Dynamic control method and platform for intelligent interception and drainage equipment

Abstract

The present invention discloses a dynamic control method and platform for intelligent interception and drainage equipment, and relates to the technical field of interception and drainage control. The method includes: analyzing drainage parameters of multi-level drainage equipment, constructing a multi-level drainage network; connecting multi-source collection channels; based on the multi-level drainage network, predicting the external drainage demand through multi-source collection data, and obtaining predicted external drainage information; extracting adaptive demand parameters with adaptive index parameters according to the predicted information; using the adaptive demand parameters as indexes, inputting an adaptive search model to search for control parameters, and obtaining dynamic control parameters; and using dynamic control parameters to control the interception and drainage of multi-level drainage equipment. The present invention solves the technical problems of poor adaptability of the interception and drainage system to complex environmental changes and inaccurate drainage demand prediction, and achieves the technical effect of dynamically controlling drainage according to weather changes, improving the accuracy of drainage demand prediction, and avoiding waterlogging in farmland.

Description

Technical Field

[0001] The present invention relates to the technical field of interception and drainage control, and in particular to a dynamic control method and platform for intelligent interception and drainage equipment. Background Art

[0002] With the continuous development of global climate change and agricultural production, drainage management plays an increasingly important role in agricultural irrigation, flood control and urban drainage. In particular, how to achieve efficient and accurate drainage control becomes particularly critical in the face of extreme climate, sudden rainfall and soil moisture changes. Existing drainage control technologies mainly focus on the adjustment of single drainage equipment or control parameters, lacking intelligent methods for the integration of multi-source data and dynamic control. In particular, in the face of changes in weather, crop growth differences and terrain complexity, it is difficult to accurately predict drainage needs and adjust drainage strategies in real time, resulting in low drainage efficiency, energy waste, and even problems such as insufficient drainage or floods, which damage large areas of farmland and affect the stability of agricultural production. Summary of the invention

[0003] The present application provides a dynamic control method and platform for intelligent interception and drainage equipment, which is used to solve the technical problems of poor adaptability of the interception and drainage system to complex environmental changes and inaccurate drainage demand prediction.

[0004] In view of the above problems, the present application provides a dynamic control method and platform for intelligent interception and drainage equipment.

[0005] The first aspect of the present application provides a dynamic control method for intelligent interception and drainage equipment, the method comprising: parsing drainage parameters of multi-level drainage equipment and constructing a multi-level drainage network; connecting multi-source acquisition channels, the multi-source acquisition channels comprising weather data acquisition channels, image data acquisition channels, and crop data acquisition channels; based on the multi-level drainage network, predicting the external drainage demand according to the multi-source acquisition data acquired by the multi-source acquisition channels, and obtaining predicted demand external drainage information; extracting index parameters according to the predicted demand external drainage information with preset adaptive index parameters, and obtaining adaptive demand parameters; using the adaptive demand parameters as indexes, inputting them into an adaptive search model to search for control parameters, obtaining dynamic control parameters, and utilizing the dynamic control parameters to perform interception and drainage control of the multi-level drainage equipment, wherein the adaptive search model is obtained based on the multi-level drainage network and optimized spatial integrated training.

[0006] The second aspect of the present application provides a dynamic control platform for intelligent interception and drainage equipment, and the platform includes: a network construction module: parsing the drainage parameters of multi-level drainage equipment and constructing a multi-level drainage network; a channel connection module: connecting multi-source acquisition channels, the multi-source acquisition channels include weather data acquisition channels, image data acquisition channels, and crop data acquisition channels; an external discharge demand prediction module: based on the multi-level drainage network, the external discharge demand is predicted according to the multi-source acquisition data obtained by the multi-source acquisition channel, and the predicted demand external discharge information is obtained; an index parameter extraction module: according to the predicted demand external discharge information, the index parameter is extracted with preset adaptive index parameters to obtain adaptive demand parameters; an interception and drainage control module: using the adaptive demand parameter as an index, inputting it into an adaptive search model to search for control parameters, obtaining dynamic control parameters, and using the dynamic control parameters to perform interception and drainage control of multi-level drainage equipment, wherein the adaptive search model is obtained based on the multi-level drainage network and optimized spatial integrated training.

[0007] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0008] Analyze the drainage parameters of multi-level drainage equipment and construct a multi-level drainage network; connect multi-source acquisition channels, which include weather data acquisition channels, image data acquisition channels, and crop data acquisition channels; based on the multi-level drainage network, predict the drainage demand according to the multi-source acquisition data obtained by the multi-source acquisition channel, and obtain the predicted demand drainage information; extract the index parameters according to the predicted demand drainage information with the preset adaptive index parameters to obtain the adaptive demand parameters; use the adaptive demand parameters as indexes, input them into the adaptive search model to search for control parameters, obtain dynamic control parameters, and use the dynamic control parameters to control the interception and drainage of multi-level drainage equipment, wherein the adaptive search model is obtained based on the multi-level drainage network and optimized spatial integrated training. The technical effect of dynamically controlling drainage according to weather changes, improving the accuracy of drainage demand prediction, and avoiding waterlogging in farmland is achieved. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of 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 creative work.

[0010] Figure 1 A schematic flow chart of a dynamic control method for intelligent interception and drainage equipment provided in an embodiment of the present application.

[0011] Figure 2A schematic diagram of the dynamic control platform structure of the intelligent interception and drainage equipment provided in an embodiment of the present application.

[0012] Explanation of the reference numerals: network construction module 1, channel connection module 2, external discharge demand prediction module 3, index parameter extraction module 4, interception and discharge control module 5. DETAILED DESCRIPTION

[0013] The present application provides a dynamic control method and platform for intelligent interception and drainage equipment, which is used to solve the technical problems of poor adaptability of the interception and drainage system to complex environmental changes and inaccurate drainage demand prediction.

[0014] The technical solutions in the embodiments of the present application will be described clearly and completely below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0015] Embodiment 1, as Figure 1 As shown, the present application provides a dynamic control method for intelligent interception and drainage equipment, the method comprising:

[0016] Analyze the drainage parameters of multi-level drainage equipment and construct a multi-level drainage network.

[0017] In one embodiment, relevant parameters of drainage equipment at different levels are first collected. These equipment include field drainage layer, secondary drainage layer, main drainage layer, etc. Each level has different drainage capacity, working intensity, flow rate and drainage path during the drainage process. Therefore, the analysis of drainage parameters not only needs to pay attention to the specific drainage capacity of each level, but also needs to consider the configuration of drainage facilities, the working status and location of equipment, etc. Then, according to the analyzed parameters, these drainage equipment are constructed into a multi-level drainage network. Each node in the network represents a control device or drainage facility, and the connection between nodes reflects the drainage path and drainage direction. Through such analysis and network construction, basic data and decision-making basis are provided for subsequent drainage scheduling and control.

