fill in missing values, and perform normalization processing. The water demand prediction model building unit is used to fit the water demand curve with a polynomial algorithm and build a model to predict the water demand in the next time window.

10. The intelligent monitoring system for urban green spaces based on the Internet of Things according to claim 7, characterized in that: The irrigation execution and model correction module includes an irrigation scheme execution unit and an irrigation result evaluation and model correction unit; The irrigation scheme execution unit is used to determine the plant growth stage and, in conjunction with meteorological data and predicted water demand, determine the actual irrigation amount. The irrigation result evaluation and model correction unit is used to evaluate the qualification of irrigation. If the qualification is not met, the water demand is recalculated and a new prediction model is constructed.

Images