Underground river basin water resource regulation and storage method based on partition collaboration

Abstract

The invention relates to an underground river basin water resource regulation and storage method based on partition collaboration, which comprises the following steps: arranging an intercepting ditch network and an earth surface water storage facility group in a target underground river basin for irrigating vegetation in the target underground river basin; in the process of irrigating the vegetation in the target underground river basin, collecting soil state information and meteorological characteristic data of soil in the target underground river basin, and inputting the soil state information and the meteorological characteristic data into a vegetation water demand model to obtain a vegetation water demand; the vegetation water demand model is obtained by training a machine learning model through a training set; the training set comprises original soil state information and original meteorological characteristic data. Scientific management and sustainable utilization of water resources in the underground river basin can be realized.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of ecological restoration and water resource management, and particularly relates to a water resource regulation and storage method for underground river basins based on zoning coordination. BACKGROUND

[0002] The southwest region of China is the largest karst area in the world with a contiguous distribution area of about 532,600 square kilometers. According to surveys and statistics, there are 3,066 karst underground rivers developed in this region. The pipe system as the main water-bearing medium constitutes an important storage space for groundwater and becomes an important water source for local residents' life and agricultural production. However, affected by the special hydrogeological structure and human activities, these underground river basins are facing a series of severe challenges: uneven seasonal distribution of water resources, soil erosion leading to underground river siltation, and continuous aggravation of stone desertification, which seriously restricts the water resource regulation and storage function and efficient utilization of underground rivers.

[0003] In the process of water resource management and utilization of underground river basins, there are many challenges. On the one hand, the distribution of water resources in underground river basins has the characteristics of temporal and spatial unevenness. Some areas are rich in water resources, while others are relatively scarce, leading to low utilization efficiency of water resources and failing to fully meet the water demand of vegetation irrigation and other uses in the basin. On the other hand, traditional water resource regulation and storage methods often focus on single engineering measures such as building reservoirs and dams, lack of systematic consideration of the whole underground river basin, and are difficult to achieve coordinated allocation of water resources between different regions, which cannot fully realize the comprehensive benefits of water resources in underground river basins.

[0004] At the same time, in terms of vegetation irrigation, the traditional irrigation method is mostly based on experience and cannot accurately irrigate according to the actual water demand of vegetation, which easily causes waste of water resources or insufficient irrigation. Moreover, for the complex soil and meteorological conditions in underground river basins, there is a lack of effective monitoring and analysis means to accurately assess the water demand of vegetation, which further affects the irrigation effect and rational utilization of water resources.

[0005] In addition, the water resource regulation and storage of underground river basins also involves ecological protection. When developing and utilizing water resources, the protection of underground river ecosystems should be fully considered to ensure that the ecological storage capacity of underground rivers is maintained at a reasonable level to maintain the balance and stability of the ecological system of underground river basins. However, at present, there is a lack of effective technical means to accurately estimate the ecological storage capacity of underground rivers and dynamically monitor and control the ecological storage capacity during water resource regulation and storage. SUMMARY

[0006] In order to solve the problems existing in the prior art, the purpose of the present application is to provide a water resource regulation and storage method for underground river basins based on zoning coordination, so as to realize the scientific management and sustainable utilization of water resources in underground river basins.

[0007] To achieve the above object, the present application provides the following scheme:

[0008] A partition coordination-based underground river basin water resource regulation and storage method, comprising:

[0009] Setting a water interception ditch network and a surface water storage facility group in a target underground river basin for irrigation of vegetation in the target underground river basin;

[0010] In the process of irrigating the vegetation in the target underground river basin, collecting soil state information and meteorological characteristic data of the soil in the target underground river basin, inputting the soil state information and the meteorological characteristic data into a vegetation water requirement model to obtain a vegetation water requirement; the vegetation water requirement model is obtained by training a machine learning model using a training set; the training set includes original soil state information and original meteorological characteristic data.

[0011] Optionally, setting the water interception ditch network comprises:

[0012] In the target underground river basin, the distribution area of the tourist and water consumption cave is dredged, and a hierarchical structure is adopted, a water diversion channel and a three-stage sand settling tank are sequentially arranged in the distribution area of the tourist and water consumption cave, a sand blocking net is arranged at the connection of the three-stage sand settling tank, and a circular water consumption pool is arranged at the entrance of the water consumption cave, and the processed rainwater is collected into the underground river from the water consumption pool after passing through the three-stage sand settling tank.