[0018] In a possible implementation, analyzing the drainage parameters of the multi-level drainage equipment and constructing the multi-level drainage network includes:

[0019] The control settings and equipment distribution of multi-level drainage equipment are obtained, and the multi-level drainage equipment includes field drainage, secondary drainage, main drainage, water storage regulation, and external drainage emergency; the control node setting position, control drainage equipment level, control parameters, drainage cross-section, drainage intensity and direction are obtained according to the control settings, and the control parameters include valve opening or pump station power; the control node is used as the network node, and the equipment distribution of the control drainage equipment level corresponding to the control node setting position is used as the connecting edge, the drainage equipment level is used as the edge attribute, and the control parameters, drainage cross-section, drainage intensity and direction are used for edge assignment to construct the multi-level drainage network.

[0020] Optionally, when constructing a multi-level drainage network, first of all, the control settings and equipment distribution of the multi-level drainage equipment are determined. The multi-level drainage equipment includes a field drainage layer, a secondary drainage layer, a main drainage layer, a water storage and regulation layer, and an external drainage emergency layer. The equipment distribution and functions of different levels are different. It is necessary to analyze the equipment parameters of each level in detail to obtain the control node setting position, control drainage equipment level, control parameters, drainage cross-section, drainage intensity and direction; among them, the field drainage layer includes drainage ditches / culverts and drainage wells / manholes. The drainage ditches or culverts are located in the farmland to collect and guide the accumulated water inside the farmland. The drainage wells / manholes are used for equipment maintenance and as nodes for network connection. The control parameter involved is the valve opening, which is used to regulate the flow of the field drainage layer; the secondary drainage layer includes secondary drainage ditches / channels and regulating gates / regulating valves. The secondary drainage ditches / channels are connected to the field drainage layer to further collect and transport the water flow to the main drainage layer. The regulating gates / regulating valves are used to control the water flow speed and flow. The control parameter involved is the valve opening; the main drainage layer includes Drainage ditches / channels and drainage pump stations, the main drainage layer is responsible for collecting and transporting the water in the secondary drainage layer, and discharging it to the outside through the pump station. The control parameter involved is the power of the pump station. The water storage and regulation layer includes reservoirs / ponds and control gates / valves. The main function is to temporarily store rainwater during the drainage process and play a role in regulating the amount of water. The control parameter involved is the valve opening. The external drainage emergency layer includes external drainage pump stations and flood discharge channels. Under extreme weather conditions, the external drainage emergency layer discharges the remaining water in the field drainage layer into the farmland area through external drainage pump stations and flood discharge channels. The control parameter involved is the power of the pump station. Subsequently, the control nodes are used as the nodes of the network, corresponding to the setting positions of drainage equipment at each level, and the control drainage equipment level is used as the edge attribute of the network, representing the connection relationship between different drainage levels. The control parameters, drainage cross-section, drainage intensity and drainage direction are used as the edge assignment parameters to describe the relationship and status between the equipment at each level. Through the above analysis and network construction, a complete multi-level drainage network can be formed to provide data support and basic framework for subsequent dynamic control.

[0021] Connecting to a multi-source acquisition channel, the multi-source acquisition channel includes a weather data acquisition channel, an image data acquisition channel, and a crop data acquisition channel.

[0022] In one embodiment, a multi-source acquisition channel is activated and a communication connection is established with the multi-source acquisition channel. The multi-source acquisition channel refers to real-time acquisition of various data affecting drainage control through different data acquisition devices. The multi-source acquisition channel includes a weather data acquisition channel, an image data acquisition channel and a crop data acquisition channel, which are used to comprehensively reflect the external environment and internal state of farmland drainage. Among them, the weather data acquisition channel monitors weather changes in real time through meteorological sensors, rain gauges, anemometers and other equipment, and collects parameters such as rainfall, rainfall intensity, rainfall duration, wind speed, humidity, and air pressure. At the same time, weather forecast data can also be obtained through the data interface or network platform with the meteorological station. These weather data are used to analyze the upcoming rainfall situation and provide a basis for the prediction and regulation of drainage. The image data acquisition channel is connected to the weather station through the weather station. Through the use of cameras, drones and other equipment installed in the fields or drainage facilities to capture images, all-weather monitoring is achieved to collect information on soil waterlogging conditions in farmland and water flow conditions in drainage channels. These image data are used to identify soil waterlogging areas, waterlogging depth and other information, providing an intuitive basis for drainage demand analysis and equipment control; the crop data collection channel uses soil moisture sensors, vegetation index sensors, crop monitors and other equipment for real-time monitoring, collecting data on crop growth conditions and water demand, including crop growth stages, soil moisture content, root health, etc., and can also be combined with satellite remote sensing data for large-scale farmland crop monitoring; these crop data reflect the current crop demand for water, provide accurate data support for drainage regulation, and ensure that the drainage process does not affect the normal growth of crops.

[0023] Based on the multi-level drainage network, drainage demand forecasting is performed according to the multi-source collection data acquired by the multi-source collection channel to obtain forecast demand drainage information.

[0024] In one embodiment, real-time weather information is obtained through the connected weather data collection channel, including meteorological data such as rainfall, rainfall intensity, and duration. Through the image data collection channel, image data of farmland and drainage channels are collected to identify soil waterlogging conditions, waterlogging depth, distribution area, etc. Through the crop data collection channel, the growth status of crops, soil moisture content, and water demand characteristics at different growth stages are monitored. The above data are integrated to form multi-source collection data to provide basic data support for external drainage demand prediction; then, based on the nodes in the multi-level drainage network, the collected weather data, image data, and crop data are spatially located to determine the specific locations and data correspondences in different drainage areas. For example, the soil waterlogging characteristic distribution map is converted into a spatial map. Align with the location of drainage equipment in the multi-level drainage network to identify areas with waterlogging problems and the corresponding drainage equipment and paths in these areas. Then, perform time series processing on weather data and crop data to extract weather change trends and crop water demand changes to form dynamic data in the time dimension. Afterwards, use the time series prediction model and these data to make a time-accumulated prediction of drainage demand, determine the external drainage demand at different time nodes, and generate predicted demand external drainage information, including the external drainage demand, drainage intensity, and drainage path of each divided control area at different time nodes. These external drainage information will serve as the basic data for subsequent control parameter optimization and dynamic drainage regulation to ensure that farmland has reasonable and timely external drainage capacity under extreme weather conditions to prevent the occurrence of waterlogging disasters.

