Water-saving type precise irrigation control system and method

Abstract

The invention discloses a water-saving precise irrigation control system and method, and relates to the technical field of agricultural irrigation, and the method comprises the steps: obtaining an irrigation region and an environment data set, preprocessing the environment data set, generating credible data, carrying out fluctuation evaluation on the credible data, and generating a data quality index; executing a gating decision on an irrigation mode judgment basis based on a data quality index, generating a partition target irrigation instruction set, executing infiltration capacity online identification on historical irrigation responses in the partition target irrigation instruction set and the environment data set, and generating a pulse irrigation sequence; and executing hydraulic state analysis on the pipe network flow pressure and the pump station operation data in the environmental data set, generating a water supply constraint, scheduling a pulse irrigation sequence based on the water supply constraint, and issuing the pulse irrigation sequence to a valve and a pump station to generate an irrigation execution result. According to the method, by forming water supply constraints, the strategy performability is improved, and meanwhile the field deviation is reduced.

Description

Technical Field

[0001] This invention relates to the field of agricultural irrigation technology, and in particular to a water-saving precision irrigation control system and method. Background Technology

[0002] In recent years, agricultural irrigation technology has gradually evolved from timed and quantitative irrigation to a precision irrigation system centered on sensor monitoring, data modeling, and closed-loop control. With the development of the Internet of Things, low-power sensors, edge computing, and agricultural information platforms, irrigation control systems can continuously acquire information on soil moisture, meteorological elements, pipeline hydraulic status, and pump and valve operation. Through model inference, they can infer crop water requirements and irrigation strategies, achieving zoned water supply and process control. At the same time, water-saving engineering practices have promoted the rapid popularization of efficient irrigation methods such as drip irrigation, micro-sprinkler irrigation, and seepage irrigation in facility agriculture and large-scale planting.

[0003] However, existing technologies still have the following shortcomings: field sensor data is significantly affected by drift, noise, missing data, communication fluctuations, etc., which leads to the execution of irrigation of the same intensity when the data quality is poor, thus causing mis-irrigation and over-irrigation. Zonal modeling often relies on static empirical zoning or representative values ​​of a single sensor point, resulting in unstable estimation of zonal water state quantities. Summary of the Invention

[0004] In view of the aforementioned existing problems, the present invention is proposed.

[0005] Therefore, this invention provides a water-saving precision irrigation control method to solve the problem of performing irrigation of the same intensity even when the data quality is poor.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution: In a first aspect, the present invention provides a water-saving precision irrigation control method, comprising, Acquire irrigation area and environmental datasets, preprocess the environmental datasets to generate reliable data, perform volatility assessment on the reliable data, and generate data quality indicators. The irrigation area is divided into multiple zones and a zone index is established to generate a zone set. Based on reliable data and data quality indicators, multi-source data fusion is performed on the zone set to generate zone moisture status. Water demand is inferred from the regional moisture state data and environmental datasets to generate regional demand intervals. These regional demand intervals are then correlated with data quality indicators to generate criteria for determining irrigation patterns. Based on data quality indicators, gating decisions are made to determine irrigation patterns, generating a set of zonal target irrigation instructions. The infiltration capacity of the historical irrigation responses in the zonal target irrigation instruction set and environmental dataset is identified online, and a pulse irrigation sequence is generated. Hydraulic state analysis is performed on the pipeline flow and pressure and pump station operation data in the environmental dataset to generate water supply constraints. Based on the water supply constraints, the pulse irrigation sequence is scheduled and sent to valves and pump stations to generate irrigation execution results.

[0007] As a preferred embodiment of the water-saving precision irrigation control method of the present invention, the specific steps for generating data quality indicators are as follows: Obtain irrigation area and environmental datasets, perform time alignment and field consistency processing on the environmental dataset, and generate aligned data; The aligned data is filled with missing data and anomaly removal is performed to generate clean data. Drift identification and correction are then performed to generate reliable data. Perform data grouping processing on reliable data to generate grouped data fragments, and perform fluctuation amplitude statistics and fluctuation persistence identification to generate fluctuation characteristics; Stability grading and aggregation are performed on fluctuation characteristics to generate data quality indicators.

[0008] In a preferred embodiment of the water-saving precision irrigation control method of the present invention, the specific steps for generating the partition set are as follows: The zoning boundaries are determined by the relationship between the topology of the irrigation pipeline network and the valve layout in the irrigation area, and the area is divided to generate zoning description information; Register the partition identifier and partition index for the partition description information, and generate a partition index table; The partition index table and partition description information are combined to generate a partition set.

[0009] As a preferred embodiment of the water-saving precision irrigation control method of the present invention, the specific steps for generating the zoned water status data are as follows: Extract trusted data subsets corresponding to each partition from the partition set and aggregate them to generate partition trusted data subsets; Weights are assigned to trusted subsets of data in partitions based on data quality metrics to generate fusion weights. Using the fusion weight as a constraint, a weighted filtering is performed on the reliable subset of partitioned data to generate the partitioned moisture status.

[0010] As a preferred embodiment of the water-saving precision irrigation control method of the present invention, the specific steps for generating the zoned demand intervals are as follows: Environmental factors are extracted from the environmental dataset, and time-series feature extraction is performed on the environmental factors to generate environmental trend information; Perform water demand forecasting on environmental trend information and regional water state quantities to generate regional water demand estimation results; Based on soil characteristics and long-term climate data in the environmental dataset, upper and lower limits and confidence levels of the regional water demand estimation results are set, and the regional water demand estimation results are expressed in interval form to generate regional demand intervals.

[0011] As a preferred embodiment of the water-saving precision irrigation control method of the present invention, the method for generating irrigation mode determination criteria refers to extracting zonal demand features and fluctuation level features from zonal demand intervals and data quality indicators respectively, prioritizing the zonal demand features and fluctuation level features, generating determination rule items, and compiling the determination rule items to form irrigation mode determination criteria.