[0013] Optionally, setting the surface water storage facility group comprises:

[0014] In the underground river skylight area in the target underground river basin, two-stage regulation and storage dams are set; wherein the two-stage regulation and storage dams adopt a double-layer structure of bottom concrete and upper hydraulic steel dam;

[0015] In the underground river skylight area between the two-stage regulation and storage dams, a water energy dam is arranged; the water energy dam comprises a water diversion channel, a blocking and storage dam, and a water energy pump.

[0016] The water diversion channel is used to guide the flow direction and flow rate of underground water;

[0017] The blocking and storage dam is used to set a water level device to collect water level data, and use the water level data to determine the underground river ecological storage capacity, so as to allocate underground river water resources;

[0018] The water energy pump is used to use the potential energy generated by the water level difference to lift the underground river water to a high place for irrigation of nearby farmland.

[0019] Optionally, determining the underground river ecological storage capacity comprises:

[0020] The For loop tool and the surface volume tool in the ArcGIS software are used to construct a model for estimating the water level-storage capacity relationship of the dammed reservoir through a model manager, and a water level-storage capacity relationship curve of the dammed reservoir is drawn;

[0021] According to the water level-storage capacity relationship curve of the dammed reservoir, the water surface area under different water levels is determined, and a water surface area-storage capacity relationship curve is constructed;

[0022] The water level of the dammed reservoir in different time periods is extracted, the water surface area under the water level of the dammed reservoir is obtained according to the water level of the dammed reservoir, the water surface area is input into the water surface area-storage capacity relationship curve, and the ecological water storage capacity of the underground river in different time periods is obtained.

[0023] Optionally, the soil state information includes soil humidity, soil texture, soil temperature and soil structure.

[0024] The meteorological feature data includes daily cumulative average illumination, daily average illumination, daily maximum illumination, daily average temperature, diurnal average temperature, daily maximum temperature, daily minimum temperature, daily average humidity, diurnal average humidity, daily maximum humidity and daily minimum humidity.

[0025] Optionally, training the machine learning model using the training set comprises:

[0026] Obtaining the sampling time period corresponding to each sample data in the training set, setting the sampling rate corresponding to the sample data, and obtaining the sampled sample data;

[0027] Determining the importance value of the sample data, correcting the sampled sample data using the importance value, and obtaining the corrected sample data;

[0028] Inputting the corrected sample data into the machine learning model and combining a loss function to train the model.

[0029] Optionally, determining the importance value of each sample data comprises:

[0030] Correcting the initial importance value of the sampled sample data based on the corresponding sampling rate to obtain the importance value of the sampled sample data;

[0031] The importance value and the initial importance value are in a direct proportion relationship, and the importance value and the sampling rate of the sampled sample data are in an inverse proportion relationship.

[0032] Optionally, obtaining the corrected sample data comprises multiplying the importance value and the corresponding sampled sample data to obtain the corrected sample data.

[0033] Optionally, the irrigation of the vegetation in the target underground river basin comprises:

[0034] When the vegetation water requirement is less than the lower limit value of the preset water requirement, the surface water storage facility group is controlled to carry out irrigation;

[0035] The vegetation water requirement is updated according to the vegetation water requirement after irrigation, and when the updated vegetation water requirement is greater than the upper limit value of the preset water requirement, the surface water storage facility group is controlled to stop irrigation.

[0036] The beneficial effects of the present application are:

[0037] The present application realizes precise control of vegetation irrigation by collecting soil state information and meteorological characteristic data and inputting them into a vegetation water requirement model to obtain vegetation water requirement. Only when the vegetation water requirement is less than the lower limit value of the preset water requirement, the surface water storage facility group is controlled to carry out irrigation, and the irrigation operation is updated and adjusted in time according to the vegetation water requirement after irrigation, avoiding the problem of water resource waste or insufficient irrigation caused by traditional irrigation methods based on experience, so that water resources can be more efficiently and reasonably utilized, the water demand of vegetation in the basin is fully met, and the utilization efficiency of water resources in vegetation irrigation is improved.