[0025] In a possible implementation, based on the multi-level drainage network, the drainage demand is predicted according to the multi-source collection data acquired by the multi-source collection channel to obtain the predicted demand drainage information, including:

[0026] Based on weather data and crop data, coordinate positioning and feature time series are performed to construct weather positioning distribution, weather time series features, crop growth feature positioning distribution, and crop growth time series features; image soil waterlogging feature recognition and coordinate positioning are performed based on image data to construct a soil waterlogging feature distribution map; the soil waterlogging feature distribution map is aligned with the position positioning of the multi-level drainage network to determine the data positioning mapping relationship, and the weather positioning distribution and the crop growth feature positioning distribution are fitted to the data positioning mapping relationship; the weather time series features and crop growth time series features are used for time series alignment, the data positioning mapping relationship is constructed by time series moving window, and the data positioning mapping relationship is aligned by time series moving window relationship according to the acquisition time of the image data, wherein when the acquisition time of the image data is missing, the image data of the adjacent acquisition time or the average of the image data of the adjacent acquisition time at both ends is used to fill it; the drainage demand analysis is performed based on the data positioning mapping relationship to determine the single window drainage demand, and the single window drainage demand is predicted in time series accumulation according to the time series moving window to obtain the predicted demand external drainage information.

[0027] Optionally, the geographical location information obtained by the weather data collection equipment and the crop monitoring equipment is used to perform spatial coordinate positioning of the collected weather data and crop data, clarify the specific areas to which the weather data and crop growth data belong, and construct a weather positioning distribution map based on these spatial coordinate positioning, that is, the spatial distribution of weather information in each area, and a crop growth characteristic positioning distribution map, that is, the spatial distribution of crop status and water demand characteristics in each area. In addition, the collected weather data and crop growth data are sorted in chronological order to reflect the law of data change over time, forming weather time series characteristics and crop growth time series characteristics for subsequent time series analysis; then, through The image data of the farmland is obtained through the image data acquisition equipment, and the spatial coordinate location of each image data is determined. Then, the smoothing method such as Gaussian filtering is used to remove noise and reduce the impact of uneven lighting. The histogram equalization method is applied to improve the image contrast, making the wet area, the waterlogged area and the background area clearer. The image data is converted into HSV image data for subsequent feature analysis. In the HSV color space, the wet soil area usually presents a darker tone (low brightness) and a higher saturation. The waterlogged area has a high reflectivity, a higher brightness and a lower saturation. By setting different color thresholds, such as the waterlogged brightness threshold and the waterlogged saturation threshold, the HSV image data is segmented to generate an image segment. In the image segmentation data, the white area indicates water accumulation, the gray area indicates moisture, and the black area indicates other backgrounds; then, according to the spatial coordinate positioning of each image data, the identified soil water accumulation area is mapped to the actual spatial coordinate to form a soil water accumulation characteristic distribution map, in which the distribution of water accumulation is marked; after obtaining the soil water accumulation characteristic distribution map, the soil water accumulation characteristic distribution map is aligned with the position positioning in the multi-level drainage network to clarify the corresponding relationship between each drainage equipment and the water accumulation area, thereby determining the data positioning mapping relationship between the multi-level drainage equipment and the water accumulation area, and then the weather positioning distribution map and the crop growth characteristic positioning distribution map are simulated according to the spatial coordinate positioning. Combine and uniformly map them into the data location mapping relationship to realize the association of various types of data in space and position; then, obtain the acquisition time of each image in the data location mapping relationship, determine the time dimension of the data location mapping relationship, and align the weather time series characteristics and the crop growth time series characteristics according to this time dimension, and then build a time series moving window based on the time dimension. The time series moving window is a method of analyzing data by time segmentation, which is used to extract weather, crop growth and image features within each time window. The size of the window determines the time range of the analysis, for example, 6 hours, 1 day, etc., and the sliding step of the window determines the analysis accuracy of the time series, for example, sliding 1 hour or half a day each time;When the image data acquisition time is missing, the data needs to be supplemented to ensure the continuity of the analysis. If the image data at a certain time point is missing, the image data at the previous or next time point can be used as a supplement. When there is a significant difference between the data before and after the missing time point, the mean of the image data at both ends of the time point is used for interpolation and supplementation. For example, if the data at time point T4 is missing, the data value of T4 can be estimated by the mean of T3 and T5. This supplementation method can maximize the integrity of the data and the accuracy of the time series analysis. After completing the construction of the time series moving window, it is necessary to perform time series alignment on the data positioning mapping relationship, align the weather characteristics and crop growth characteristics with the soil waterlogging characteristic distribution map in spatial position, and use the mapping relationship to connect various feature data with the control nodes and The equipment is associated, and in each time series moving window, according to the position of the drainage equipment, the distribution characteristics of the waterlogged area are matched with the corresponding drainage equipment to generate the final data positioning mapping relationship. For example, the waterlogging in a certain area is connected to the pump station P1 through the control node N1. The characteristic data in the window include parameters such as rainfall, moisture content of moist soil, and degree of waterlogging. Finally, based on the data positioning mapping relationship, in each time series moving window, the soil waterlogging, weather changes and crop growth characteristics are analyzed to comprehensively judge the drainage needs of each area. Then, based on the time series moving window, the single window drainage demand is accumulated over time through the time series prediction model to predict the change trend of drainage demand in the future period of time, obtain the drainage demand at each time node, form the predicted demand outflow information, and provide a basis for subsequent dynamic drainage regulation. ;

[0028] In a possible implementation, performing drainage demand analysis based on the data location mapping relationship to determine the single window drainage demand, performing time series cumulative prediction on the single window drainage demand according to the time series moving window, and obtaining the predicted demand drainage information includes:

[0029] The acquisition time of the image data is obtained, and an entity time series moving window is determined, where the entity time series moving window is a time series moving window corresponding to the acquisition time closest to the occurrence of weather and rainfall; according to the entity time series moving window, the drainage demand of each area is calculated based on the soil water accumulation characteristics and crop growth characteristics of each area according to the segmented control area of ​​the data positioning mapping relationship, and the single window drainage demand is obtained, and the single window drainage demand is used to represent the drainage demand of each segmented control area in the entity time series node; the single window drainage demand is used as the node state, and the state change amount is fitted according to the weather time series characteristics of the time series moving window to construct a time series prediction model; the time series cumulative prediction of the time series moving window is performed through the time series prediction model, and the drainage demand of each segmented control area corresponding to each time node is obtained as the predicted demand external drainage information.