[0012] As a preferred embodiment of the water-saving precision irrigation control method of the present invention, the specific steps for generating the zoned target irrigation instruction set are as follows: Based on historical fluctuation levels, missing drift characteristics, and sensor performance parameters, a quality judgment threshold is set. The data quality indicators are compared with the quality judgment threshold to generate gating discrimination results. When the gating judgment result shows that the data quality is available, the irrigation mode is determined to be the normal irrigation mode by the judgment rule item in the irrigation mode judgment basis, and the median value of the interval is selected from the partition demand interval as the normal target demand. When the gating judgment result shows that the data quality is in a limited state, the irrigation mode is determined to be a degraded irrigation mode by the judgment rule item in the irrigation mode judgment basis, and the lower limit value of the interval is selected from the partition demand interval as the degraded target demand. The normal target demand, the degraded target demand, and the partition index are combined to generate a partition target irrigation instruction set.

[0013] In a preferred embodiment of the water-saving precision irrigation control method of the present invention, the specific steps for generating the pulse irrigation sequence are as follows: The historical irrigation responses in the environmental dataset are aggregated according to the partition index to generate a partitioned historical response sequence. The historical response sequence of the partition is matched with the target irrigation instruction set of the partition, and the water change features before and after irrigation are extracted to generate an identification sample set. The partition infiltration characteristic parameters are calculated based on the identified sample set, and the duration and interval of the pulse are determined by the partition infiltration characteristic parameters, and then the parameters are aggregated to form a partition pulse parameter set. The target irrigation instruction set for each zone is pulsed according to the zone pulse parameter set to generate a pulse irrigation sequence.

[0014] As a preferred embodiment of the water-saving precision irrigation control method of the present invention, the specific steps for generating water supply constraints are as follows: The pipeline flow and pressure data in the environmental dataset are time-aligned and the pump station operation data are processed to generate hydraulic condition data. The core nodes of the pipeline network are determined by the location of pressure monitoring devices and flow monitoring devices on the pipeline network. Based on hydraulic condition data, the pressure level and flow distribution status of the core nodes of the pipeline network are identified, and the risks of insufficient pressure and flow congestion are determined, aggregated, and risk markers are generated. Based on pump station operation data, the available water supply capacity is determined, and the number of zones that can be irrigated concurrently and the available water supply intensity are determined by the available water supply capacity and risk markers. The number of zones that can be irrigated concurrently and the available water supply intensity are combined to form a water supply constraint.

[0015] Secondly, the present invention provides a water-saving precision irrigation control system, comprising, The data quality module is used to acquire irrigation area and environmental datasets, preprocess the environmental datasets to generate reliable data, perform volatility assessment on the reliable data, and generate data quality indicators. The partition index module is used to divide the irrigation area into multiple partitions and establish partition indexes, generate a partition set, perform multi-source data fusion on the partition set based on reliable data and data quality indicators, and generate partition moisture status. The irrigation mode module is used to infer water demand from the regional water status and environmental dataset, generate regional demand intervals, and associate the regional demand intervals with data quality indicators to generate the basis for irrigation mode determination. The gating decision module is used to perform gating decisions based on data quality indicators to determine irrigation patterns, generate a set of zonal target irrigation instructions, perform online identification of infiltration capacity on the historical irrigation responses in the zonal target irrigation instruction set and environmental dataset, and generate pulse irrigation sequences. The irrigation execution module is used to perform hydraulic state analysis on the pipeline flow and pressure and pump station operation data in the environmental dataset, generate water supply constraints, schedule the pulse irrigation sequence based on the water supply constraints, and send it to valves and pump stations to generate irrigation execution results.

[0016] The beneficial effects of this invention are as follows: by generating data quality indicators, it improves control robustness and reduces the risk of mis-irrigation; by generating zonal water state quantities, it reduces the influence of single-point noise and drift on zonal states; by generating irrigation mode determination criteria, it reduces human experience intervention and makes the decision-making chain clearer and more interpretable; by determining normal irrigation mode and degraded irrigation mode, it can still avoid over-irrigation and maintain safe water supply when data is abnormal; by generating pulsed irrigation sequences, it makes the irrigation rhythm match the infiltration characteristics of the zonal area and reduces runoff and deep infiltration caused by a large volume of water at one time; by forming water supply constraints, it improves the executability of the strategy and reduces on-site deviations. Attached Figure Description

[0017] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a flowchart of a water-saving precision irrigation control method.

[0019] Figure 2 This is a schematic diagram of a water-saving precision irrigation control system.

[0020] Figure 3 A flowchart for generating data quality metrics.

[0021] Figure 4 This is a flowchart for generating a pulsed irrigation sequence. Detailed Implementation

[0022] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0023] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0024] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0025] Reference Figures 1-4 As an embodiment of the present invention, this embodiment provides a water-saving precision irrigation control method, comprising the following steps: S1: Obtain irrigation area and environmental datasets, preprocess the environmental datasets to generate reliable data, perform volatility assessment on the reliable data, and generate data quality indicators.

[0026] S1.1: Obtain irrigation area and environmental datasets, perform time alignment and field consistency processing on the environmental dataset, and generate aligned data.

[0027] Irrigation areas are obtained through plot boundary records. The environmental dataset for these areas is then read, including soil moisture monitoring records, meteorological monitoring records, pipeline flow and pressure data, and pump station operation data. The timestamp field of each data point in the environmental dataset is read, and its representation type is determined. If the timestamp field is a string, the year, month, day, hour, minute, and second are extracted segment by segment and concatenated according to a fixed separator order to obtain a unified year-month-day-hour-minute-second timestamp format. If the timestamp field is a count type, the count unit is converted to the corresponding date and time, and then converted to a unified year-month-day-hour-minute-second timestamp format. If the timestamp field contains a time zone marker, the date and time are converted to a unified time zone date and time and a unified year-month-day-hour-minute-second timestamp format, and the unified year-month-day-hour-minute-second timestamp is written back to the timestamp field, resulting in a unified timestamp environmental dataset. This unified timestamp environmental dataset is then sorted by the timestamp field to obtain the timestamps. The earliest and latest timestamp values ​​are sequentially listed from the earliest timestamp value to the latest timestamp value according to a fixed sampling period, forming a target timeline. The fixed sampling period is on the minute level. The unified timestamp environment dataset is matched against the target timeline field by field and time point by time. When a corresponding time point record exists in the target timeline, the field value of the corresponding time point record is directly taken. When a corresponding time point record is missing in the target timeline, the value of the most recent time point is used to extend and fill in the missing field value. The filled field value is then written to the corresponding time point in the target timeline to form time-aligned data. The field name, field unit, and field value type of the time-aligned data are read field by field. Synonymous field names are converted to unified field names, different field units are converted to unified field units, and string values ​​are converted to numeric values ​​while uniformly retaining precision, such as two decimal places, to form field-consistent data. The field-consistent data is then collected and organized in ascending order of timestamp field to generate aligned data.