[0038] The present application takes into account the characteristics of spatial and temporal uneven distribution of water resources in underground river basins, realizes the collaborative allocation of water resources between different regions by setting up water interception ditch networks and surface water storage facility groups, and accurately estimating and dynamically monitoring and regulating the ecological water storage capacity of underground rivers. For example, in the middle reaches of the water consumption cave distribution area, the water diversion channel, sand pool and water consumption pool are set up, the treated rainwater is collected into the underground river, and the water resources of the underground river are increased; water energy dam is set up in the underground river skylight area, and water energy pump is used to lift the underground river water to a high place for irrigation of nearby farmland, effectively solving the problem of water resource shortage in some areas and unable to meet the water demand, fully utilizing the comprehensive benefits of water resources in underground river basins, and improving the utilization efficiency of water resources in the whole underground river basin.

[0039] In the present application, ArcGIS software is used to construct a model to estimate the water level-storage capacity relationship of the dam reservoir, and the ecological water storage capacity of the underground river in different time periods is obtained through the water surface area-storage capacity relationship curve, which can accurately determine the ecological water storage capacity of the underground river. In the process of water resources regulation and storage, the ecological water storage capacity can be dynamically monitored and regulated according to the accurate ecological water storage capacity, so as to maintain the ecological water storage capacity of the underground river at a reasonable level, thereby providing stable water source for the ecological system of the underground river basin and maintaining the balance and stability of the ecological system of the underground river basin, effectively avoiding the destruction of the ecological system caused by excessive development and utilization of water resources. BRIEF DESCRIPTION OF DRAWINGS

[0040] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0041] Figure 1 This is a method for regulating water resources in underground river basins based on zone collaboration according to an embodiment of the present invention;

[0042] Figure 2 A schematic diagram of a drainage ditch network according to an embodiment of the present invention;

[0043] Figure 3 A schematic diagram of a sand retaining net according to an embodiment of the present invention;

[0044] Figure 4 A schematic diagram of a surface water storage facility group according to an embodiment of the present invention;

[0045] Figure 5 Schematic diagram of a water level monitoring device according to an embodiment of the present invention;

[0046] Figure 6 Schematic diagram of a water pump according to an embodiment of the present invention. DETAILED DESCRIPTION

[0047] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0048] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0049] like Figure 1 As shown, this embodiment discloses a method for regulating water resources in an underground river basin based on zone collaboration, including: setting up a diversion ditch network and a surface water storage facility group in the target underground river basin to irrigate the vegetation in the target underground river basin; in the process of irrigating the vegetation in the target underground river basin, collecting soil state information and meteorological characteristic data of the soil in the target underground river basin, inputting the soil state information and meteorological characteristic data into a vegetation water demand model to obtain vegetation water demand; the vegetation water demand model is obtained by training a machine learning model using a training set; the training set includes: original soil state information and original meteorological characteristic data.

[0050] Specifically, the embodiment discloses a water resource regulation and storage method for underground river basins based on zoning cooperation, which comprises: implementing a comprehensive project of "intercepting, storing, supplementing and irrigating" in a concentrated distribution area of stone desertification in an upstream area of a basin; the comprehensive project comprises a network of intercepting ditches, a group of surface water storage facilities and an intelligent irrigation system. A rainfall observation station is built in a low-lying area, three-stage intercepting ditches are built along the top of a slope and the middle of a slope surface, slope surface runoff is intercepted and led to a water storage pool, the water storage pool is provided with an intelligent irrigation control system, the intelligent irrigation control system comprises drip irrigation facilities, a soil humidity sensor and an intelligent processing platform, soil humidity is monitored in real time, vegetation water demand is predicted through a machine learning algorithm, the intelligent processing platform controls the drip irrigation system to realize precise irrigation, a virtuous cycle system of "storing water during rainfall and irrigating during drought" is formed, the water source conservation capacity of the upstream area can be enhanced, the ecological water demand of vegetation and crops in the dry season can be ensured, the vegetation coverage rate can be effectively increased, and ecological restoration can be realized.

