AI-based irrigation district water resource multi-scale dynamic allocation method and system

Abstract

The invention discloses an AI-based irrigation district water resource multi-scale dynamic allocation method and system, and relates to the technical field of irrigation district water resource allocation, and the method comprises the steps: monitoring each field in an irrigation district, evaluating the terrain difference degree and the irrigation difference degree, judging whether the adjustment is needed, matching the field needing to be adjusted with historical records in a database, and carrying out the adjustment. The method comprises the following steps: calculating prediction stability, screening optimal adjustment data and executing the optimal adjustment data, monitoring in regulation and control, judging whether secondary allocation is needed or not, aiming at a field parcel subjected to secondary allocation, positioning an abnormal region, matching historical secondary adjustment cases, calculating optimal secondary adjustment data and executing the optimal secondary adjustment data, and then performing secondary adjustment monitoring and evaluation feedback, so that the adaptability of allocation and terrain is improved; the water distribution uniformity in the field is ensured, the growth effect of crops in the field is improved, the dynamic adjustment is ensured to accurately respond to the demand change of the local field, the actual demand is met, and the flexibility and accuracy of the dynamic adjustment are improved.

Description

Technical Field

[0001] This invention relates to the field of irrigation district water resource allocation technology, specifically to an AI-based multi-scale dynamic allocation method and system for irrigation district water resources. Background Technology

[0002] Dynamic allocation of irrigation water resources refers to a scheme that dynamically adjusts the amount, timing, and regional distribution of irrigation water based on real-time supply and demand data, such as topography and uneven irrigation conditions. Changes in field topography and uneven irrigation can directly disrupt the coordinated balance of water, crops, and soil, leading to decreased irrigation efficiency and resource waste. Therefore, targeted adjustments are needed to achieve precision irrigation.

[0003] Existing technologies, such as the multi-source joint allocation method for oasis agricultural irrigation areas with complex canal systems disclosed in application CN118297331A, firstly simulate the typical annual water inflow process of high, average, low, and extremely low water years based on the P-III frequency analysis method using long-term series of water intake data. Secondly, based on the system coupling principle, the method couples the distribution of irrigation system engineering with the calculation process of hydraulic coefficients, and calculates the net and gross water requirements of each crop in the metering unit according to the planting plan and irrigation system. Finally, based on the self-iterative optimization principle, the method allocates water resources in each metering unit according to the water balance of the entire process of water demand, groundwater quality, and water allocation principles, with the goal of minimizing the degree of water demand disruption throughout the entire process. This method has the advantages of strong robustness, easy global convergence, and fast calculation speed, and it fits well with the actual water allocation process. The calculation results can be used to guide the scientific water allocation in irrigation area agriculture.

[0004] Existing technologies, such as the real-time dynamic water allocation method and system for multi-level canals in irrigation districts disclosed in application CN118521431A, include: acquiring basic data and water supply and demand information of the irrigation district; setting an irrigation priority level division strategy based on important water diversion gates; setting a dynamic water allocation method with "variable flow and variable duration" based on water balance and flow constraints; constructing a database table structure for a real-time dynamic water allocation model of the irrigation district; constructing a multi-level canal adaptive real-time dynamic water allocation model based on basic data, water supply and demand information, irrigation priority level division strategy, dynamic water allocation method, and constructed database table structure; and running the real-time dynamic water allocation model to simulate the real-time dynamic water allocation process of multi-level target canals in the irrigation district. This application proposal can quickly and accurately simulate the water allocation time, flow rate, and allocated water volume of each water diversion gate on the target canal, providing technical support for the scientific and rational formulation of dynamic water allocation schemes for canal systems at all levels in irrigation districts, and providing a scientific basis for promoting efficient agricultural water use.

[0005] The aforementioned scheme specifically discloses the optimization of water allocation at the canal system level, but it fails to incorporate changes in field topography and actual irrigation conditions into the control logic. This results in a significant deviation in the calculation of gross water demand in the water allocation scheme, causing the total water allocation to become disconnected from actual field needs. Spatial water allocation cannot adapt to the terrain, leading to uneven water distribution in the fields and reducing the growth effect of crops. Furthermore, the dynamic allocation of the aforementioned scheme is based solely on supply and demand information at the canal system level, without constructing a feedback loop between field topography and irrigation differences. This results in the dynamic adjustment failing to accurately respond to changes in demand at the field scale, causing water allocation adjustments to lag behind actual needs, reducing the flexibility and accuracy of dynamic adjustments, while also increasing water waste and reducing water resource utilization.

[0006] Real-time monitoring during water distribution and identification of hazardous areas, along with targeted adjustments based on these areas, can effectively ensure the effectiveness and safety of the distribution. The aforementioned scheme lacks monitoring and secondary adjustments during distribution, resulting in the water distribution scheme being unable to cope with dynamic disturbances during execution, unable to correct water distribution deviations in real time, reducing the control effect, and the lack of monitoring makes it impossible to identify risk areas, increasing the probability of risk spread and causing safety risks to get out of control. Summary of the Invention

[0007] To address the aforementioned technical shortcomings, the present invention aims to provide an AI-based method and system for multi-scale dynamic allocation of irrigation water resources.

[0008] To solve the above-mentioned technical problems, the present invention adopts the following technical solution: In the first aspect, the present invention provides an AI-based method for multi-scale dynamic allocation of irrigation water resources, including the following steps: S1, Irrigation area monitoring: monitoring the topography and irrigation of each field in the irrigation area, determining whether each field needs irrigation adjustment, and if at least one field needs irrigation adjustment, executing S2.

[0009] S2, First Adjustment: Record each field that needs irrigation adjustment as a target field, obtain the size and terrain data of each target field, and use the historical adjustment records of irrigation districts in the database to analyze the optimal adjustment data of each target field. Adjustment is carried out according to the optimal adjustment data, and the terrain and irrigation are monitored during adjustment. The effect of the first adjustment for each target field is analyzed. If at least one target field has a poor effect, S3 is executed.