[0028] S1.2: Perform missing data completion and anomaly removal on the aligned data to generate clean data, and perform drift identification and correction to generate reliable data.

[0029] The aligned data is read and split into time-series data arranged by timestamp field, based on field name. When null values ​​or missing records appear at adjacent time points, the most recent valid values ​​before and after the missing position are read and linear interpolation is used for completion. When the consecutive missing length of the field time-series data exceeds a fixed missing window, such as more than ten minutes, the most recent valid value is used for extension completion. The allowed value range for each field name in the completed data is set according to the sensor performance parameters. For example, the allowed value range for the field name in soil moisture monitoring records is 0-1, and the allowed value range for the temperature field name in meteorological monitoring records is -40°C to 60°C. The rate of change range for each field name in the completed data is set according to the physical change rate and engineering fluctuation rate under minute-level sampling periods. For example, the rate of change range for the field name in soil moisture monitoring records is no more than 5%-10% between adjacent sampling points, and the rate of change range for the temperature field name is... The range is defined as no more than 5° between adjacent sampling points. When the field value in the completed data exceeds the allowed range, the out-of-range value is deleted and the corresponding time point is marked as an outlier. When the change in the field value in the completed data between adjacent time points exceeds the rate of change range, the abrupt change value is deleted and the corresponding time point is marked as an outlier. Outliers are backfilled using the same missing data completion process to obtain cleaned data. The cleaned data is organized into field time series data according to field names. The field time series data is divided into multiple window segments according to a fixed time window, such as 30 minutes. The field values ​​of each window segment are read and the mean and variance of the window segment are calculated. When the mean of a window segment maintains the same direction of shift more than twice between adjacent windows, it is marked as a drift segment. The mean of the stable segment before the start of the drift segment is read as the baseline mean, and the difference between the drift segment mean and the baseline mean is calculated to obtain the offset. The offset is deducted from the field values ​​of the drift segment to complete the drift correction. The cleaned data after drift correction is collected and organized to obtain reliable data.

[0030] S1.3: Perform data grouping processing on the reliable data to generate grouped data fragments, and perform fluctuation amplitude statistics and fluctuation persistence identification to generate fluctuation characteristics.

[0031] The reliable data is split into field time series data according to the field name. The field time series data is divided and aggregated according to fixed time windows, such as 5 group time windows, to form group data segments corresponding to the time windows. The maximum and minimum values ​​of each group data segment are read, and the difference between the maximum and minimum values ​​is calculated to obtain the fluctuation amplitude. The fluctuation judgment threshold is set according to the standard deviation of the historical stable segment corresponding to the field name and combined with the sensor performance parameters. The value range of the fluctuation judgment threshold is two to four times the standard deviation of the historical stable segment. The value range of the fluctuation judgment threshold is set according to the upper limit of the noise of the sensor performance parameters and the upper limit of the physical change rate under the sampling period. Using a value range of two to four times the standard deviation of the historical stable segment can identify continuous fluctuations earlier, which is suitable for field names that are sensitive to irrigation decisions and reduces the probability of missed detection. The fluctuation amplitude of adjacent group data segments is read and the fluctuation amplitude is compared with the fluctuation judgment threshold. The number of time windows in which the fluctuation amplitude continuously exceeds the fluctuation judgment threshold is read to obtain the fluctuation persistence. The fluctuation amplitude and fluctuation persistence are aggregated and organized according to the field name to form the fluctuation characteristics.

[0032] S1.4: Perform stability grading and aggregation on fluctuation characteristics to generate data quality indicators.

[0033] The fluctuation amplitude and fluctuation duration are combined into a fluctuation intensity value by field name. The fluctuation intensity values ​​of the same field name within a continuous time period are sorted and classified into stability levels according to quantiles. For example, if the continuous time period is the most recent day or the most recent week, the stability level is determined to be stable when the fluctuation intensity value is in the low quantile and unstable when the fluctuation intensity value is in the high quantile. The stability levels are aggregated by field name to obtain the data quality index.

[0034] S2: Divide the irrigation area into multiple zones and establish a zone index to generate a zone set. Perform multi-source data fusion on the zone set based on reliable data and data quality indicators to generate zone moisture status.

[0035] S2.1: Determine the zoning boundaries by analyzing the topology of the irrigation network and the valve arrangement in the irrigation area, and then divide the area to generate zoning description information.

[0036] The irrigation network topology and valve layout relationships of the irrigation area are read, and the main pipes, branch pipes, branch nodes, and terminal nodes are listed. The network node number corresponding to each valve is read through the valve layout relationship. When the valve layout relationship does not include the network node number, the installation position coordinates of the valve are read, and the network node matching the installation position coordinates is located in the irrigation network topology. A correspondence is established between the valve and the located network node, and the valve mark is recorded in the irrigation network topology to obtain the network topology with valve marks. In the network topology with valve marks, the valve control range is used as the candidate partition unit. The independent valve control ranges are identified along the water supply path from the main pipe to the branch pipe to form the connection position between the candidate partition units, which serves as the partition boundary. The plot boundary records of the irrigation area are matched and organized with the network topology with valve marks through the partition boundary. Specifically, the plot boundary segments covered by the same candidate partition unit are collected to form the partition boundary line. The irrigation area is divided into multiple partition areas according to the partition boundary line. For each partition area, the partition boundary line, corresponding valve mark, corresponding water supply path, and partition area information are registered and organized to form the partition description information.