[0051] Further, in the middle reaches of the target underground river basin, the silted area of the water sink is dredged, and a hierarchical structure is adopted, a water diversion channel and three-stage sand settling pools are arranged in the water sink distribution area in sequence, a sand blocking net is arranged at the connection of the three-stage sand settling pools, and a circular water sink is arranged at the water sink opening, and the treated rainwater is collected into the underground river from the water sink after passing through the three-stage sand settling pools.

[0052] Specifically, in the middle reaches of the underground river basin, in order to reduce the risk that the underground river is blocked by the mud brought into the underground river by surface water scouring in the rainy season, the blocked water sink in the middle reaches of the underground river basin is excavated and dredged, the soil layer in the water sink is completely removed, and the bedrock at the water sink opening is exposed. In order to prevent the water sink from being blocked again, three-stage sand settling pools are built near the water sink, the sand settling pools adopt a hierarchical structure, a water diversion channel is built first, and then three-stage sand settling pools are built, the sand settling pools have a square structure, the length and width are 5 m, and the depth is 4 m, and finally a circular water sink is built at the water sink opening, the water sink has a diameter of 8 m and a height of 4 m, the surface water is collected into the underground river from the water sink after passing through the three-stage sand settling pools, and a stainless steel sand blocking net is arranged at the connection of the sand settling pools, which can effectively block larger objects such as branches, the construction of the water sink sand settling pool can not only reduce the risk of waterlogging in the middle reaches, but also effectively reduce the mud entering the underground river, so that the water resource regulation and storage function of the underground river is improved, as shown in Figures 2-3 .

[0053] The floating soil layer at the water sink opening in the middle reaches of the underground river is cleaned, then three-stage sand settling pools and a water sink are built, most of the mud in the surface stream is deposited in the sand settling pool after the surface stream passes through the three-stage sand settling pool, then the surface stream is collected into the underground river from the water sink through a water diversion channel, a stainless steel sand blocking net is arranged at the connection of the sand settling pools, the mud carried by the rainwater into the underground river is effectively reduced, the risk of blockage of the underground river pipeline is reduced, and the water resource regulation and storage capacity of the underground river is enhanced.

[0054] Further, the ground surface water storage facility group comprises: two-stage regulation and storage dams are arranged in the underground river sky window area in the target underground river basin; wherein, the two-stage regulation and storage dams adopt a double-layer structure of a bottom concrete and an upper hydraulic steel dam; a water energy dam is arranged in the underground river sky window area between the two-stage regulation and storage dams; the water energy dam comprises: a water diversion channel, a retaining dam and a water energy pump; the water diversion channel is used to guide the underground water flow; the retaining dam is used to set a water level device to collect water level data, and the water level data are used to determine the underground river ecological storage capacity, so as to allocate the underground river water resources; and the water energy pump is used to use the potential energy generated by the river water level difference to lift the underground river water to a high place for irrigation of nearby farmland.

[0055] Specifically, due to the uneven seasonal distribution of underground river water resources, two-stage regulation and storage dams are built in the area where the underground river sky window develops in the underground river basin, the regulation and storage dams adopt a double-layer structure of a bottom concrete and an upper hydraulic steel dam, the hydraulic steel dam also adopts an intelligent platform control, the downstream regulation and storage dam is pre-buried with a water drainage pipeline at the bottom of the concrete, the waterproof pipeline device is controlled through the intelligent platform, an underground water level monitoring device is built upstream of the regulation and storage dam, and the indoor processing platform can solve the problem of reduced or interrupted outlet flow of the underground river in the dry season due to uneven rainfall according to the collected water level data, effectively increasing the regulation and storage capacity of the underground river water resources, such as shown in Figures 4-5 .

[0056] A water energy dam is built in the position where the underground river sky window develops between the two-stage regulation and storage dams, the water energy dam comprises a water diversion channel, a retaining dam and a water energy pump, the potential energy generated by the underground river water level difference is used to drive the water energy pump to work, the underground river water is lifted to a high place for irrigation of nearby farmland, so as to realize efficient utilization of the underground river water resources, such as shown in Figure 6 . The underground river basin is taken as a research unit, work is carried out at different positions in the basin, and problems such as rocky desertification caused by degradation of water source conservation function and low utilization rate of water resources caused by poor regulation and storage function of underground river water resources in the basin are solved scientifically and effectively, so as to realize regulation and storage, efficient utilization and ecological restoration of the underground river water resources.