[0010] S3. Secondary Adjustment: Mark the target fields with poor performance as marked fields, extract abnormal terrain data and abnormal irrigation data of deformed areas in each marked field, and use the historical adjustment records of irrigation districts in the database to confirm the secondary adjustment data of each marked field and send it to the irrigation district control center. The irrigation district control center performs corresponding adjustments, monitors the terrain and irrigation during the adjustment, analyzes the effect of the secondary adjustment on each marked field, and then performs corresponding operations based on the effect of the secondary adjustment.

[0011] Secondly, the present invention provides an AI-based multi-scale dynamic allocation system for irrigation district water resources, including: an irrigation district monitoring module, used to monitor the topography and irrigation of each field in the irrigation district, determine whether each field needs irrigation adjustment, and if at least one field needs irrigation adjustment, execute the allocation module once.

[0012] The first allocation module is used to mark each field that needs irrigation adjustment as a target field, and at the same time obtain the size and terrain data of each target field. It also uses the historical adjustment records of irrigation districts in the database to analyze the optimal adjustment data for each target field, and adjusts the field according to the optimal adjustment data. During the adjustment, the terrain and irrigation are monitored, and the effect of the first adjustment for each target field is analyzed. If at least one target field is not effective, the second allocation module is executed.

[0013] The secondary adjustment module is used to mark the target fields with poor performance as marked fields, extract abnormal terrain data and abnormal irrigation data of deformed areas in each marked field, and use the historical adjustment records of the irrigation district in the database to confirm the secondary adjustment data of each marked field and send it to the irrigation district control center. The irrigation district control center performs corresponding adjustments, monitors the terrain and irrigation during the adjustment, analyzes the effect of the secondary adjustment on each marked field, and then performs corresponding operations based on the effect of the secondary adjustment.

[0014] The beneficial effects of this invention are as follows: 1. This invention provides an AI-based multi-scale dynamic allocation method and system for irrigation district water resources. By monitoring each field in the irrigation district, assessing the degree of topographic and irrigation differences, and determining whether adjustments are needed, the fields requiring adjustment are matched with historical records in the database. By calculating the predicted stability, the optimal adjustment data is selected and executed. Monitoring is conducted during the adjustment process. By monitoring and analyzing the effect value of the first adjustment, it is determined whether a second adjustment is needed. For marked fields with poor first adjustment effects, abnormal areas are located and the deviation values ​​of topographic and irrigation data are calculated. Historical second adjustment cases are matched, the optimal second adjustment data is calculated and executed. After adjustment, second adjustment monitoring is conducted in conjunction with normal areas. If the effect is poor, manual intervention is prompted. If the effect is good, key monitoring is conducted. This improves the adaptability of allocation to topography, ensures the uniformity of water distribution in the fields, and improves the growth effect of crops in the fields. At the same time, real-time monitoring is carried out during adjustment to construct a feedback closed loop of field topography-irrigation differences, ensuring that dynamic adjustment accurately responds to the changes in the needs of local fields, so that the adjustment can meet the actual needs in a timely manner, improve the flexibility and accuracy of dynamic adjustment, and reduce the risks in the allocation process.

[0015] 2. The one-time allocation in this application is applicable to the allocation optimization of the entire field, and based on the reuse of historical data and AI quantitative analysis, it achieves accurate matching of anomaly type and adjustment data, improves the targeting of adjustment data matching, avoids one-size-fits-all regulation, and thus improves the scientific nature of decision-making.

[0016] 3. Secondary allocation precisely locates abnormal areas in marked fields, rather than controlling the entire field, reducing ineffective water resource consumption and equipment wear and tear. At the same time, the accuracy of allocation is improved by reusing historical cases and AI quantitative analysis.

[0017] 4. Both primary and secondary allocations are monitored in real time to achieve a closed-loop logic of monitoring-allocation-effect evaluation-iterative optimization. Before control, the equipment status is checked to avoid execution risks. During control, high-frequency monitoring is used for dynamic tracking. After control, quantitative evaluation is conducted and a manual intervention mechanism is linked to improve the control effect and ensure the safety of control. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. 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.

[0019] Figure 1 This is a schematic diagram of the implementation steps of the method of the present invention.

[0020] Figure 2 This is a schematic diagram of the system structure connection of the present invention. Detailed Implementation

[0021] 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.

[0022] See Figure 1 As shown, an AI-based method for multi-scale dynamic allocation of irrigation water resources includes the following steps: S1, Irrigation monitoring: Monitor the topography and irrigation of each field in the irrigation area, determine whether each field needs irrigation adjustment, and if at least one field needs irrigation adjustment, execute S2.

[0023] In a specific embodiment, the specific process of S1 is as follows: S1.1, Sensor devices are deployed in each field in the irrigation area, and the sensor devices are used to detect the terrain data and irrigation data of each field at preset time intervals to obtain the terrain data and irrigation data of each field at each time.

[0024] The preset duration for each interval of detection is set and adjusted by the supervisors according to the irrigation adjustment needs, and no specific numerical limit is specified here.

[0025] It should be noted that topographic data is quantitative data reflecting the surface morphology of the field, including topographic parameters of each monitoring point, such as slope and elevation.

[0026] Irrigation data are quantitative data reflecting the irrigation status of a field, including irrigation parameters at each monitoring point, such as soil moisture content and flow rate.

[0027] The sensor device includes several sensor devices or monitoring devices for collecting topographic and irrigation data, such as lidar sensors, tilt monitors, Doppler flow meters, and soil moisture sensors.

[0028] S1.2. Based on the topographic and irrigation data of each field at each time, analyze the change status of each field. If the change status is abnormal, irrigation adjustment is required. In this way, it is determined whether irrigation adjustment is required for each field.

[0029] Preferably, the specific process of analyzing the change status of each field is as follows: plot a time-topography line graph with time as the horizontal axis and topographic data as the vertical axis, extract the slope corresponding to each point on the line graph, and then select the maximum value of the difference between the slopes of each point as the topographic difference degree, thereby obtaining the topographic difference degree corresponding to each field.

[0030] The irrigation data of each field at each time point are used to calculate the irrigation difference of each field according to the method of calculating the topographic difference of each field.

[0031] The topographic and irrigation differences of each field are compared with the preset topographic and irrigation difference thresholds in the database. If the topographic or irrigation difference is greater than the corresponding threshold, it indicates an abnormal change. If the topographic and irrigation differences are less than or equal to the corresponding preset thresholds, it indicates a normal change. This is how the change status of each field is obtained.