[0037] S2.2: Register the partition identifier and partition index for the partition description information, and generate a partition index table; combine the partition index table and partition description information to generate a partition set.

[0038] The partition description information is read, and multiple partition areas are numbered one by one according to the partition boundary line. A unique partition identifier is registered for each partition area, and the correspondence between the partition identifier, the partition boundary line, and the corresponding valve mark is recorded. The partition index is assigned to the partition identifier by the control order of the water supply path and the valve arrangement relationship of the irrigation pipeline topology. The partition identifier and the partition index are written into the index record in pairs to form a partition index table. The partition index on the same water supply path is sequentially increased according to the water supply direction. The partition index table and the partition description information are associated and aggregated according to the partition identifier to obtain a set of partitions containing partition identifier, partition index, partition boundary line, corresponding valve mark, corresponding water supply path, and partition area information.

[0039] S2.3: Extract the trusted data subsets corresponding to each partition from the partition set and aggregate them to generate the partition trusted data subset.

[0040] Read the partition boundary line and corresponding valve marker for each partition index in the partition set. Read the acquisition location identifier or device identifier corresponding to each data entry in the trusted data. When the trusted data contains acquisition location coordinates, draw a ray along a fixed direction from the acquisition location coordinates. Determine whether the ray intersects with the boundary line segment formed by connecting two adjacent points in the boundary point sequence, and count the number of intersections. If the number of intersections is odd, it is determined that the acquisition location coordinates are within the area enclosed by the partition boundary line, and the corresponding data is marked as trusted data of the corresponding partition index. The trusted data contains valve markers, water supply path markers, and pump station branch markers carried by each data entry. When the trusted data carries valve markers, the valve markers of the trusted data are associated with the partition index. The corresponding valve tags of each partition index in the set are compared one by one. When the valve tags match, the partition index corresponding to the trusted data is determined, and the trusted data is marked as trusted data of the corresponding partition index. When the trusted data carries water supply path tags or pump station branch tags, the water supply path tags or pump station branch tags are compared one by one with the corresponding water supply paths of each partition index in the partition set. When the comparison matches, the partition index corresponding to the trusted data is determined, and the trusted data is marked as trusted data of the corresponding partition index. The trusted data marked as the same partition index are aggregated in ascending order of the timestamp field to form a subset of trusted data corresponding to the partition index. The subsets of trusted data corresponding to multiple partition indexes are merged to obtain the partition trusted data subset.

[0041] S2.4: Perform weight allocation on the reliable data subset of the partition through data quality indicators to generate fusion weights; use the fusion weights as constraints to perform weighted filtering on the reliable data subset of the partition to generate the partition moisture status quantity.

[0042] The reliable subset of partitioned data is organized according to partition index and field name to obtain partitioned field time series data. Fields with a stable stability level in the data quality indicators are assigned a stable weight. The stable weight is set based on the fluctuation amplitude and duration in the fluctuation characteristics, for example, 0.8. Fields with an unstable stability level are assigned an unstable weight. The ratio between the stable and unstable weights is set based on the impact of the noise upper bound and drift risk in the sensor performance parameters, for example, the stable weight is four times the unstable weight, i.e., the unstable weight is 0.2. When there are multiple source values ​​under the same timestamp field, the ratio of the stable weight, the sum of the stable weight and the unstable weight under the same timestamp field is calculated to obtain a normalized weight, which is used as the fusion weight. The fusion weight and the partitioned field time series data are read. Multiple source values ​​of the partitioned field time series data under the same timestamp field are weighted and summed according to the fusion weight to obtain a weighted value. The weighted value is then weighted and averaged in time order to obtain a smoothed value. The smoothed value is then aggregated according to the partition index to form the partition moisture state quantity.

[0043] S3: Infer water demand from the regional moisture status data and environmental datasets to generate regional demand intervals. Then, correlate these regional demand intervals with data quality indicators to generate the basis for determining irrigation patterns.

[0044] S3.1: Extract environmental factors from the environmental dataset, perform time-series feature extraction on the environmental factors, and generate environmental trend information; perform water demand prediction on the environmental trend information and the regional water state quantity, and generate regional water demand estimation results.

[0045] Environmental factors related to crop water consumption are selected from the environmental dataset. These environmental factors include the names of the temperature, relative humidity, wind speed, and rainfall fields from meteorological monitoring records. The environmental factors are aggregated according to the timestamp field to obtain time series data. The time series data of environmental factors is divided into multiple window segments according to a fixed time window, such as one hour. The mean, maximum, minimum, and rate of change of each window segment are read and aggregated to form environmental trend information. The regional water status quantities are organized into regional water status time series data according to the regional index. The environmental trend information and the regional water status time series data are aligned and combined according to the timestamp field. The water demand conversion factor is set according to the correspondence between irrigation water volume and post-irrigation water change in historical irrigation response. The evapotranspiration enhancement is determined according to the environmental trend information, and the water consumption of the regional water status time series data is corrected by the evapotranspiration enhancement. The corrected water consumption is multiplied by the water demand conversion factor to obtain the regional water demand estimation result.

[0046] S3.2: Based on the soil characteristics and long-term climate data in the environmental dataset, set the upper and lower limits and confidence levels of the water demand estimation results for each region, and express the water demand estimation results for each region in interval form to generate the regional demand interval.