[0030] Optionally, first, obtain the acquisition time of the image data, and determine the entity time series moving window according to these acquisition time points. The definition range of the entity time series moving window includes the period before the weather rainfall occurs and the time series moving window closest to the image acquisition time. By determining the entity time series moving window, it can be ensured that the calculation of drainage demand is based on the latest image data and weather data; in each entity time series moving window, combined with the data positioning mapping relationship, the analysis range is divided into multiple independent segmented control areas, and each area corresponds to one or more drainage equipment; in each segmented control area, the influence of soil water accumulation characteristics and crop growth characteristics on drainage demand is calculated. Specifically, the crop growth characteristics within the entity time series moving window are mapped to the drainage demand. The time series characteristics of crop growth (such as growth stage, soil moisture content, and root health) are input into the water requirement depth prediction model. This water requirement depth prediction model can be constructed based on long short-term memory networks, etc. Through the analysis of the water requirement depth prediction model, the required water requirement depth in the segmented control area can be obtained, and then the water accumulation depth and water requirement depth of the segmented control area are used to calculate the drainage demand, that is, the single-window drainage demand of the segmented control area is obtained by difference, that is, in the physical time series node, the drainage demand of the segmented control area. If the single-window drainage demand is positive, it means that the current area has too much water and needs to be drained. If it is negative, it means that the current area does not need to be drained and may even need to retain water. ; Then, the single-window drainage demand in each entity time series moving window is defined as the node state of each control area. Based on these node states, the state change amount is fitted in combination with the weather time series characteristics of each entity time series moving window, that is, according to the same method as above, the node state of each time series moving window is calculated, and the node state of each time series moving window is calculated The difference between the node state of the corresponding entity time series moving window is calculated to obtain the state change amount, and then the time series prediction model is constructed based on the state change amount using LSTM (long short-term memory network), Markov chain and other methods; Taking LSTM as an example, all node states, weather time series characteristics and state changes are counted to construct training sets and test sets. Then, the time series prediction model structure is set based on LSTM, including the input layer, LSTM hidden layer, and output layer; after that, the training set is input into the time series prediction model, the predicted state change is calculated through forward propagation, the loss value between the predicted state change and the actual state change is calculated through the mean square error, the loss gradient is calculated through back propagation, and then the parameters of the time series prediction model are optimized based on the loss gradient by the optimizer to minimize the error between the predicted state change and the actual state change; after the training is completed, that is, the error reaches the preset value or reaches the maximum number of iterations, the current time series prediction model is tested with the test set, and if the test result meets expectations, the current time series prediction model is output;After the time series prediction model is constructed, the data of the time series moving window is input into the trained time series prediction model to predict the changes in drainage demand at multiple time nodes in the future for each segmented control area. The predicted drainage demand is accumulated by time node to form a time series cumulative prediction result. Specifically, at each time node, the predicted drainage demand values ​​at the current moment and the previous moment are accumulated to form the predicted demand external drainage information. This information not only includes the future drainage demand of each segmented control area, but also reflects the time change trend of the demand, providing an accurate reference for the dynamic control of multi-level drainage equipment, and avoiding farmland waterlogging or water shortage problems caused by insufficient or excessive drainage. ;

[0031] According to the predicted demand outflow information, index parameters are extracted using preset adaptive index parameters to obtain adaptive demand parameters.

[0032] In one embodiment, drainage demand data of each segmented control area is extracted by predicting demand outflow information, including the size and distribution characteristics of demand, and then combined with the control path and related parameters of drainage equipment in the multi-level drainage network, the preset adaptive index parameters are used to screen and match data. The preset adaptive index parameters are key parameters set based on the characteristics of the drainage network, including control parameters such as valve opening and pump station power. Subsequently, the drainage equipment corresponding to each area and its related control parameters are determined by matching the segmented control area in the predicted demand outflow information with the control path of the drainage network, and then the correlation of the control parameters is analyzed in combination with the regional drainage demand, for example, evaluating whether the valve opening or pump station power of the current equipment meets the demand, and whether the drainage flow direction conforms to the predicted drainage pressure distribution. After completing the above matching and analysis, the preset adaptive index parameter analysis and extraction are performed according to the segmented control area and demand of the predicted demand outflow information to obtain adaptive demand parameters. These adaptive demand parameters dynamically reflect the matching degree between the regional drainage demand and the equipment capacity, and can provide direct data support for subsequent equipment scheduling and operation optimization. Through this process, drainage control can achieve refined regulation based on real-time demand to ensure the maximum operation efficiency and drainage effect of the multi-level drainage network.

[0033] In a possible implementation, according to the predicted demand outsourcing information, extracting index parameters with preset adaptive index parameters to obtain adaptive demand parameters includes:

[0034] According to the control parameters, the control paths of the multi-level drainage network are marked with control parameters; the control paths of the multi-level drainage network are divided and matched with the predicted demand outflow information to obtain the control parameters of the control areas; the control parameters of the control areas are mined for correlation with outflow demand parameters to determine control association parameters; the control association parameters are used as preset adaptive index parameters, and preset adaptive index parameter analysis and extraction are performed according to the divided control areas and demand of the predicted demand outflow information to obtain the adaptive demand parameters.