[0057] Further, the determination of the underground river ecological storage capacity comprises: a model for estimating the relationship between the water level and the storage capacity of the dammed reservoir is constructed by using the For loop tool and the surface volume tool in the ArcGIS software through the model manager, and a relationship curve between the water level and the storage capacity of the dammed reservoir is drawn; the water surface area under different water levels is determined according to the relationship curve between the water level and the storage capacity of the dammed reservoir, and a relationship curve between the water surface area and the storage capacity is constructed; the water level of the dammed reservoir under different time periods is extracted, the water surface area under the water level of the dammed reservoir is obtained according to the water level of the dammed reservoir, the water surface area is input into the relationship curve between the water surface area and the storage capacity, and the underground river ecological storage capacity under different time periods is obtained.

[0058] Specifically, the water level-storage capacity data output by the model is exported to professional drawing software (such as Excel, Origin, etc.) using the drawing function in ArcGIS software. The water level-storage capacity relationship curve is drawn according to the data points calculated by the model, with the water level value as the horizontal coordinate and the storage capacity as the vertical coordinate. This curve intuitively reflects the change of the storage capacity of the dam reservoir at different water levels, and the shape of the curve can reflect the morphological characteristics of the reservoir and the law of the storage capacity changing with the water level. The water surface area-storage capacity relationship curve is constructed: according to the water level-storage capacity relationship curve of the dam reservoir at different water levels, the water surface area at different water levels is determined. The specific method is as follows: assuming that a point a on the water level-storage capacity relationship curve corresponds to the water level at the time, a tangent line of the water level-storage capacity relationship curve is drawn through point a, intersecting the vertical axis at point b, and the angle between the tangent line and the vertical axis is θ. Through geometric relationship or mathematical calculation method, the corresponding water surface area is calculated by using the slope of the tangent line and other information.

[0059] The water surface area data at different water levels is collected, and the water surface area-storage capacity relationship curve is drawn with the water surface area as the horizontal coordinate and the storage capacity as the vertical coordinate. This curve further reveals the relationship between the water surface area and the storage capacity, providing a basis for obtaining the storage capacity according to the water surface area in the subsequent process.

[0060] The actual water level data of the dam reservoir at different time periods is obtained from the monitoring data or related data sources. These water level data are the key information reflecting the change of the water level of the reservoir in the actual operation process, and are the basis for extracting the underground river ecological storage capacity at the corresponding time period. According to the dam reservoir water level at different time periods extracted, the corresponding water surface area is found in the dam reservoir water level-storage capacity relationship curve or the water surface area-storage capacity relationship curve. This step is to convert the actual water level data into water surface area data by using the relationship curve constructed in the previous step, so as to prepare for the next step of calculating the storage capacity.

[0061] The obtained water surface area data is input into the water surface area-storage capacity relationship curve. Through the function relationship or interpolation method of the relationship curve, the storage capacity under the corresponding water surface area is calculated. This storage capacity is the underground river ecological storage capacity at different time periods, which can reflect the ecological water storage of the underground river at different times, and has important reference value for the protection of the underground river ecosystem and the rational utilization of water resources.

[0062] The whole process of extracting the storage capacity is a comprehensive process of using geographic information system (GIS) technology and mathematical analysis method. Through the steps of model construction, relationship curve drawing and data conversion, the key water resource information of the underground river ecological storage capacity is finally obtained.

[0063] Further, the soil state information includes soil humidity, soil texture, soil temperature and soil structure; and the meteorological feature data includes daily cumulative average illumination, daily average illumination, daily maximum illumination, daily average temperature, diurnal average temperature, daily maximum temperature, daily minimum temperature, daily average humidity, diurnal average humidity, daily maximum humidity, daily minimum humidity.