[0047] Read soil characteristics and long-term climate data from the environmental dataset. Soil characteristics include soil texture, field capacity, and permeability coefficient fields. Long-term climate data includes multi-year seasonal rainfall statistics and evapotranspiration statistics. Determine the water demand fluctuation coefficient based on soil characteristics. Sandy soil texture is directly proportional to the fluctuation coefficient. Determine the climate uncertainty coefficient based on the long-term climate data. List the seasonal rainfall series for the past ten years by year. Calculate the mean and standard deviation from the seasonal rainfall series, and use the ratio of the standard deviation to the mean as the rainfall variability coefficient. The climate uncertainty coefficient is between 1 and 2 times the rainfall variability coefficient. For example, when the rainfall variability coefficient is 0.2 and the multiple is 1.5, the climate uncertainty coefficient is... With an uncertainty coefficient of 0.3, the fluctuation of rainfall in the same season over many years is proportional to the climate uncertainty coefficient. The interval width coefficient is obtained by combining the water demand fluctuation amplitude coefficient and the climate uncertainty coefficient. The interval width coefficient is then multiplied by the regional water demand estimation result to obtain the interval expansion, which is 10%-30% of the regional water demand estimation result. The difference between the regional water demand estimation result and the interval expansion is used as the lower limit of the regional water demand estimation result, and the sum of the regional water demand estimation result and the interval expansion is used as the upper limit of the regional water demand estimation result. The confidence level is obtained by converting the interval width coefficient, with a confidence level of 70%-95%. The upper and lower limits of the regional water demand estimation result and the confidence level are combined according to the regional index to form the regional demand interval.

[0048] S3.3: Extract partition demand features and fluctuation level features from the partition demand range and data quality indicators respectively, prioritize the partition demand features and fluctuation level features, generate judgment rule items, and compile the judgment rule items to form the basis for irrigation mode judgment.

[0049] The difference between the upper and lower limits of the partitioned demand interval is calculated to obtain the interval width. The mean of the lower and upper limits of the partitioned demand interval is calculated as the demand center value. The demand center value, interval width, and confidence level are combined to form the partitioned demand feature. The stability level in the data quality index is converted into a volatility level feature. When the stability level is stable, the volatility level feature is assigned a low volatility value. When the stability level is unstable, the volatility level feature is assigned a high volatility value. The partitioned demand feature and volatility level feature are associated and organized according to the partition index. The demand center values ​​are sorted from high to low. When the demand center values ​​are the same, they are sorted according to the volatility level feature. First, the partition indexes with high volatility are sorted, and then the partition indexes with low volatility are sorted. The partition priority is registered for each partition index according to the sorting result. The partition priority, partitioned demand feature, and volatility level feature are combined to form multiple judgment items. The multiple judgment items are combined to form a judgment rule item. The judgment rule item is combined to obtain the basis for determining the irrigation mode.

[0050] S4: Based on data quality indicators, perform gating decisions on irrigation mode determination criteria, generate a set of zonal target irrigation instructions, perform online infiltration capacity identification on the historical irrigation responses in the zonal target irrigation instruction set and environmental dataset, and generate pulse irrigation sequences.

[0051] S4.1: Set a quality judgment threshold based on historical fluctuation levels, missing drift characteristics, and sensor performance parameters. Compare the data quality indicators with the quality judgment threshold to generate gating discrimination results.

[0052] A quality judgment threshold is set based on the standard deviation of historical stable segments, combined with the missing rate, the proportion of drift segments, and sensor accuracy. The quality judgment threshold ranges from 0.6 to 0.8. The range of the quality judgment threshold is further determined by the distribution of historical stable segments within the comprehensive quality score interval and the acceptable upper limit of the missing rate and the proportion of drift segments. Using a range of 0.6-0.8 ensures the stability of the gating results when there are short-term fluctuations in the data, and also allows for timely transition to a restricted state when missing data and drift accumulate, reducing the risk of misjudgment due to data distortion. Data quality indicators are read and a comprehensive quality score is obtained. The comprehensive quality score is compared with the quality judgment threshold. When the comprehensive quality score is greater than or equal to the quality judgment threshold, the gating result is determined to be in a usable state; when the comprehensive quality score is less than the quality judgment threshold, the gating result is determined to be in a restricted state.

[0053] S4.2: When the gating judgment result shows that the data quality is in an available state, the irrigation mode is determined to be the normal irrigation mode by the judgment rule item in the irrigation mode judgment basis, and the midpoint of the partition demand interval is selected as the normal target demand quantity; when the gating judgment result shows that the data quality is in a restricted state, the irrigation mode is determined to be the degraded irrigation mode by the judgment rule item in the irrigation mode judgment basis, and the lower limit of the partition demand interval is selected as the degraded target demand quantity.

[0054] The gating results are distinguished into available and restricted states based on the partition index. When the gating result is available, the irrigation mode determination criteria are read, and the partition priority and fluctuation level characteristics in the determination rule items are extracted according to the partition index. The irrigation mode is determined to be the normal irrigation mode based on the partition priority and fluctuation level characteristics. The average of the lower and upper limits of the partition demand interval is calculated as the interval midpoint and the interval midpoint is registered as the normal target demand. Similarly, when the gating result is restricted, the partition priority and fluctuation level characteristics are extracted from the determination rule items in the irrigation mode determination criteria according to the partition index. The degraded irrigation mode is preferentially adopted for partition indexes with high fluctuation level characteristics, and the degraded irrigation mode is continued to be adopted according to the partition priority for partition indexes with low fluctuation level characteristics. The lower limit of the partition demand interval is registered as the degraded target demand. The normal target demand, degraded target demand, and partition index are combined to generate the partition target irrigation instruction set.

[0055] S4.3: Collect historical irrigation responses in the environmental dataset by partition index to generate partition historical response sequences; match the partition historical response sequences with the partition target irrigation instruction set, and extract the water change features before and after irrigation to generate an identification sample set.

[0056] Historical irrigation responses are retrieved from the environmental dataset. These responses include partition indexes, irrigation start times, irrigation end times, and irrigation water volume records. The historical irrigation responses are then aggregated in ascending order by timestamp field according to the partition index, forming a partitioned historical response sequence. The partitioned target irrigation instruction set is read, and its partition index and target demand are obtained. The partitioned target irrigation instruction set and the partitioned historical response sequence are then correlated according to the partition index. The irrigation water volume record corresponding to the target demand, along with the corresponding irrigation start and end times, are located within the partitioned historical response sequence to obtain the response matching results. The pre-irrigation time window is located in the field time series data using the irrigation start time, and the pre-irrigation average water content is calculated. Similarly, the post-irrigation time window is located in the field time series data using the irrigation end time, and the post-irrigation average water content is calculated. The difference between the post-irrigation average water content and the pre-irrigation average water content is used as the water content change feature. The target demand, irrigation water volume records, irrigation start time, irrigation end time, and water content change feature are then aggregated and organized according to the partition index to form an identification sample set.