[0032] The preset terrain difference threshold and irrigation difference threshold are critical values ​​used to determine whether terrain changes and irrigation differences are significant. These can be set and adjusted by the supervisors according to the irrigation adjustment needs, and no specific numerical limit is specified here.

[0033] The slope extraction in the above analysis can cover all data points of the line chart, ensuring that no change at any moment is missed. Selecting the maximum slope difference as the difference index can focus on the most significant changes and fluctuations. For example, if a sudden collapse causes a sudden change in the slope over a certain period of time, the maximum slope difference will increase significantly, avoiding misjudging small local fluctuations as anomalies. Both terrain anomalies and irrigation anomalies can affect irrigation effects. If they occur alone, adjustments are required. The "OR logic" is used to fit the actual situation of the irrigation area, while also enabling timely detection of anomalies and corresponding adjustments, thereby improving the efficiency of irrigation area management.

[0034] S2, First Adjustment: Record each field that needs irrigation adjustment as a target field, obtain the size and terrain data of each target field, and use the historical adjustment records of irrigation districts in the database to analyze the optimal adjustment data of each target field. Adjustment is carried out according to the optimal adjustment data, and the terrain and irrigation are monitored during adjustment. The effect of the first adjustment for each target field is analyzed. If at least one target field has a poor effect, S3 is executed.

[0035] In a specific embodiment, the specific process of S2 is as follows: S2.1, classify each target field into anomaly types, namely irrigation anomaly type, terrain anomaly type and composite anomaly type; extract the terrain data and irrigation data of each target field at the last moment as the current terrain data and current irrigation data.

[0036] The anomaly types are categorized as follows: if only the terrain difference is less than or equal to the preset terrain difference threshold, it is an irrigation anomaly; if only the irrigation difference is less than or equal to the irrigation difference threshold, it is a terrain anomaly; if both the terrain difference and the irrigation difference are less than or equal to the corresponding preset terrain difference threshold and irrigation difference threshold, respectively, it is a composite anomaly.

[0037] S2.2 Obtain the anomaly type, pre-adjustment topographic data, pre-adjustment irrigation data, post-adjustment topographic data, post-adjustment irrigation data, adjustment data, and adjustment effect data of the corresponding fields for each adjustment from the historical adjustment records of the irrigation district. Select the adjustments for which the anomaly type, pre-adjustment topographic data, and pre-adjustment irrigation data of the fields correspond to the same anomaly type, current topographic data, and current irrigation data of each target field as the target adjustments for each adjustment.

[0038] S2.3 Extract the pre-adjustment topographic data, pre-adjustment irrigation data, post-adjustment topographic data, post-adjustment irrigation data, adjustment data, and adjustment effect data of each target field for each target adjustment, select the best adjustment data for each target field, and send it to the irrigation district control center. Preferably, the process of selecting the best adjustment data for each target field is as follows: based on the pre-adjustment terrain data, pre-adjustment irrigation data, post-adjustment terrain data, post-adjustment irrigation data, adjustment data, and adjustment effect data of each target field for each adjustment, the pre-adjustment terrain data, pre-adjustment irrigation data, post-adjustment terrain data, post-adjustment irrigation data, and adjustment effect data of each target field when using each adjustment data are statistically analyzed.

[0039] In the above, the adjustment data is a set of quantitative parameters for the first adjustment of the irrigation system for the field. It includes the execution instructions that directly affect the irrigation equipment and the time logic of the irrigation action. The execution instructions that directly affect the irrigation equipment include, for example, the opening percentage of the main gate of the canal and the branch gate of the field, as well as the flow rate of the canal and the branch canal of the field. The time logic of the irrigation action includes, for example, "open the main gate first, then open the branch gate after a 10-second delay" and "adjust the opening in three steps".

[0040] Using the pre-adjustment topographic data, pre-adjustment irrigation data, post-adjustment topographic data, and post-adjustment irrigation data of each target field when using each adjustment data, calculate the adjustment stability of each target field when using each adjustment data.

[0041] It should be noted that the terrain data and irrigation data before each target adjustment are the last terrain data and irrigation data collected before the adjustment; the terrain data and irrigation data after each target adjustment are the last terrain data and irrigation data collected before the end of the adjustment.

[0042] The calculation process for the stability adjustment is as follows: Construct a line graph of the terrain data before and after each adjustment for each target field when using each adjustment data. The slope between the terrain data before and after the adjustment is the rate of change of terrain before and after each adjustment for each target field when using each adjustment data. Then, select the maximum difference from the differences between the rates of change of terrain before and after each adjustment for each target field when using each adjustment data as the degree of change of terrain before and after each adjustment for each target field.

[0043] Similarly, each target field is constructed separately with a line graph of the adjusted terrain data after each adjustment when using each adjustment data. The slope of each point in the line graph is extracted, and the maximum difference between the slopes of each point is selected as the degree of terrain change after each adjustment when using each adjustment data for each target field.

[0044] Based on the analysis methods of changes in topography before and after irrigation and changes in topography after adjustment, the changes in irrigation before and after irrigation and changes in irrigation after adjustment are obtained.

[0045] The changes in topography before and after, the changes in topography after adjustment, the changes in irrigation before and after, and the changes in irrigation after adjustment are normalized to the range [0, 1] and denoted as a1, a2, a3, and a4, respectively.

[0046] Adjust stability = 1 - 0.25 × (a1 + a2 + a3 + a4).

[0047] The core of the above-mentioned stability adjustment is that the data changes smoothly before and after the adjustment, and there are no drastic fluctuations in the data after the adjustment. The degree of change before and after the adjustment focuses on the magnitude of the impact of the adjustment on the data, whether there is a sudden change from before irrigation to after irrigation, and avoids the adjustment data causing a large deviation from the terrain or irrigation status. The degree of change after the adjustment focuses on the continuous stability after the adjustment, whether the data after irrigation is stable, and avoids the problem of imbalance after the adjustment.