[0035] Optionally, the control path of the multi-level drainage network is marked according to the control parameters. The key parameters in the control path mainly include valve opening and pump station power. These parameters are mapped to the nodes and edges in the drainage network by analyzing the physical layout and operating characteristics of the multi-level drainage equipment. For example, in the field drainage layer, the valve opening controls the amount of water flowing through, while in the main drainage layer, the pump station power determines the drainage flow rate and intensity. By marking these control parameters, a complete parameterized multi-level drainage network is formed, which provides a basis for subsequent analysis; then, the control path of the multi-level drainage network is matched with the predicted demand outflow information to determine the segmentation. The control parameters corresponding to the control area and the predicted demand external drainage information provide the drainage demand of each area, and the control path annotation describes the correlation between the control node and the drainage equipment. By matching the segmented control area and the control path, the drainage demand and the corresponding control parameters of each area can be clarified; then, the correlation between the control parameters of the segmented control area and the external drainage demand parameters is mined, focusing on analyzing the key factors affecting the drainage efficiency, such as the height difference of the terrain. These characteristics determine the specific adjustment range of the control parameters in the drainage process. For example, when the terrain difference is large, the natural discharge of water is better, and the power of the pump can be appropriately reduced, while in areas with relatively high terrain, the water flow is discharged naturally. For flat areas, the power of the pump station needs to be increased to overcome the resistance. After completing the above correlation mining, the control parameter relationships under the influence of these characteristics are integrated into control correlation parameters. The control correlation parameters dynamically reflect the comprehensive relationship between the terrain characteristics and drainage demand of the segmented control area. The control correlation parameters are used as preset adaptive index parameters. For the segmented control area and demand in the predicted demand outflow information, the correlation analysis and parameter extraction between demand and control force are further combined with the specific conditions of different drainage equipment, pipeline characteristics and drainage paths, and finally the adaptive demand parameters are obtained. Specifically, the segmented control is determined based on the predicted demand outflow information. The drainage demand of the region. For example, in the field drainage layer, the drainage demand of a certain area is 10 cubic meters. In the secondary drainage layer, the regional drainage demand is greater, and the water in the field needs to be guided to the secondary drainage layer. In the main drainage layer, the drainage demand is further expanded, and it is responsible for collecting the water from the secondary drainage to the main drainage layer for rapid discharge. In each divided control area, according to the current drainage demand and the functional characteristics of the drainage equipment level, analyze the key pipeline characteristics that affect the drainage process, including the length, distance, diameter, terrain drop and flow resistance of the pipeline. These pipeline characteristics directly affect the resistance of water flow and the control force required by the drainage equipment;According to the drainage demand of the segmented control area and the physical characteristics of the pipeline, the control force required to meet the demand, such as the power of the pump or the opening of the valve, is calculated. For example, in the primary drainage layer or the secondary drainage layer, if the pipeline is short and the terrain difference is large, the natural drainage effect of the water flow is better, and the control demand for the pump power is small. When the pipeline is long, the terrain is flat, and the flow resistance is large, in order to ensure that the drainage demand reaches 10 cubic meters, it is necessary to increase the pump power or increase the valve opening to overcome the resistance of water flow transmission; after completing the correlation analysis between the demand and the control force, the results are converted into adaptive demand parameters. These parameters include specific control values, such as the pump power setting required for the primary drainage layer, the optimal opening of the valve in the secondary drainage layer, etc.; in the segmented control area, the adaptive demand parameters can combine the functional characteristics of each drainage equipment level to achieve accurate drainage control in different segmented control areas under actual pipeline conditions, ensure that the drainage demand is met, and improve the operation efficiency of the entire multi-level drainage network. ;

[0036] In a possible implementation, mining the correlation of the control parameters of the control area with the external demand parameters to determine the control correlation parameters includes:

[0037] Establish a drainage relationship between the control parameters and drainage parameters; based on the multi-level drainage network, extract the drainage path, drainage section, drainage intensity and direction of the control area, and identify the regional terrain difference according to the location of the control area; based on the hydrodynamic simulation model, explore the correlation between the regional terrain difference, drainage path, drainage section, drainage intensity and direction and the drainage parameters; transfer parameter relationships according to the correlation and drainage relationship, determine the control relationship between the control parameters and the regional terrain difference, drainage path, drainage section, drainage intensity and direction, and determine the control correlation parameters.

[0038] Optionally, a drainage relationship between control parameters and drainage parameters is established. Control parameters include operation settings of drainage equipment, such as pump power, valve opening, etc., while drainage parameters include water flow velocity, water pressure, flow rate, etc. These parameters jointly determine the effect of the drainage process, that is, how to achieve effective discharge of water according to the control behavior of the equipment; then, based on the multi-level drainage network, the drainage path, drainage cross-section, drainage intensity and direction of the control area are extracted in detail. The drainage path is determined by analyzing the connection relationship of the multi-level drainage network, from the field drainage layer to the secondary drainage layer, and then to the main drainage layer. Each control area corresponds to a specific drainage path. The drainage cross section refers to the cross-sectional area of ​​a pipe or ditch, which affects the amount of water flowing through it, while the drainage intensity describes the rate and discharge capacity of the water flow. By controlling the location of the area, the terrain difference of the area can be further identified. For example, in areas with large terrain differences, water flows more easily by relying on gravity to flow naturally, while areas with flatter or low-lying terrain require external forces such as pump station lifting power to complete drainage. On this basis, based on the hydrodynamic simulation model, the correlation between regional terrain difference, drainage path, drainage cross section, drainage intensity and direction and drainage parameters is explored. The hydrodynamic simulation model can simulate the water flow in different pipes, channels and terrain conditions. The flow behavior under different terrain differences and the interaction between various parameters are analyzed based on fluid mechanics equations (such as the Navier-Stokes equations) and boundary conditions. By describing the velocity, pressure, flow rate and other characteristics of the water flow, and combining the actual terrain conditions, pipeline structure, cross-sectional dimensions and boundary conditions such as friction resistance, the dynamic calculation and simulation of the flow behavior in the drainage path are realized. During the simulation process, the drainage path, drainage cross-section, drainage intensity and direction are input into the hydrodynamic simulation model. The flow of water under different terrain differences and drainage paths is simulated by numerical calculation to obtain the variation law of drainage parameters such as water pressure, flow velocity and flow resistance. For example, When the terrain difference is large, the water flow rate increases and the flow resistance decreases. When the drainage path is long and the cross-sectional area is small, the water pressure loss increases and the flow rate decreases, requiring a higher-power pump station or a valve with a larger opening to compensate. Through multiple iterative simulations, the relationship between regional terrain difference, drainage path, drainage cross-section, drainage intensity and direction and drainage parameters is analyzed, and key influencing factors are extracted, such as the relationship between specific path length and flow resistance, the mapping of terrain difference and water flow pressure distribution, and the restriction of cross-sectional size on flow. Through these relationships, the dynamic matching relationship between control parameters (such as pump power and valve opening) and drainage efficiency can be quantified.By mining the above-mentioned association and drainage relationships, parameter relationship transfer is performed, and the control relationship between control parameters and regional terrain difference, drainage path, drainage section, drainage intensity and water flow direction is gradually determined. For example, in areas with large terrain differences, the flow velocity is affected by gravity acceleration, the power demand of the pump station is reduced, and the valve opening can be appropriately reduced. In areas with relatively flat or low-lying terrain, the gravity effect is insufficient, the water flow resistance increases, and it is necessary to increase the pump station power or adjust the valve opening to ensure the drainage velocity and flow. When the drainage section is large, the flow resistance of the water flow is small, and efficient drainage can be achieved by reducing the pump station power. In areas with small drainage sections, the water flow is prone to blockage or pressure increase. At this time, it is necessary to increase the pump station power or optimize the valve control to ensure smooth water flow. Finally, based on these analyses, the control association parameters are determined. These parameters include specific control equipment adjustment values ​​(such as pump power setting, valve opening range) and control adjustment rules related to regional characteristics, which can dynamically reflect the operation requirements of the control equipment under different regional conditions and provide data basis for precise drainage control. ;

[0039] The adaptive demand parameter is used as an index and input into an adaptive search model to search for control parameters, and dynamic control parameters are obtained. The dynamic control parameters are used to control the interception and discharge of multi-stage drainage equipment, wherein the adaptive search model is obtained based on the multi-stage drainage network and optimized spatial integrated training.