[0057] S4.4: Calculate the infiltration characteristic parameters of the partition based on the identified sample set, and determine the duration and interval of the pulse through the infiltration characteristic parameters of the partition, and collect them to form a partition pulse parameter set; arrange the partition target irrigation instruction set according to the partition pulse parameter set to generate a pulse irrigation sequence.

[0058] The irrigation water volume records and water change characteristics in the identification sample set are read. The ratio of the irrigation water volume records to the water change characteristics is used as the infiltration value. The mean and dispersion of the infiltration values ​​are extracted within the same partition index and combined to obtain the partition infiltration characteristic parameters. The pulse duration and pulse interval of a single irrigation are determined by the partition infiltration characteristic parameters. The partition infiltration characteristic parameters indicate that the smaller the infiltration value, the shorter the pulse duration and the longer the pulse interval. Conversely, the larger the infiltration value, the longer the pulse duration and the shorter the pulse interval. For example, the pulse duration is five to twenty minutes, and the pulse interval is ten to forty minutes. The partition index, The pulse duration and pulse interval duration are combined to form a partitioned pulse parameter set. The target demand in the partitioned target irrigation instruction set is converted into a single pulse water volume according to the pulse duration, resulting in the number of pulses. Pulse numbers are listed according to the number of pulses, and the single pulse water volume and pulse duration are recorded for each pulse number. The pulse duration and pulse interval duration are accumulated segment by segment according to the start time of the first pulse to obtain the start time and end time corresponding to each pulse number. The pulse number, start time, end time and single pulse water volume are combined to form multiple pulse execution entries. Pulse interval durations are inserted between adjacent pulse execution entries. The multiple pulse execution entries are combined and organized in chronological order to form a pulse irrigation sequence.

[0059] S5: Perform hydraulic state analysis on the pipeline flow and pressure and pump station operation data in the environmental dataset, generate water supply constraints, schedule the pulse irrigation sequence based on the water supply constraints, and send it to valves and pump stations to generate irrigation execution results.

[0060] S5.1: Align the pipeline flow and pressure data with the pump station operation data in the environmental dataset with the time data and process the operating conditions to generate hydraulic operating condition data.

[0061] The timestamp fields of pipeline flow and pressure data and pump station operation data in the environmental dataset are converted into a unified year-month-day-hour-minute-second timestamp format. The pipeline flow and pressure data and pump station operation data are then sorted in ascending order by timestamp field to obtain sorted records. These sorted records are then matched against the target timeline point by point. If a record for a corresponding time point exists on the target timeline, the names and values ​​of the flow, pressure, pump station frequency, and operation status fields for that time point are retrieved. If a record for a corresponding time point is missing from the target timeline, the most recent time point value is used to pad the record and write it to the corresponding time point on the target timeline. The time-aligned records are obtained by dividing them into operating segments, stopping segments, and switching segments according to the pump station operating status field name. The pressure, flow, and pump station frequency field names in the operating segments are aggregated to form operating condition data. The pressure, flow, and pump station frequency field names in the stopping segments are aggregated to form stopping condition data. The pressure, flow, and pump station frequency field names in the switching segments are aggregated to form switching condition data. The operating condition data, stopping condition data, and switching condition data are then aggregated and organized to obtain hydraulic condition data.

[0062] S5.2: Determine the core nodes of the pipeline network by the location of pressure monitoring devices and flow monitoring devices on the pipeline network, identify the pressure level and flow distribution status of the core nodes of the pipeline network based on hydraulic condition data, determine and aggregate the risks of insufficient pressure and flow congestion, and generate risk markers.

[0063] The deployment location identifiers of pressure monitoring devices and flow monitoring devices are obtained from the pipeline network. A correspondence is established between these location identifiers and the pipeline node numbers in the irrigation pipeline network topology. Pipeline node numbers with corresponding pressure and flow monitoring device numbers are registered as core nodes of the pipeline network, and a core node list is formed. The pressure field name and flow field name values ​​of each core node at each time point are extracted according to the core node list, and a core node hydraulic sequence is formed by the timestamp field. The minimum and average values ​​of the pressure field name of the core nodes are statistically analyzed through the core node hydraulic sequence. The minimum and average values ​​of the pressure field name of the core nodes are used as the pressure level, and the average and peak values ​​of the flow field name of the core nodes are used as the flow distribution status. The minimum allowable pressure and maximum allowable flow corresponding to each core node in the pipeline network design parameters are obtained. When the pressure level is lower than the minimum allowable pressure, the pressure insufficiency risk is registered. When the flow distribution status exceeds the maximum allowable flow, the flow congestion risk is registered. The pressure insufficiency risk and flow congestion risk are aggregated according to the timestamp field and the core node list to form a risk marker.

[0064] S5.3: Determine the available water supply capacity based on pump station operation data, and limit the number of zones that can be irrigated concurrently and the available water supply intensity through the available water supply capacity and risk markers; combine the number of zones that can be irrigated concurrently and the available water supply intensity to form a water supply constraint.