[0048] The coefficient 0.25 corresponds to the equal weighting of topography before and after, topography after adjustment, irrigation before and after, and irrigation after adjustment. This reflects the assessment logic that topography and irrigation are equally important, and the state before and after adjustment is equally critical. The formula for calculating the stability of the adjustment shows that the smaller the degree of change, the closer the stability is to 1.

[0049] It should be noted that all normalizations in the embodiments can be processed using the Min-Max normalization technique, which is an existing technology and will not be described in detail here.

[0050] The adjustment effect data includes the first adjustment effect and the second adjustment effect. The number of times the first adjustment effect and the number of times the second adjustment effect are good when each target field uses each adjustment data are counted, and the stability of the effect when each target field uses each adjustment data is analyzed.

[0051] It should be noted that each adjustment includes the first adjustment and the second adjustment.

[0052] Stability of effect: The number of times the first adjustment and the number of times the second adjustment are good are normalized and then recorded as b1 and b2 respectively. Stability of effect = 0.6×b1 + 0.4×b2.

[0053] The core of effectiveness stability is that the same adjustment data can consistently produce good results in multiple applications, rather than achieving the best result in a single instance. The number of times the first adjustment shows good results covers the effectiveness of the adjustment data in the initial generalized allocation scenario, meeting the core requirements of the first allocation in S2. The number of times the second adjustment shows good results covers the effectiveness of the adjustment data in the second targeted attack scenario, meeting the core requirements of the second allocation in S3. These two types of results together constitute the effectiveness evaluation of the entire allocation scenario, avoiding the one-sidedness caused by statistics from a single scenario. The higher weight of b1 is the core requirement of prioritizing the stability of regular scenarios.

[0054] The mean of the adjustment stability and effect stability of each target field using each adjustment data is calculated, and the calculation result is used as the predicted stability of each target field using each adjustment data. The adjustment data with the highest predicted stability is selected as the optimal adjustment data.

[0055] S2.4 Before regulation, the irrigation district control center conducts a status check. After the status is normal, the irrigation district control center controls according to the optimal adjustment data of each target field. During regulation, the topography and irrigation of each target field are monitored to obtain the topography and irrigation data of each target field at each regulation time. Then, the effect of the first regulation for each target field is analyzed. If at least one target field has a poor effect, S3 is executed.

[0056] Preferably, the process of S2.4 is as follows: S2.4-1, Before regulation, the irrigation district control center connects to the irrigation district intelligent management and control platform to check the equipment status of the canal gates and field branch gates of each target field. When the equipment status is normal, the canal gates and field branch gates are controlled according to the optimal adjustment data of each target field.

[0057] It should be noted that the specific verification process is as follows: The irrigation district intelligent management and control platform verifies the status of the gates and equipment through equipment self-inspection. The verification content differs for different types of gates and equipment. Taking the gate as an example: a voltage sensor built into the gate collects the power supply voltage, and a vibration sensor and limit switch are built into the gate to detect the mechanical status; a flow sensor from the distribution canal is used to assist in detecting the seal; when all the following conditions are met—power supply voltage within ±5% of the rated value, vibration amplitude ≤0.5mm / s (no jamming), normal limit switch signal (accurate feedback when fully open / fully closed), and flow rate close to 0 when there is no water flow (no leakage)—the status is considered normal. The rated values ​​can be obtained from the gate's instruction manual.

[0058] S2.4-2. During regulation, the sensor devices of each target field are used to monitor the terrain data and irrigation data at a preset monitoring frequency. The terrain data and irrigation data of each target field at each regulation time are obtained. Then, the adjustment stability and irrigation uniformity of each target field are analyzed, and the effect of each target field on the first regulation is judged.

[0059] The preset monitoring frequency can be set and adjusted by the supervisors according to the irrigation adjustment needs, and no specific numerical limit is specified here.

[0060] The above-mentioned judgment of the effect of the first adjustment is as follows: according to the calculation method of adjustment stability in S2.3, the adjustment stability of the first adjustment of each target field is calculated. The irrigation parameters of each monitoring point are obtained from the irrigation data of each target field at each control time. Then, the maximum and minimum irrigation parameters at each control time are selected. The irrigation difference at each control time is calculated as the difference between the maximum and minimum irrigation parameters at each control time divided by the preset allowable irrigation parameter difference. The irrigation uniformity at each control time is calculated as 1 - the irrigation difference at each control time. The average value of the irrigation uniformity at each control time is then calculated to obtain the irrigation uniformity.

[0061] Among them, the preset permissible irrigation parameter difference refers to the maximum irrigation difference in irrigation parameters that are allowed for crop growth in the irrigation area. It is set and adjusted by professionals according to crop characteristics, and no specific numerical limit is imposed here.

[0062] When both the adjustment stability and irrigation uniformity are greater than the preset adjustment stability threshold and irrigation uniformity threshold, the adjustment effect is judged to be good; otherwise, it is judged to be poor.

[0063] Field monitoring data is transformed into quantitative features. In the first adjustment effect judgment stage, AI quantifies "adjustment stability" and "irrigation uniformity" into a two-dimensional matrix and automatically marks "unresolved abnormal fields" through clustering algorithms. This AI-driven quantitative identification ensures that field needs can be transmitted to the irrigation district level in "precise values" rather than "fuzzy descriptions", providing calculable input basis for multi-scale resource scheduling and is a prerequisite for multi-scale linkage.

[0064] Among them, the stability threshold is the critical value for determining whether the terrain and irrigation are stable after adjustment, and the irrigation uniformity threshold is the critical value for determining whether the irrigation of the field is uniform after adjustment. The stability threshold and the irrigation uniformity threshold are set and adjusted by the supervisors according to the allocation needs of the irrigation area, and no specific numerical restrictions are imposed here.

[0065] This application constructs a closed-loop linkage mechanism of "micro-level field anomaly triggering → meso-level canal system resource scheduling → micro-level field effect feedback". This solution, through the full-process design of "field-scale anomaly identification - irrigation district-scale canal system scheduling - field-scale effect verification", fully embeds multi-scale allocation logic: the core of multi-scale allocation is "different management scales based on data linkage to realize the closed loop of 'problem upward reporting - resource scheduling - effect downward feedback'". The solution, through the design of "anomaly triggering point - scheduling execution layer - effect feedback layer", fully matches this essence.