[0064] Further, training the machine learning model using the training set comprises: obtaining a sampling time period corresponding to each sample data in the training set, setting a sampling rate corresponding to the sample data, and obtaining the sample data after sampling; determining an importance value of the sample data, correcting the sample data after sampling by using the importance value to obtain the sample data after correction; inputting the sample data after correction into the machine learning model, and training the model in combination with a loss function.

[0065] Determining the importance value of each sample data comprises: correcting the initial importance value of the sample data after sampling based on the corresponding sampling rate to obtain the importance value of the sample data after sampling; wherein the importance value and the initial importance value are in direct proportion, and the importance value and the sampling rate of the sample data after sampling are in inverse proportion.

[0066] Obtaining the sample data after correction comprises: multiplying the importance value and the corresponding sample data after sampling to obtain the sample data after correction.

[0067] Specifically, first, a complete training set is collected, which contains a large amount of sample data. These sample data may come from different data sources and cover various features and target variables. For each sample data in the training set, its corresponding sampling time period is obtained. The sampling time period refers to the time range in which the sample data is collected, which is very important for subsequent sampling rate setting and data processing.

[0068] Setting the sampling rate and obtaining the sample data after sampling: according to the sampling time period of the sample data, set a suitable sampling rate. The sampling rate refers to the frequency of collecting data in unit time, for example, collecting once per second, collecting once per minute, etc. The choice of sampling rate depends on the characteristics of the data, computing resources and the needs of model training. According to the set sampling rate, the original sample data is sampled to obtain the sample data after sampling. The sampling process can use simple uniform sampling method, or use more complex sampling strategy such as random sampling, stratified sampling, etc., to ensure that the sampled data can represent the distribution of the original data.

[0069] For each sample data after sampling, its initial importance value is determined. The initial importance value can be determined based on the characteristics of the sample data, the distribution of the target variable, the rarity of the data, etc. For example, if a certain sample data has unique characteristics in the target variable or belongs to a minority category, its initial importance value may be higher.

[0070] The initial importance value of the sampled sample data is modified according to the sampling rate. The modification rule is that the importance value is proportional to the initial importance value and inversely proportional to the sampling rate. This means that the lower the sampling rate (i.e. the lower the data acquisition frequency), the higher the importance value of the sample data. This is because the data at a low sampling rate may contain more information or be more representative in the time series.

[0071] The modified importance value is calculated, and the formula can be expressed as: importance value = initial importance value × sampling rate. In this way, the importance value of each sampled sample data is obtained.

[0072] The importance value is multiplied by the sampled sample data. For each sampled sample data, it is multiplied by the corresponding importance value. The purpose of this step is to weight the sample data according to its importance, so that in the model training process, important sample data has a greater impact on the model.

[0073] The modified sample data can be represented as: modified sample data = sampled sample data × importance value; in this way, each sample data is given different weights, reflecting its importance in model training.

[0074] Train the machine learning model: input the modified sample data into the machine learning model, and input the modified sample data into the machine learning model as input data. The machine learning model can be any type of model, such as linear regression, decision tree, neural network, etc., the specific choice depends on the nature of the problem and the characteristics of the data. In the model training process, a loss function is used to evaluate the performance of the model. The loss function measures the difference between the model's predicted value and the actual target value, common loss functions include mean square error (MSE), cross-entropy loss, etc. Through optimization algorithms such as gradient descent, the parameters of the model are adjusted to minimize the value of the loss function. In the training process, the importance value of the modified sample data will affect the calculation of the loss function, so that the model pays more attention to important sample data. The training process will be iterated until the performance of the model reaches a satisfactory level or meets the preset training conditions.

[0075] During training, the validation set is used regularly to validate the model and evaluate its generalization ability. According to the validation results, the parameters, structure or training strategy of the model may need to be adjusted. For example, if the model performs poorly on the validation set, the complexity of the model may need to be adjusted, the form of the loss function may need to be changed, or the training hyperparameters may need to be adjusted, etc.

[0076] After the training is completed, the model is comprehensively evaluated using the test set. Evaluation indicators can include accuracy, recall rate, F1 score, mean square error, etc., and the specific choice depends on the type of problem and business needs. Through the evaluation results, it is judged whether the model has reached the expected performance target. If the model performance is not ideal, it may be necessary to return to the data preparation stage, re-collect data or adjust data processing methods, and then re-train the model. If the model evaluation result is satisfactory, the trained model is deployed to actual application. In actual application, the model can be used for prediction, classification, decision support, etc., to provide valuable output for business. At the same time, the performance of the model needs to be continuously monitored to ensure that it can work stably and accurately in actual operation. If the performance of the model is found to be declining, the model needs to be updated and optimized in a timely manner.