[0065] The pump station operation status and frequency fields are read from the pump station operation data. Records with the operation status field name indicating operation are selected and aggregated by timestamp field to form an operational pump station sequence. Pump station design parameters are obtained, and rated flow and rated frequency parameters are extracted from these parameters. The pump station frequency field name at each time point in the operational pump station sequence is compared with the rated frequency parameter to obtain a frequency ratio coefficient. The rated flow parameter is multiplied by the frequency ratio coefficient to obtain the water supply flow capacity at that time point, and aggregated by timestamp field to form the available water supply capacity. Pressure is extracted from risk markers and by timestamp field. When the risks of insufficient pressure and flow congestion exist, the available water supply capacity is reduced to obtain the available water supply intensity. When the risk of flow congestion exists, the available water supply intensity is adjusted to obtain the available water supply strength. When neither the risk of insufficient pressure nor the risk of flow congestion exists, the available water supply capacity is registered as the available water supply intensity. The allocation base of a unit partition is set by using the available water supply intensity and the partition area information in the partition set. The available water supply intensity is compared with the allocation base and the integer is taken to obtain the number of partitions that can be irrigated concurrently. The number of partitions that can be irrigated concurrently and the available water supply intensity are aggregated by the timestamp field to form the water supply constraint.

[0066] The partition index, pulse number, start time, end time, and single pulse water volume are extracted from the pulsed irrigation sequence. The pulse execution entries within the same time period are sorted according to the partition priority in the irrigation mode determination criteria. The number of concurrently executed partition indexes is counted within each time period and compared with the number of partitions that can be concurrently irrigated. If the number of concurrently executed partition indexes exceeds the number of partitions that can be concurrently irrigated, the start and end times of the pulse execution entry are postponed to the next available time period. If the number of concurrently executed partition indexes does not exceed the number of partitions that can be concurrently irrigated, the start and end times remain unchanged. The single pulse water volume of the concurrently executed entries is summarized within each time period to obtain the water supply demand intensity for that time period. The water supply demand intensity for that time period is compared with the available water supply intensity. If the water supply demand intensity for that time period exceeds the available water supply intensity, the ratio of the available water supply intensity to the water supply demand intensity for that time period is used as the minimum water supply intensity. The reduction ratio is used to proportionally reduce the single pulse water volume of each pulse execution item while keeping the pulse duration unchanged. For example, the reduction ratio is 0.7-0.9. The valve marker and water supply path corresponding to the partition index are read from the partition set. The pulse irrigation sequence after scheduling is converted into valve opening and closing instructions and pump station operation instructions one by one and written into the control issuance record. The valve opening and closing instructions include the partition index, valve marker, valve opening time and valve closing time. The pump station operation instructions include the pump station frequency field name and operation status field name corresponding to the timestamp field. After the control issuance record is issued to the valves and pump stations, the valve status field name, pump station operation status field name and pipeline flow and pressure monitoring records are extracted from the environmental dataset. The actual valve opening time, actual valve closing time, actual pump station frequency and actual flow and pressure are collected according to the timestamp field and matched with the pulse irrigation sequence after scheduling to obtain the irrigation execution result.

[0067] This embodiment also provides a water-saving precision irrigation control system, including: The data quality module is used to acquire irrigation area and environmental datasets, preprocess the environmental datasets to generate reliable data, perform volatility assessment on the reliable data, and generate data quality indicators. The partition index module is used to divide the irrigation area into multiple partitions and establish partition indexes, generate a partition set, perform multi-source data fusion on the partition set based on reliable data and data quality indicators, and generate partition moisture status. The irrigation mode module is used to infer water demand from the regional water status and environmental dataset, generate regional demand intervals, and associate the regional demand intervals with data quality indicators to generate the basis for irrigation mode determination. The gating decision module is used to perform gating decisions based on data quality indicators to determine irrigation patterns, generate a set of zonal target irrigation instructions, perform online identification of infiltration capacity on the historical irrigation responses in the zonal target irrigation instruction set and environmental dataset, and generate pulse irrigation sequences. The irrigation execution module is used to perform hydraulic state analysis on the pipeline flow and pressure and pump station operation data in the environmental dataset, generate water supply constraints, schedule the pulse irrigation sequence based on the water supply constraints, and send it to valves and pump stations to generate irrigation execution results.

[0068] This embodiment also provides a computer device applicable to the water-saving precision irrigation control method, including: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the water-saving precision irrigation control method proposed in the above embodiment.

[0069] The computer device can be a terminal, comprising a processor, memory, communication interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, carrier networks, NFC (Near Field Communication), or other technologies. The display screen can be an LCD screen or an e-ink screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad on the computer device's casing, or an external keyboard, touchpad, or mouse.

[0070] This embodiment also provides a storage medium storing a computer program, which, when executed by a processor, implements the water-saving precision irrigation control method proposed in the above embodiments. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Red-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0071] In summary, this invention improves control robustness and reduces the risk of mis-irrigation by generating data quality indicators; reduces the influence of single-point noise and drift on the regional water status by generating regional water status metrics; reduces human experience intervention by generating irrigation mode determination criteria, making the decision-making chain clearer and more interpretable; avoids over-irrigation and maintains safe water supply even when data is abnormal by determining normal irrigation mode and degraded irrigation mode; generates pulsed irrigation sequences to make the irrigation rhythm match the regional infiltration characteristics and reduce runoff and deep infiltration caused by large volumes of water at once; and improves the executability of the strategy and reduces field deviations by forming water supply constraints.

[0072] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A water-saving precision irrigation control method, characterized in that: include, Acquire irrigation area and environmental datasets, preprocess the environmental datasets to generate reliable data, perform volatility assessment on the reliable data, and generate data quality indicators. The irrigation area is divided into multiple zones and a zone index is established to generate a zone set. Based on reliable data and data quality indicators, multi-source data fusion is performed on the zone set to generate zone moisture status. Water demand is inferred from the regional moisture state data and environmental datasets to generate regional demand intervals. These regional demand intervals are then correlated with data quality indicators to generate criteria for determining irrigation patterns. Based on data quality indicators, gating decisions are made to determine irrigation patterns, generating a set of zonal target irrigation instructions. The infiltration capacity of the historical irrigation responses in the zonal target irrigation instruction set and environmental dataset is identified online, and a pulse irrigation sequence is generated. Hydraulic state analysis is performed on the pipeline flow and pressure and pump station operation data in the environmental dataset to generate water supply constraints. Based on the water supply constraints, the pulse irrigation sequence is scheduled and sent to valves and pump stations to generate irrigation execution results.