[0066] S2.4-2 clearly states that irrigation anomalies are captured in real time through "topography / irrigation data of each monitoring point in the field", and unresolved abnormal fields are marked as "marked fields" through "first-level adjustment effect judgment" - this is the core action of transmitting "resource allocation needs" from the field scale to a higher level, which corresponds to the basic link of "making micro-problems explicit" in multi-scale allocation.

[0067] The "equipment status verification and optimal adjustment data generation and transmission of the main valve of the irrigation canal and the branch valve of the field" is not an isolated field operation, but a resource coordination based on the irrigation district scale: the irrigation district control center controls the water resource allocation of the entire irrigation canal by controlling the main valve of the irrigation canal, and the branch valve of the field corresponds to the flow regulation of a single field. The linkage control of the two is essentially a three-level canal system resource scheduling of "irrigation district-irrigation canal-field".

[0068] The key to multi-scale allocation in irrigation districts is to "avoid resource waste caused by independent scheduling of each field". The solution achieves this goal through two core designs: a unified data platform to ensure that the irrigation district level can obtain full-scale data in real time and provide a basis for resource allocation; and a linkage control logic: the opening adjustment of the main valve of the canal needs to be synchronized with the status of the branch valves of all subordinate fields, which is a typical "meso-micro" linkage, thus achieving multi-scale scheduling effect.

[0069] S3. Secondary Adjustment: Mark the target fields with poor performance as marked fields, extract abnormal terrain data and abnormal irrigation data of deformed areas in each marked field, and use the historical adjustment records of irrigation districts in the database to confirm the secondary adjustment data of each marked field and send it to the irrigation district control center. The irrigation district control center performs corresponding adjustments, monitors the terrain and irrigation during the adjustment, analyzes the effect of the secondary adjustment on each marked field, and then performs corresponding operations based on the effect of the secondary adjustment.

[0070] In a specific embodiment, the specific process of S3 is as follows: S3.1, extract the terrain data and irrigation data of each marked field at each control time, then locate the abnormal area and normal area from each marked field, and obtain the abnormal terrain data and abnormal irrigation data corresponding to the abnormal area, and compare them with the terrain data and irrigation data of the normal area to obtain the abnormal terrain data deviation value and irrigation data deviation value corresponding to the abnormal area.

[0071] The process involves locating abnormal and normal areas: marking fields are divided into grids, and the average topographic and irrigation parameters of each monitoring point in each grid are used as the topographic and irrigation data for each grid. Then, the analysis method of the first-level survey effect for each target field in S2.4 is used to analyze the first-level survey effect of each grid. Grids with poor first-level survey effects are clustered to obtain abnormal areas, and grids with good first-level survey effects are clustered to obtain normal areas.

[0072] The deviation values ​​of abnormal terrain data and irrigation data are the differences between the abnormal terrain data and abnormal irrigation data corresponding to the abnormal area and the terrain data and irrigation data of the normal area.

[0073] S3.2 Obtain the deviation values ​​of abnormal terrain data before adjustment, the deviation values ​​of irrigation data before adjustment, the deviation values ​​of abnormal terrain data after adjustment, the deviation values ​​of irrigation data after adjustment, the second adjustment data and the second adjustment effect from the historical adjustment records of the irrigation district for each abnormal area corresponding to the second adjustment, and filter out the second adjustment of each marked field abnormal area.

[0074] Among them, the secondary adjustments that correspond to the same abnormal terrain data deviation value and irrigation data deviation value as the abnormal terrain data deviation value and irrigation data deviation value of the abnormal area of ​​each marked field are regarded as the secondary adjustments of each mark.

[0075] S3.3. Based on the secondary adjustment of each marker in the abnormal area of ​​each marked field, calculate the adjustment characteristic value of each marker in the abnormal area using each secondary survey data, select the secondary survey data with the largest adjustment characteristic value as the secondary survey data of each marked field, and carry out corresponding secondary regulation.

[0076] The process described above for calculating the adjustment feature values ​​of each marked field anomaly area using each second survey data is as follows: extract the pre-adjustment abnormal terrain data deviation value, pre-adjustment irrigation data deviation value, second survey effect value, second survey data and second survey effect of each marked field anomaly area, and statistically analyze the pre-adjustment abnormal terrain data deviation value, pre-adjustment irrigation data deviation value, second survey effect value and second survey effect of each marked field anomaly area when using each second survey data.

[0077] Extract the second-order adjustment effect values ​​of each marked field with good second-order adjustment effect when using each second-order data for abnormal areas, and calculate the adjustment stability of each marked field with good second-order adjustment effect when using each second-order data for abnormal areas.

[0078] It should be noted that the calculation methods for the adjustment stability of each marked field's abnormal area when using each second survey data with good and poor results are the same as those in S2.4-2, and will not be repeated here.

[0079] Based on the adjustment stability, number of times the second national land survey results were good, and number of times the second national land survey results were poor when using the second national land survey data for the abnormal areas of each marked field, the adjustment characteristic value of each second national land survey data for the abnormal areas of each marked field was calculated.

[0080] Divide the number of times the second survey had a good effect by the sum of the number of times the second survey had a good effect and the number of times the second survey had a poor effect to obtain the proportion of the second survey with a good effect.

[0081] The stability of the second-level adjustment with good results and the proportion of the second-level adjustment with good results are normalized and denoted as d1 and d2 respectively. The adjustment characteristic value is 0.6×d1+0.4×d2.

[0082] S3.4 During the second-stage regulation, the topography and irrigation of the abnormal areas of each marked field are monitored, and the topography and irrigation data of the abnormal areas of each marked field at each monitoring time are obtained. The second-stage regulation effect value of each marked field is analyzed, the effect of the second-stage regulation corresponding to each marked field is determined, and then the corresponding operation is performed according to the effect of the second-stage regulation.

[0083] Preferably, the specific process of S3.4 is as follows: S3.4-1, using the sensor devices in each marked field to monitor the terrain data and irrigation data of the abnormal area and the normal area according to the preset monitoring frequency, to obtain the terrain data and irrigation data of the abnormal area and the normal area of ​​each marked field at each monitoring time, and then to obtain the terrain data deviation value and irrigation data deviation value of the abnormal area of ​​each marked field at each monitoring time.