[0040] In one embodiment, the adaptive demand parameter is used as an index and input into the adaptive search model. The adaptive search model uses this index as a constraint to guide the model to search for the optimal control parameters in the complex environment of the multi-level drainage network. The adaptive search model is constructed based on the characteristics of the multi-level drainage network and is obtained through optimization space integrated training. The model contains multiple sub-models for evaluating the operating efficiency of the drainage equipment and the regulation effect of the control parameters. By searching in the optimization space, the model iteratively adjusts the control parameters (such as pump power and valve opening) to ensure that the drainage demand is accurately met, avoid insufficient or excessive drainage, optimize the operating efficiency of the equipment, and reduce unnecessary energy consumption. During the search process, the model is searched multiple times. Through simulation and optimization, each set of control parameters is dynamically adjusted to maximize the output effect of adaptive demand parameters. For example, in areas with high drainage demand, the model may give priority to increasing the power of the pump station and expanding the valve opening, while in areas with low drainage demand, it will reduce the equipment output and save resources. Through the optimization of the adaptive search model, dynamic control parameters are finally obtained. These parameters include specific equipment operation settings for each segmented control area (such as the power adjustment value of the pump station and the opening range of the valve), which can guide the real-time operation of the drainage equipment. Using these dynamic control parameters, multi-stage drainage equipment realizes precise interception and discharge control to meet the drainage needs of the segmented control area, while ensuring the efficient operation of the drainage network as a whole.

[0041] In a possible implementation, the adaptive demand parameter is used as an index and input into an adaptive search model to search for control parameters. Before obtaining the dynamic control parameter, the following steps are included:

[0042] Based on the hydrodynamic simulation model and the multi-level drainage network, a simulation model is constructed. The simulation model has a built-in integrated learning space, and the integrated learning space includes a control parameter learning space, an evaluation learning space, and an optimization learning space; wherein the evaluation learning space performs evaluation relationship learning based on the multi-level equipment drainage evaluation target, constructs multiple evaluation weak models, and configures weak model integration weights according to the model verification results of the multiple evaluation weak models; the multiple evaluation weak models are integrated and learned according to the weak model integration weights to construct an evaluation model; the adaptive demand parameters are used as input constraints, and the optimization learning space is optimized iteratively. Learning; based on the integrated learning space, spatial learning models are fused and connected to construct the adaptive search model.

[0043] Optionally, a comprehensive simulation model is constructed based on the hydrodynamic simulation model and the multi-level drainage network. The model has an embedded integrated learning space that can effectively combine the dynamic characteristics of the drainage equipment and the network operation requirements to optimize and control the drainage system. The integrated learning space consists of three subspaces, namely the control parameter learning space, the evaluation learning space, and the optimization learning space; the control parameter learning space focuses on analyzing the relationship between the control parameters (such as pump station power, valve opening) and the performance of the drainage equipment in the multi-level drainage network. Through hydrodynamic simulation, the model learns the influence of different control parameters on the drainage velocity, flow rate and pressure in this space, thereby establishing the relationship between the control parameters and the drainage. The mapping relationship between parameters provides an accurate control basis for subsequent evaluation learning and optimization learning; the core of the evaluation learning space is to conduct relationship learning through the evaluation objectives of multi-level equipment drainage, analyze the drainage efficiency of each equipment, the adaptability of control parameters and the drainage effect of the overall network. In order to achieve this goal, the evaluation learning space constructs multiple weak evaluation models, which evaluate the performance of the drainage network from different angles, such as the flow utilization rate of a certain path, the output efficiency of the pump station power or the adjustment effect of the valve opening. These models can be constructed based on neural networks, and the training process is the same as mentioned above, which is carried out through forward propagation, loss calculation, back propagation, parameter optimization and other steps. The verification results of these weak models are analyzed through test data, and weights are assigned according to the accuracy of the models, so as to perform weighted integration of the weak models and form a stronger comprehensive evaluation model; the optimization learning space takes the maximization of the output value of the evaluation model as the goal (that is, the evaluation score is greater than or equal to the preset score), and performs iterative learning of the optimal control parameters. By taking the adaptive demand parameters as input constraints and combining the feedback of the evaluation model, the control parameters in the simulation model (such as pump station power and valve opening) are continuously adjusted, and the operating configuration of the equipment is iteratively optimized to meet the drainage needs of different control areas. During the optimization process, the optimization learning space will comprehensively consider the resource utilization efficiency of the equipment operation, the balance of the drainage network, and the dynamic characteristics of the regional drainage demand, and gradually approach the optimal control. Control parameters, for example, according to the demand and pipeline flow resistance, increase or decrease the power output and opening, or redistribute the load among multiple pumping stations or valves to optimize the operation efficiency. In addition, genetic algorithms, particle swarm optimization algorithms, etc. can be used to optimize the control parameters in the optimization learning space to accelerate the convergence process of the control parameters; finally, the control parameter learning space, evaluation learning space and optimization learning space are fused and connected through the spatial learning model to construct a complete adaptive search model. The fused model can not only dynamically adapt to the operation needs of the multi-level drainage network, but also comprehensively analyze the influence of different control parameters on the drainage effect, and provide the optimal control parameter configuration for the drainage equipment operation, ensuring that the drainage equipment can meet the needs of each region while maximizing the overall drainage efficiency.

[0044] Embodiment 2, based on the same inventive concept as the dynamic control method of the intelligent interception and drainage equipment in the above embodiment, Figure 2 As shown, the present application provides a dynamic control platform for intelligent interception and drainage equipment, and the platform and method embodiments in the embodiments of the present application are based on the same inventive concept. Among them, the platform includes: network construction module 1: parse the drainage parameters of multi-level drainage equipment and build a multi-level drainage network; channel connection module 2: connect multi-source acquisition channels, and the multi-source acquisition channels include weather data acquisition channels, image data acquisition channels, and crop data acquisition channels; external discharge demand prediction module 3: based on the multi-level drainage network, the external discharge demand is predicted according to the multi-source acquisition data obtained by the multi-source acquisition channel, and the predicted demand external discharge information is obtained; index parameter extraction module 4: according to the predicted demand external discharge information, the index parameter is extracted with the preset adaptive index parameter to obtain the adaptive demand parameter; interception and drainage control module 5: using the adaptive demand parameter as the index, inputting the control parameter search into the adaptive search model, obtaining the dynamic control parameter, and using the dynamic control parameter to control the interception and drainage of multi-level drainage equipment, wherein the adaptive search model is obtained based on the multi-level drainage network and optimized space integrated training.