2. The water-saving precision irrigation control method as described in claim 1, characterized in that: The specific steps for generating the data quality indicators are as follows: Obtain irrigation area and environmental datasets, perform time alignment and field consistency processing on the environmental dataset, and generate aligned data; The aligned data is filled with missing data and anomaly removal is performed to generate clean data. Drift identification and correction are then performed to generate reliable data. Perform data grouping processing on reliable data to generate grouped data fragments, and perform fluctuation amplitude statistics and fluctuation persistence identification to generate fluctuation characteristics; Stability grading and aggregation are performed on fluctuation characteristics to generate data quality indicators.

3. The water-saving precision irrigation control method as described in claim 2, characterized in that: The specific steps for generating the partition set are as follows. The zoning boundaries are determined by the relationship between the topology of the irrigation pipeline network and the valve layout in the irrigation area, and the area is divided to generate zoning description information; Register the partition identifier and partition index for the partition description information, and generate a partition index table; The partition index table and partition description information are combined to generate a partition set.

4. The water-saving precision irrigation control method as described in claim 3, characterized in that: The specific steps for generating the regional moisture state data are as follows. Extract trusted data subsets corresponding to each partition from the partition set and aggregate them to generate partition trusted data subsets; Weights are assigned to trusted subsets of data in partitions based on data quality metrics to generate fusion weights. Using the fusion weight as a constraint, a weighted filtering is performed on the reliable subset of partitioned data to generate the partitioned moisture status.

5. The water-saving precision irrigation control method as described in claim 4, characterized in that: The specific steps for generating the required partition range are as follows. Environmental factors are extracted from the environmental dataset, and time-series feature extraction is performed on the environmental factors to generate environmental trend information; Perform water demand forecasting on environmental trend information and regional water state quantities to generate regional water demand estimation results; Based on soil characteristics and long-term climate data in the environmental dataset, upper and lower limits and confidence levels of the regional water demand estimation results are set, and the regional water demand estimation results are expressed in interval form to generate regional demand intervals.

6. The water-saving precision irrigation control method as described in claim 5, characterized in that: The criteria for determining irrigation patterns refer to extracting partition demand features and fluctuation level features from the partition demand range and data quality indicators, prioritizing the partition demand features and fluctuation level features, generating judgment rule items, and compiling the judgment rule items to form the criteria for determining irrigation patterns.

7. The water-saving precision irrigation control method as described in claim 6, characterized in that: The specific steps for generating the target irrigation instruction set for the designated area are as follows: Based on historical fluctuation levels, missing drift characteristics, and sensor performance parameters, a quality judgment threshold is set. The data quality indicators are compared with the quality judgment threshold to generate gating discrimination results. When the gating judgment result shows that the data quality is available, the irrigation mode is determined to be the normal irrigation mode by the judgment rule item in the irrigation mode judgment basis, and the median value of the interval is selected from the partition demand interval as the normal target demand. When the gating judgment result shows that the data quality is in a limited state, the irrigation mode is determined to be a degraded irrigation mode by the judgment rule item in the irrigation mode judgment basis, and the lower limit value of the interval is selected from the partition demand interval as the degraded target demand. The normal target demand, the degraded target demand, and the partition index are combined to generate a partition target irrigation instruction set.

8. The water-saving precision irrigation control method as described in claim 7, characterized in that: The specific steps for generating the pulsed irrigation sequence are as follows: The historical irrigation responses in the environmental dataset are aggregated according to the partition index to generate a partitioned historical response sequence. The historical response sequence of the partition is matched with the target irrigation instruction set of the partition, and the water change features before and after irrigation are extracted to generate an identification sample set. The partition infiltration characteristic parameters are calculated based on the identified sample set, and the duration and interval of the pulse are determined by the partition infiltration characteristic parameters, and then the parameters are aggregated to form a partition pulse parameter set. The target irrigation instruction set for each zone is pulsed according to the zone pulse parameter set to generate a pulse irrigation sequence.

9. The water-saving precision irrigation control method as described in claim 8, characterized in that: The specific steps for generating water supply constraints are as follows: The pipeline flow and pressure data in the environmental dataset are time-aligned and the pump station operation data are processed to generate hydraulic condition data. The core nodes of the pipeline network are determined by the location of pressure monitoring devices and flow monitoring devices on the pipeline network. Based on hydraulic condition data, the pressure level and flow distribution status of the core nodes of the pipeline network are identified, and the risks of insufficient pressure and flow congestion are determined, aggregated, and risk markers are generated. Based on pump station operation data, the available water supply capacity is determined, and the number of zones that can be irrigated concurrently and the available water supply intensity are determined by the available water supply capacity and risk markers. The number of zones that can be irrigated concurrently and the available water supply intensity are combined to form a water supply constraint.

10. A water-saving precision irrigation control system, based on the water-saving precision irrigation control method according to any one of claims 1 to 9, characterized in that: include, The data quality module is used to acquire irrigation area and environmental datasets, preprocess the environmental datasets to generate reliable data, perform volatility assessment on the reliable data, and generate data quality indicators. The partition index module is used to divide the irrigation area into multiple partitions and establish partition indexes, generate a partition set, perform multi-source data fusion on the partition set based on reliable data and data quality indicators, and generate partition moisture status. The irrigation mode module is used to infer water demand from the regional water status and environmental dataset, generate regional demand intervals, and associate the regional demand intervals with data quality indicators to generate the basis for irrigation mode determination. The gating decision module is used to perform gating decisions based on data quality indicators to determine irrigation patterns, generate a set of zonal target irrigation instructions, perform online identification of infiltration capacity on the historical irrigation responses in the zonal target irrigation instruction set and environmental dataset, and generate pulse irrigation sequences. The irrigation execution module is used to perform hydraulic state analysis on the pipeline flow and pressure and pump station operation data in the environmental dataset, generate water supply constraints, schedule the pulse irrigation sequence based on the water supply constraints, and send it to valves and pump stations to generate irrigation execution results.

Images