[0084] The preset monitoring frequency can be set and adjusted by the supervisors according to the irrigation adjustment needs, and no specific numerical limit is specified here.

[0085] S3.4-2. Based on the topographic and irrigation data of the abnormal and normal areas of each marked field at each monitoring time, calculate the second-order impact value of the normal area of ​​each marked field. Based on the topographic and irrigation data deviation values ​​of the abnormal areas of each marked field at each monitoring time, calculate the second-order characteristic value of the abnormal areas of each marked field.

[0086] Specifically, following the calculation methods for adjustment stability and irrigation uniformity in S2.4-2, the adjustment stability and irrigation uniformity of the normal area of ​​each marked field are calculated. The adjustment stability and irrigation uniformity of the normal area of ​​each marked field before the second adjustment are also obtained (the topographic data and irrigation data of the normal area at each control time can be calculated according to the calculation methods for adjustment stability and irrigation uniformity in S2.4-2). The difference between the adjustment stability of the normal area of ​​each marked field and the adjustment stability before the second adjustment is divided by the adjustment stability before the second adjustment to obtain the adjustment stability fluctuation rate. The irrigation uniformity fluctuation rate is obtained according to the calculation method for adjustment stability fluctuation rate. The impact value of the second adjustment is the result of summing the adjustment stability fluctuation rate and the irrigation uniformity fluctuation rate and then averaging them.

[0087] The adjustment stability and irrigation uniformity of each marked field abnormal area were calculated according to the calculation method of adjustment stability and irrigation uniformity in S2.4-2. After normalization, the average value was calculated to obtain the second adjustment characteristic value.

[0088] S3.4-3. Using the second-round survey impact value of the normal area and the second-round survey characteristic value of the abnormal area of ​​each marked field, calculate the second-round survey effect value of each marked field, determine the effect of the second-round survey for each marked field, send the marked fields with poor second-round survey effect to the irrigation district control center, prompting manual inspection and adjustment of field topography and irrigation, and send the marked fields with good second-round survey effect to the irrigation district control center, prompting key monitoring.

[0089] Wherein, the second-order effect value = 0.5 × the second-order characteristic value of the abnormal area + 0.5 × (1 - the second-order influence value of the normal area).

[0090] In the above, the essence of the second-level survey impact value is "the quantified value of the disturbance received by the normal area due to the second-level survey," and its value range is [0,1]—the closer the value is to 1, the more severe the disturbance in the normal area; the closer it is to 0, the smaller the disturbance. The second-level survey characteristic value reflects the quantified value of the abnormal improvement effect, and its value range is [0,1]—the closer the value is to 1, the better the improvement effect in the abnormal area.

[0091] By converting "1-influence value", the converted value is made to be completely consistent with the logical direction of the second-order characteristic value of the abnormal area. That is, the closer the value is to 1, the better the protection effect of the normal area. This forms a unified "optimal value orientation" with "the closer the second-order characteristic value is to 1, the better the improvement effect of the abnormal area", which provides a basis for weighted fusion. Weighted fusion of the second-order characteristic value of the abnormal area and the second-order influence value of the normal area can prevent the sacrifice of the normal area in pursuit of improvement of the abnormal area and avoid solving one problem and creating another problem. The balanced evaluation logic of abnormal improvement + normal protection avoids the one-sidedness of the second-order survey.

[0092] When the second-order effect value is greater than the preset second-order effect value threshold, the second-order effect is judged to be good; otherwise, it is judged to be poor.

[0093] Among them, the threshold value of the second survey effect value is the critical value for judging whether the second survey effect is good. It is set and adjusted by the supervisors according to the allocation needs of the irrigation area, and no specific numerical limit is imposed here.

[0094] By adjusting the mean values ​​of stability fluctuation rate and irrigation uniformity fluctuation rate, the relative change rate is used to quantify the disturbance of the second national land survey to normal areas, which can objectively reflect the degree of impact of the second national land survey. The second national land survey should not only solve abnormal problems, but also not disrupt the irrigation status of normal areas. Both are equally important. The effect value of the second national land survey assesses both the improvement effect in abnormal areas and the degree of impact on normal areas, realizing a two-dimensional assessment and avoiding one-sidedness.

[0095] The effect evaluation of S3.4 is a key link in the AI ​​quantitative feedback: In the calculation of the feature value of the second survey in the abnormal area, the AI ​​automatically identifies the quantitative contribution of "adjustment stability" (weight 0.6) and "irrigation uniformity" (weight 0.4) through normalization and weight allocation; In the feedback link, the AI ​​quantifies the field effect and associates it with the irrigation district scheduling parameters to generate and trigger the next round of precise scheduling. This AI logic of "quantitative feedback-parameter optimization" ensures that the multi-scale closed loop is not a simple repetition, but a continuous iterative process of precision.

[0096] See Figure 2 As shown, an AI-based multi-scale dynamic allocation system for irrigation district water resources includes: an irrigation district monitoring module, used to monitor the topography and irrigation of each field in the irrigation district, determine whether each field needs irrigation adjustment, and execute the allocation module once if at least one field needs irrigation adjustment.

[0097] The first allocation module is used to mark each field that needs irrigation adjustment as a target field, and at the same time obtain the size and terrain data of each target field. It also uses the historical adjustment records of irrigation districts in the database to analyze the optimal adjustment data for each target field, and adjusts the field according to the optimal adjustment data. During the adjustment, the terrain and irrigation are monitored, and the effect of the first adjustment for each target field is analyzed. If at least one target field is not effective, the second allocation module is executed.

[0098] The secondary adjustment module is used to mark the target fields with poor performance as marked fields, extract abnormal terrain data and abnormal irrigation data of deformed areas in each marked field, and use the historical adjustment records of the irrigation district in the database to confirm the secondary adjustment data of each marked field and send it to the irrigation district control center. The irrigation district control center performs corresponding adjustments, monitors the terrain and irrigation during the adjustment, analyzes the effect of the secondary adjustment on each marked field, and then performs corresponding operations based on the effect of the secondary adjustment.