[0045] Furthermore, the network construction module 1 also includes:

[0046] The control settings and equipment distribution of multi-level drainage equipment are obtained, and the multi-level drainage equipment includes field drainage, secondary drainage, main drainage, water storage regulation, and external drainage emergency; the control node setting position, control drainage equipment level, control parameters, drainage cross-section, drainage intensity and direction are obtained according to the control settings, and the control parameters include valve opening or pump station power; the control node is used as the network node, and the equipment distribution of the control drainage equipment level corresponding to the control node setting position is used as the connecting edge, the drainage equipment level is used as the edge attribute, and the control parameters, drainage cross-section, drainage intensity and direction are used for edge assignment to construct the multi-level drainage network.

[0047] Furthermore, the outbound demand prediction module 3 also includes:

[0048] Based on weather data and crop data, coordinate positioning and feature time series are performed to construct weather positioning distribution, weather time series features, crop growth feature positioning distribution, and crop growth time series features; image soil waterlogging feature recognition and coordinate positioning are performed based on image data to construct a soil waterlogging feature distribution map; the soil waterlogging feature distribution map is aligned with the position positioning of the multi-level drainage network to determine the data positioning mapping relationship, and the weather positioning distribution and the crop growth feature positioning distribution are fitted to the data positioning mapping relationship; the weather time series features and crop growth time series features are used for time series alignment, the data positioning mapping relationship is constructed by time series moving window, and the data positioning mapping relationship is aligned by time series moving window relationship according to the acquisition time of the image data, wherein when the acquisition time of the image data is missing, the image data of the adjacent acquisition time or the average of the image data of the adjacent acquisition time at both ends is used to fill it; the drainage demand analysis is performed based on the data positioning mapping relationship to determine the single window drainage demand, and the single window drainage demand is predicted in time series accumulation according to the time series moving window to obtain the predicted demand external drainage information.

[0049] Furthermore, the outbound demand prediction module 3 also includes:

[0050] The acquisition time of the image data is obtained, and an entity time series moving window is determined, where the entity time series moving window is a time series moving window corresponding to the acquisition time closest to the occurrence of weather and rainfall; according to the entity time series moving window, the drainage demand of each area is calculated based on the soil water accumulation characteristics and crop growth characteristics of each area according to the segmented control area of ​​the data positioning mapping relationship, and the single window drainage demand is obtained, and the single window drainage demand is used to represent the drainage demand of each segmented control area in the entity time series node; the single window drainage demand is used as the node state, and the state change amount is fitted according to the weather time series characteristics of the time series moving window to construct a time series prediction model; the time series cumulative prediction of the time series moving window is performed through the time series prediction model, and the drainage demand of each segmented control area corresponding to each time node is obtained as the predicted demand external drainage information.

[0051] Furthermore, the index parameter extraction module 4 also includes:

[0052] According to the control parameters, the control paths of the multi-level drainage network are marked with control parameters; the control paths of the multi-level drainage network are divided and matched with the predicted demand outflow information to obtain the control parameters of the control areas; the control parameters of the control areas are mined for correlation with outflow demand parameters to determine control association parameters; the control association parameters are used as preset adaptive index parameters, and preset adaptive index parameter analysis and extraction are performed according to the divided control areas and demand of the predicted demand outflow information to obtain the adaptive demand parameters.

[0053] Furthermore, the index parameter extraction module 4 also includes:

[0054] Establish a drainage relationship between the control parameters and drainage parameters; based on the multi-level drainage network, extract the drainage path, drainage section, drainage intensity and direction of the control area, and identify the regional terrain difference according to the location of the control area; based on the hydrodynamic simulation model, explore the correlation between the regional terrain difference, drainage path, drainage section, drainage intensity and direction and the drainage parameters; transfer parameter relationships according to the correlation and drainage relationship, determine the control relationship between the control parameters and the regional terrain difference, drainage path, drainage section, drainage intensity and direction, and determine the control correlation parameters.

[0055] Furthermore, the interception and discharge control module 5 also includes:

[0056] Based on the hydrodynamic simulation model and the multi-level drainage network, a simulation model is constructed. The simulation model has a built-in integrated learning space, and the integrated learning space includes a control parameter learning space, an evaluation learning space, and an optimization learning space; wherein the evaluation learning space performs evaluation relationship learning based on the multi-level equipment drainage evaluation target, constructs multiple evaluation weak models, and configures weak model integration weights according to the model verification results of the multiple evaluation weak models; the multiple evaluation weak models are integrated and learned according to the weak model integration weights to construct an evaluation model; the adaptive demand parameters are used as input constraints, and the optimization learning space is optimized iteratively. Learning; based on the integrated learning space, spatial learning models are fused and connected to construct the adaptive search model.

[0057] It should be noted that the above-mentioned sequence of the embodiments of the present application is only for description and does not represent the advantages and disadvantages of the embodiments. And the above-mentioned specific embodiments of this specification are described. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0058] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application should be included in the protection scope of the present application.

[0059] This specification and the drawings are merely exemplary illustrations of the present application and are deemed to cover any and all modifications, variations, combinations or equivalents within the scope of the present application. Obviously, a person skilled in the art may make various modifications and variations to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalents, the present application intends to include these modifications and variations.

Claims

1. A dynamic control method for intelligent interception and drainage equipment, characterized in that: The method comprises: Analyze the drainage parameters of multi-level drainage equipment and build a multi-level drainage network; Connecting a multi-source acquisition channel, wherein the multi-source acquisition channel includes a weather data acquisition channel, an image data acquisition channel, and a crop data acquisition channel; Based on the multi-level drainage network, the drainage demand is predicted according to the multi-source collection data obtained by the multi-source collection channel to obtain the predicted demand drainage information; According to the predicted demand outflow information, extract index parameters with preset adaptive index parameters to obtain adaptive demand parameters; The adaptive demand parameter is used as an index, input into an adaptive search model to search for control parameters, obtain dynamic control parameters, and use the dynamic control parameters to control the interception and discharge of multi-stage drainage equipment, wherein the adaptive search model is obtained based on the multi-stage drainage network and optimized spatial integrated training; According to the predicted demand outflow information, extracting index parameters with preset adaptive index parameters to obtain adaptive demand parameters includes: According to the control parameters, the control paths of the multi-level drainage network are marked with control parameters; According to the control path of the multi-level drainage network and the predicted demand drainage information, the control area is divided and matched to obtain the control parameters of the control area; Performing external demand parameter correlation mining on the control parameters of the control area to determine the control correlation parameters; The control association parameter is used as a preset adaptive index parameter, and the preset adaptive index parameter is analyzed and extracted according to the segmented control area and demand of the predicted demand outflow information to obtain the adaptive demand parameter; The control parameters include valve opening or pump station power.