[0077] The entire model training process is a systematic engineering involving data preparation, sample data importance evaluation, sample data correction, model training, model verification and adjustment, model evaluation and application, etc. Through reasonable data processing and model training strategies, the performance and generalization ability of the model can be improved, so that it can be better applied to practical problems.

[0078] Further, the irrigation of the vegetation in the target underground river basin includes: when the vegetation water requirement is less than the lower limit of the preset water requirement, controlling the ground water storage facility group to irrigate; updating the vegetation water requirement according to the vegetation water requirement after irrigation, and when the updated vegetation water requirement is greater than the upper limit of the preset water requirement, controlling the ground water storage facility group to stop irrigation.

[0079] Specifically, the monitored vegetation water requirement is compared with the preset water requirement lower limit value; wherein the preset water requirement lower limit value is preset according to the physiological needs of the vegetation, the growth stage, the climate conditions and other factors, and is used to judge whether the vegetation needs irrigation.

[0080] Determine the irrigation condition: if the monitored vegetation water requirement is less than the lower limit of the preset water requirement, it means that the vegetation is in a water shortage state and needs to be irrigated. At this time, the irrigation start condition is triggered.

[0081] Control the ground water storage facility group to irrigate: according to the irrigation start condition, the automatic control system controls the ground water storage facility group to start irrigation. The automatic control system can automatically adjust the flow, time and area of the irrigation facilities according to the preset irrigation strategy, to ensure the efficiency and accuracy of the irrigation process.

[0082] During the irrigation process, the irrigation flow, irrigation time and other parameters are monitored in real time to ensure that the irrigation process meets the preset irrigation plan. During the irrigation process, the water requirement of the vegetation is continuously monitored in real time. The monitoring device can feed back the change of the water requirement of the vegetation after irrigation in real time, so as to timely adjust the irrigation strategy. The water requirement of the vegetation after irrigation is compared with the preset upper limit value of the water requirement. The preset upper limit value of the water requirement is also preset according to the physiological demand, growth stage, climate condition and other factors of the vegetation, and is used to judge whether the vegetation has been irrigated sufficiently. According to the monitored water requirement of the vegetation after irrigation, the record of the water requirement of the vegetation is updated. The updated water requirement of the vegetation will be used as the basis for subsequent irrigation decision. If the updated water requirement of the vegetation is greater than the upper limit value of the preset water requirement, it means that the vegetation has been irrigated sufficiently, and it is not necessary to continue irrigation. At this time, the irrigation stop condition is triggered. According to the irrigation stop condition, the automatic control system controls the ground water storage facility group to stop irrigation. The automatic control system can automatically close the valve of the irrigation facility, stop the water pump and other operations to ensure the timely stop of the irrigation process. After stopping irrigation, the total time, total water volume and other parameters of the irrigation are recorded to evaluate the irrigation effect and adjust the subsequent irrigation plan.

[0083] The above-described embodiments are only descriptions of the preferred modes of the present application, and do not limit the scope of the present application. Without departing from the design spirit of the present application, various modifications and improvements to the technical solutions of the present application made by those skilled in the art shall fall within the protection scope determined by the claims of the present application.

Claims

1. A method for water resources regulation and storage of an underground river basin based on zoning cooperation, characterized in that, The method comprises the following steps: setting a network of intercepting ditches and a group of surface water storage facilities in a target underground river basin for irrigation of vegetation in the target underground river basin; during the irrigation of the vegetation in the target underground river basin, collecting soil state information and meteorological characteristic data of the soil in the target underground river basin, inputting the soil state information and the meteorological characteristic data into a vegetation water requirement model to obtain a vegetation water requirement; the vegetation water requirement model is obtained by training a machine learning model using a training set; the training set comprises original soil state information and original meteorological characteristic data.