[0099] The database is used to store historical adjustment records of irrigation districts, topographic difference thresholds, and irrigation difference thresholds.

[0100] The examples described in this invention are not limited to the specific embodiments listed above. The examples are merely illustrative to facilitate understanding of the invention and do not constitute a limitation on the scope of protection of this invention. Any modifications, equivalent substitutions, etc., made within the spirit and principles of this invention should be included within the scope of protection.

[0101] The above description is merely an example and illustration of the concept of the present invention. Those skilled in the art can make various modifications or additions to the specific embodiments described or use similar methods to replace them, as long as they do not deviate from the concept of the invention or exceed the scope defined in this specification, they should all fall within the protection scope of the present invention.

Claims

1. A multi-scale dynamic allocation method for irrigation district water resources based on AI, characterized in that, Includes the following steps: S1. Irrigation District Monitoring: Monitor the topography and irrigation of each field in the irrigation district to determine whether each field needs irrigation adjustment. If at least one field needs irrigation adjustment, execute S2. S2, First Adjustment: Each field requiring irrigation adjustment is designated as a target field. The size and topography data of each target field are obtained. The optimal adjustment data for each target field is analyzed using historical adjustment records of the irrigation district in the database. The optimal adjustment data is then sent to the irrigation district control center. The irrigation district control center adjusts the irrigation according to the optimal adjustment data. During the adjustment, the topography and irrigation are monitored, and the effect of the first adjustment for each target field is analyzed. If at least one target field has a poor effect, S3 is executed. S3. Secondary Adjustment: Mark the target fields with poor performance as marked fields, extract abnormal terrain data and abnormal irrigation data of deformed areas in each marked field, and use the historical adjustment records of irrigation districts in the database to confirm the secondary adjustment data of each marked field and send it to the irrigation district control center. The irrigation district control center performs corresponding adjustments, monitors the terrain and irrigation during the adjustment, analyzes the effect of the secondary adjustment on each marked field, and then performs corresponding operations based on the effect of the secondary adjustment.

2. The AI-based multi-scale dynamic allocation method for irrigation district water resources according to claim 1, characterized in that, The specific process of S1 is as follows: S1.

1. Sensor devices are installed in each field in the irrigation area. The sensor devices are used to detect the topographic data and irrigation data of each field at preset time intervals to obtain the topographic data and irrigation data of each field at each time. S1.

2. Based on the topographic and irrigation data of each field at each time, analyze the change status of each field. If the change status is abnormal, irrigation adjustment is required. In this way, it is determined whether irrigation adjustment is required for each field.

3. The AI-based multi-scale dynamic allocation method for irrigation district water resources according to claim 2, characterized in that, The specific process for analyzing the changing status of each field is as follows: Plot a time-topography line graph with time as the horizontal axis and topographic data as the vertical axis. Then extract the slope corresponding to each point on the line graph. Finally, select the maximum value of the difference between the slopes of each point as the topographic difference degree to obtain the topographic difference degree corresponding to each field. The irrigation data of each field at each time point are used to calculate the irrigation difference of each field according to the method of calculating the topographic difference of each field. The topographic and irrigation differences of each field are compared with the preset topographic and irrigation difference thresholds in the database. If the topographic or irrigation difference is greater than the corresponding threshold, it indicates an abnormal change. If the topographic and irrigation differences are less than or equal to the corresponding preset thresholds, it indicates a normal change. This is how the change status of each field is obtained.

4. The AI-based multi-scale dynamic allocation method for irrigation district water resources according to claim 1, characterized in that, The specific process of S2 is as follows: S2.1 Classify the anomalies of each target field into irrigation anomaly type, terrain anomaly type, and composite anomaly type; extract the terrain data and irrigation data of each target field at the last moment as the current terrain data and current irrigation data; S2.2 Obtain the anomaly type, pre-adjustment topographic data, pre-adjustment irrigation data, post-adjustment topographic data, post-adjustment irrigation data, adjustment data, and adjustment effect data of the corresponding fields for each adjustment from the historical adjustment records of the irrigation district. Select the adjustments for which the anomaly type, pre-adjustment topographic data, and pre-adjustment irrigation data of the fields correspond to the same anomaly type, current topographic data, and current irrigation data of each target field as the target adjustments for each adjustment. S2.3 Extract the pre-adjustment topographic data, pre-adjustment irrigation data, post-adjustment topographic data, post-adjustment irrigation data, adjustment data, and adjustment effect data of each target field for each target adjustment, select the best adjustment data for each target field, and send it to the irrigation district control center. S2.4 Before regulation, the irrigation district control center conducts a status check. After the status is normal, the irrigation district control center controls according to the optimal adjustment data of each target field. During regulation, the topography and irrigation of each target field are monitored to obtain the topography and irrigation data of each target field at each regulation time. Then, the effect of the first regulation for each target field is analyzed. If at least one target field has a poor effect, S3 is executed.

5. The AI-based multi-scale dynamic allocation method for irrigation water resources according to claim 4, characterized in that, The specific process for selecting the optimal adjustment data for each target field is as follows: Based on the pre-adjustment topographic data, pre-adjustment irrigation data, post-adjustment topographic data, post-adjustment irrigation data, adjustment data, and adjustment effect data of each target field for each target adjustment, the pre-adjustment topographic data, pre-adjustment irrigation data, post-adjustment topographic data, post-adjustment irrigation data, and adjustment effect data of each target field when using each adjustment data are statistically analyzed. Using the pre-adjustment topographic data, pre-adjustment irrigation data, post-adjustment topographic data, and post-adjustment irrigation data of each target field when using each adjustment data, calculate the adjustment stability of each target field when using each adjustment data. The adjustment effect data includes the first adjustment effect and the second adjustment effect. The number of times the first adjustment effect and the second adjustment effect were good when each target field used each adjustment data are counted, and the stability of the effect when each target field used each adjustment data is analyzed. The mean of the adjustment stability and effect stability of each target field using each adjustment data is calculated, and the calculation result is used as the predicted stability of each target field using each adjustment data. The adjustment data with the highest predicted stability is selected as the optimal adjustment data.