2. The dynamic control method of the intelligent interception and drainage equipment according to claim 1, characterized in that: The method of analyzing the drainage parameters of the multi-level drainage equipment and constructing the multi-level drainage network includes: Obtain control settings and equipment distribution of multi-stage drainage equipment, including field drainage, secondary drainage, main drainage, water storage regulation, and external drainage emergency; According to the control setting, the control node setting position, the control drainage equipment level, the control parameters, the drainage section, the drainage intensity and the direction are obtained; The control node is used as the network node, the equipment distribution of the control drainage equipment level corresponding to the control node setting position is used as the connecting edge, the drainage equipment level is used as the edge attribute, and the control parameter, drainage section, drainage intensity and direction are used to assign edges to construct the multi-level drainage network.

3. The dynamic control method of the intelligent interception and drainage equipment according to claim 2, characterized in that: Based on the multi-level drainage network, the drainage demand is predicted according to the multi-source collection data acquired by the multi-source collection channel to obtain the predicted demand drainage information, including: Based on weather data and crop data, coordinate positioning and feature time series are performed to construct weather positioning distribution, weather time series features, crop growth feature positioning distribution, and crop growth time series features; Based on the image data, the image soil waterlogging features are identified and the coordinates are located, and a soil waterlogging feature distribution map is constructed; Aligning the soil waterlogging characteristic distribution map with the position positioning of the multi-level drainage network, determining a data positioning mapping relationship, and fitting the weather positioning distribution and the crop growth characteristic positioning distribution into the data positioning mapping relationship; The weather time series characteristics and crop growth time series characteristics are used to perform time series alignment, a time series moving window is constructed for the data positioning mapping relationship, and the data positioning mapping relationship is aligned with the time series moving window relationship according to the acquisition time of the image data, wherein when the acquisition time of the image data is missing, the image data of the adjacent acquisition time or the average of the image data of the acquisition time at both ends is used to fill it; Based on the data positioning mapping relationship, a drainage demand analysis is performed to determine the single window drainage demand, and a time series cumulative prediction is performed on the single window drainage demand according to the time series moving window to obtain the predicted demand external drainage information.

4. The dynamic control method of the intelligent interception and drainage equipment according to claim 3, characterized in that: Performing drainage demand analysis based on the data positioning mapping relationship to determine the single window drainage demand, performing time series cumulative prediction on the single window drainage demand according to the time series moving window, and obtaining the predicted demand drainage information, including: Acquire the acquisition time of the image data and determine the entity time series moving window, where the entity time series moving window is the time series moving window corresponding to the acquisition time closest to the occurrence of rainfall; According to the physical time series moving window, the drainage demand is calculated based on the soil water accumulation characteristics and crop growth characteristics of each area according to the segmented control area of ​​the data positioning mapping relationship, and the single window drainage demand is obtained. The single window drainage demand is used to represent the drainage demand of each segmented control area in the physical time series node; Taking the single window drainage demand as the node state, fitting the state change amount according to the weather time series characteristics of the time series moving window, and constructing a time series prediction model; The time series prediction model is used to perform time series cumulative prediction of the time series moving window to obtain the drainage demand of each divided control area corresponding to each time node as the predicted demand external drainage information.

5. The dynamic control method of intelligent interception and drainage equipment according to claim 1, characterized in that: Mining the correlation between the control parameters of the control area and the external demand parameters to determine the control correlation parameters includes: Establishing a drainage relationship between the control parameter and the drainage parameter; The drainage parameters include water flow velocity, water pressure, and flow rate; Based on the multi-level drainage network, the drainage path, drainage section, drainage intensity and direction of the control area are extracted, and the regional terrain difference is identified according to the location of the control area; Based on the hydrodynamic simulation model, the correlation between the regional terrain difference, drainage path, drainage section, drainage intensity and direction and the drainage parameters is explored; Parameter relationship transfer is performed according to the association relationship and drainage relationship, and the control relationship between the control parameter and the regional terrain difference, drainage path, drainage cross section, drainage intensity and direction is determined to determine the control association parameter.

6. The dynamic control method of the intelligent interception and drainage equipment according to claim 2, characterized in that: The adaptive demand parameter is used as an index and input into the adaptive search model to search for control parameters. Before obtaining the dynamic control parameters, the following steps are included: Based on the hydrodynamic simulation model and the multi-level drainage network, a simulation model is constructed, wherein the simulation model has a built-in integrated learning space, and the integrated learning space includes a control parameter learning space, an evaluation learning space, and an optimization learning space; The evaluation learning space performs evaluation relationship learning based on multi-level equipment drainage evaluation targets, constructs multiple evaluation weak models, and configures weak model integration weights based on model verification results of the multiple evaluation weak models; Performing integrated learning on the multiple evaluation weak models according to the weak model integration weights to construct an evaluation model; Taking the adaptive demand parameters as input constraints and maximizing the output value of the evaluation model as the goal, performing iterative learning to optimize the learning space; The spatial learning model is fused and connected based on the integrated learning space to construct the adaptive search model.

7. Dynamic control platform of intelligent interception and drainage equipment, characterized in that: The platform is used to execute the dynamic control method of the intelligent interception and drainage equipment according to any one of claims 1 to 6, comprising: Network construction module: Analyze the drainage parameters of multi-level drainage equipment and build a multi-level drainage network; Channel connection module: connects multi-source acquisition channels, including weather data acquisition channel, image data acquisition channel, and crop data acquisition channel; An outflow demand prediction module: based on the multi-level drainage network, an outflow demand prediction is performed according to the multi-source collection data acquired by the multi-source collection channel to obtain the predicted demand outflow information; Index parameter extraction module: extracts index parameters according to the predicted demand outflow information with preset adaptive index parameters to obtain adaptive demand parameters; Interception and discharge control module: using the adaptive demand parameter as an index, inputting it into an adaptive search model to search for control parameters, obtaining dynamic control parameters, and utilizing the dynamic control parameters to perform interception and discharge control of multi-stage drainage equipment, wherein the adaptive search model is obtained based on the multi-stage drainage network and optimized spatial integrated training.

Images