2. The zoning-collaboration-based underground river basin water resource regulation method according to claim 1, characterized in that, The setting of the network of intercepting ditches comprises: dredging in a distribution area of tourist and consumption sinkholes in a target underground river basin, and adopting a hierarchical structure to sequentially set a water diversion channel and three-stage sand settling tanks in the distribution area of the tourist and consumption sinkholes, and disposing a sand blocking net at the connection of the three-stage sand settling tanks, and setting a circular sinkhole pool at the sinkhole opening, and causing the treated rainwater to flow into the underground river from the sinkhole pool after passing through the three-stage sand settling tanks.

3. The zoning-collaboration-based underground river basin water resource regulation method according to claim 1, characterized in that, The setting of the group of surface water storage facilities comprises: setting two-stage regulating and storing dams in a sky window area of an underground river in a target underground river basin; wherein the two-stage regulating and storing dams adopt a double-layer structure of a bottom concrete and an upper hydraulic steel dam; setting a water energy dam in the sky window area of the underground river between the two-stage regulating and storing dams; the water energy dam comprises a water diversion channel, a blocking and storing dam, and a water energy pump; the water diversion channel is used for guiding the flow direction and flow rate of underground water; the blocking and storing dam is used for setting a water level device to collect water level data, and determining an underground river ecological storage capacity by using the water level data, so as to allocate underground river water resources; the water energy pump is used for using the potential energy generated by the water level difference to lift the underground river water to a high place for irrigation of nearby farmland.

4. The zoning-collaboration-based underground river basin water resource regulation method according to claim 3, characterized in that, The determination of the underground river ecological storage capacity comprises: using a For loop tool and a surface volume tool in ArcGIS software to construct a model for estimating the water level-storage capacity relationship of a dammed reservoir through a model manager, and drawing a water level-storage capacity relationship curve of the dammed reservoir; determining the water surface area under different water levels according to the water level-storage capacity relationship curve of the dammed reservoir, and constructing a water surface area-storage capacity relationship curve; extracting the water level of the dammed reservoir under different time periods, obtaining the water surface area under the water level of the dammed reservoir according to the water level of the dammed reservoir, inputting the water surface area into the water surface area-storage capacity relationship curve, and obtaining the underground river ecological storage capacity under different time periods.

5. The zoning-collaboration-based underground river basin water resource regulation method according to claim 1, characterized in that, The soil state information comprises soil humidity, soil texture, soil temperature, and soil structure; The meteorological characteristic data comprises daily cumulative average illumination, daily average illumination, daily maximum illumination, daily average temperature, diurnal average temperature, daily maximum temperature, daily minimum temperature, daily average humidity, diurnal average humidity, daily maximum humidity, and daily minimum humidity.

6. The zoning-collaboration-based underground river basin water resource regulation method according to claim 1, characterized in that, The training of the machine learning model using the training set comprises: obtaining the sampling time period corresponding to each sample data in the training set, setting the sampling rate corresponding to the sample data, and obtaining the sampled sample data; determining the importance value of the sample data, correcting the sampled sample data using the importance value, and obtaining the corrected sample data. The machine learning model is input with the corrected sample data, and model training is performed in combination with a loss function.

7. The zoning-collaboration-based underground river basin water resource regulation method according to claim 6, characterized in that, The importance value of each sample data is determined by: The initial importance value of the sampled sample data is corrected based on the corresponding sampling rate to obtain the importance value of the sampled sample data. The importance value and the initial importance value are in a direct proportion relationship, and are in an inverse proportion relationship with the sampling rate of the sampled sample data.

8. The zoning-collaboration-based underground river basin water resource regulation method according to claim 6, characterized in that, The corrected sample data is obtained by multiplying the importance value by the corresponding sampled sample data.

9. The zoning-collaboration-based underground river basin water resource regulation method according to claim 1, characterized in that, Irrigating the vegetation in the target underground river basin includes: When the vegetation water requirement is less than the lower limit of the preset water requirement, the surface water storage facility group is controlled to irrigate; The vegetation water requirement is updated according to the irrigation of the vegetation, and when the updated vegetation water requirement is greater than the upper limit of the preset water requirement, the surface water storage facility group is controlled to stop irrigation.

Images