6. The AI-based multi-scale dynamic allocation method for irrigation district water resources according to claim 4, characterized in that, The process in S2.4 is as follows: S2.4-1 Before regulation, the irrigation district control center connects to the irrigation district intelligent management and control platform to check the equipment status of the branch gates of the irrigation canals and the branch gates of the fields where each target field is located. When the equipment status is normal, the branch gates of the irrigation canals and the branch gates of the fields are controlled according to the optimal adjustment data of each target field. S2.4-2. During regulation, the sensor devices of each target field are used to monitor the terrain data and irrigation data at a preset monitoring frequency. The terrain data and irrigation data of each target field at each regulation time are obtained. Then, the adjustment stability and irrigation uniformity of each target field are analyzed, and the effect of each target field on the first regulation is judged.

7. The AI-based multi-scale dynamic allocation method for irrigation district water resources according to claim 1, characterized in that, The specific process of S3 is as follows: S3.1 Extract the topographic and irrigation data of each marked field at each control time, then locate the abnormal and normal areas from each marked field, obtain the abnormal topographic and irrigation data corresponding to the abnormal areas, and compare them with the topographic and irrigation data of the normal areas to obtain the abnormal topographic and irrigation data deviation values ​​corresponding to the abnormal areas. S3.2 Obtain the deviation values ​​of abnormal terrain data before adjustment, the deviation values ​​of irrigation data before adjustment, the deviation values ​​of abnormal terrain data after adjustment, the deviation values ​​of irrigation data after adjustment, the second adjustment data and the second adjustment effect from the historical adjustment records of the irrigation district for each abnormal area corresponding to the second adjustment, and filter out the second adjustment of each marked field abnormal area. S3.

3. Based on the secondary adjustment of each marker in the abnormal area of ​​each marked field, calculate the adjustment characteristic value of each marker in the abnormal area using each secondary survey data, select the secondary survey data with the largest adjustment characteristic value as the secondary survey data of each marked field, and perform corresponding secondary regulation. S3.4 During the second-stage regulation, the topography and irrigation of the abnormal areas of each marked field are monitored, and the topography and irrigation data of the abnormal areas of each marked field at each monitoring time are obtained. The second-stage regulation effect value of each marked field is analyzed, the effect of the second-stage regulation corresponding to each marked field is determined, and then the corresponding operation is performed according to the effect of the second-stage regulation.

8. The AI-based multi-scale dynamic allocation method for irrigation district water resources according to claim 7, characterized in that, The process of calculating the adjusted feature values ​​of each marked field's abnormal area using each second land survey data is as follows: Extract the deviation values ​​of the abnormal terrain data before the adjustment, the deviation values ​​of the irrigation data before the adjustment, the effect value of the second adjustment, the data of the second adjustment, and the effect of the second adjustment for each marked abnormal area. Statistically calculate the deviation values ​​of the abnormal terrain data before the adjustment, the deviation values ​​of the irrigation data before the adjustment, the effect value of the second adjustment, and the effect of the second adjustment for each marked abnormal area when using each second adjustment data. Extract the second-order adjustment effect value of each marked field with good second-order adjustment effect when using each second-order data for the abnormal areas of each marked field, and calculate the adjustment stability of each marked field with good second-order adjustment effect when using each second-order data for the abnormal areas of each marked field. Based on the adjustment stability, number of times the second national land survey results were good, and number of times the second national land survey results were poor when using the second national land survey data for the abnormal areas of each marked field, the adjustment characteristic value of each second national land survey data for the abnormal areas of each marked field was calculated.

9. A multi-scale dynamic allocation method for irrigation water resources based on AI according to claim 7, characterized in that, The specific process of S3.4 is as follows: S3.4-1. Using the sensor devices in each marked field, the terrain data and irrigation data of the abnormal area and the normal area are monitored according to the preset monitoring frequency. The terrain data and irrigation data of the abnormal area and the normal area of ​​each marked field are obtained at each monitoring time. Then, the deviation values ​​of the terrain data and the deviation values ​​of the irrigation data of the abnormal area of ​​each marked field are obtained at each monitoring time. S3.4-2. Based on the topographic and irrigation data of the abnormal and normal areas of each marked field at each monitoring time, calculate the second-level survey impact value of the normal area of ​​each marked field. Based on the topographic and irrigation data deviation values ​​of the abnormal areas of each marked field at each monitoring time, calculate the second-level survey characteristic value of the abnormal areas of each marked field. S3.4-3. Using the second-round survey impact value of the normal area and the second-round survey characteristic value of the abnormal area of ​​each marked field, calculate the second-round survey effect value of each marked field, determine the effect of the second-round survey for each marked field, send the marked fields with poor second-round survey effect to the irrigation district control center, prompting manual inspection and adjustment of field topography and irrigation, and send the marked fields with good second-round survey effect to the irrigation district control center, prompting key monitoring.

10. A system for multi-scale dynamic allocation of irrigation district water resources based on the AI-based method for multi-scale dynamic allocation of irrigation district water resources according to any one of claims 1-9, characterized in that, include: The irrigation district monitoring module is used to monitor the topography and irrigation of each field in the irrigation district, and determine whether each field needs irrigation adjustment. If at least one field needs irrigation adjustment, the allocation module is executed once. The first allocation module is used to mark each field that needs irrigation adjustment as a target field, and at the same time acquire the size and topography data of each target field. It also uses the historical adjustment records of the irrigation district in the database to analyze the optimal adjustment data for each target field and sends the optimal adjustment data to the irrigation district control center. The irrigation district control center adjusts according to the optimal adjustment data. During the adjustment, the topography and irrigation are monitored and the effect of the first adjustment for each target field is analyzed. If at least one target field is not effective, the second allocation module is executed. The secondary adjustment module is used to mark the target fields with poor performance as marked fields, extract abnormal terrain data and abnormal irrigation data of deformed areas in each marked field, and use the historical adjustment records of the irrigation district in the database to confirm the secondary adjustment data of each marked field and send it to the irrigation district control center. The irrigation district control center performs corresponding adjustments, monitors the terrain and irrigation during the adjustment, analyzes the effect of the secondary adjustment on each marked field, and then performs corresponding operations based on the effect of the secondary adjustment.

Images