Waterlogging intelligent prediction method, system and device based on drainage pipe network internet-of-things perception

Abstract

The invention provides an intelligent waterlogging prediction method, system and device based on drainage pipe network Internet of Things perception, and relates to the technical field of urban waterlogging prediction, and the method mainly comprises the steps: collecting the topographic data, land utilization data, soil data, river data, road data, building data and rainfall data of a city; obtaining the corrected terrain grid data; constructing a ground surface model; constructing a drainage pipe network generalization model; constructing a basic catchment area; dividing sub catchment areas; constructing an urban inland inundation prediction model; performing intelligent calibration on the urban inland inundation prediction model through a genetic algorithm based on drainage pipe network Internet of Things sensing data to obtain optimal parameters of the urban inland inundation prediction model; and collecting current rainfall data of the city, inputting the rainfall data into the city waterlogging prediction model, and outputting to obtain a waterlogging prediction result. According to the scheme, the urban inland inundation prediction model can be effectively optimized, the urban inland inundation prediction precision is improved, and the prediction result is more practical.

Description

Technical Field

[0001] The present invention relates to the technical field of urban waterlogging prediction, and in particular to an intelligent waterlogging prediction method, system and device based on the Internet of Things perception of drainage pipe networks. Background Art

[0002] Currently, urban waterlogging prediction methods mainly rely on basic data such as terrain data and precipitation data, and fail to use the Internet of Things perception information of real drainage pipe networks for model calibration. Moreover, conventional terrain data usually has problems such as low resolution and poor timeliness, resulting in the prediction results of the model not conforming to the actual situation and low prediction accuracy, thus generating potential urban waterlogging hazards. Summary of the Invention

[0003] The purpose of the present invention is to provide an intelligent waterlogging prediction method, system and device based on the Internet of Things perception of drainage pipe networks to solve at least one of the above technical problems existing in the prior art.

[0004] In a first aspect, to solve the above technical problems, an intelligent waterlogging prediction method based on the Internet of Things perception of drainage pipe networks provided by the present invention includes the following steps: Step 1, collect terrain data, land use data, soil data, river data, road data, building data, precipitation data, etc. of the city; the terrain data includes drainage pipe network census data and drainage pipe network Internet of Things perception data, etc.; the drainage pipe network census data includes elevation values of manually measured pipeline manholes, etc.; the drainage pipe network Internet of Things perception data includes water level values and flow values of each drainage pipe well node, etc.; the precipitation data includes precipitation intensity, etc.; In a feasible implementation manner, the drainage pipe network census data, the land use data, the soil data, the river data, the road data and the building data are obtained by accessing the public resource data server; the precipitation data is obtained by accessing the meteorological resource data server.

[0005] In a feasible implementation manner, the drainage pipe network Internet of Things perception data is obtained by setting Internet of Things-based water level gauges and flow meters at each drainage pipe well node.

[0006] Step 2, preprocess the elevation values of the manually measured pipeline manholes, screen out data noise points exceeding the elevation threshold to obtain terrain grid data, so as to ensure that the elevation values conform to the actual situation; based on the terrain grid data, perform interpolation processing through a radial basis function (RBF) to obtain corrected terrain grid data, so that the terrain grid data is more accurate.

[0007] In a feasible implementation manner, the interpolation processing includes the following steps: Step a1: Use the Gaussian kernel function as the RBF kernel function. The specific formula is: ; where represents the Euclidean distance between two elevation nodes in the terrain grid data; represents the width parameter of the kernel function, which can be set as the average value of the Euclidean distances between all elevation nodes in the terrain grid data; Step a2: Construct an interpolation matrix based on the elevation node positions and elevation values. For example, if the coordinates corresponding to the measured elevation nodes in the terrain grid data are known, the elements of the interpolation matrix can be expressed as: ; where represents the element between the th elevation node and the th elevation node in the interpolation matrix; represents the x - coordinate of the th elevation node; represents the x - coordinate of the th elevation node; represents the y - coordinate of the th elevation node; represents the y - coordinate of the th elevation node; Calculate the weight coefficients through a system of linear equations. The specific formula is: ; where represents the weight coefficient; represents the vector of known elevation values; Step a3: Taking each measured elevation node as the center, extend the grid indices by a preset number in eight directions respectively to obtain all interpolation target points. The eight directions are north, northeast, east, southeast, south, southwest, west, and northwest. Calculate the Euclidean distance between the interpolation target points and the elevation nodes. The specific formula is: ; where represents the Euclidean distance between the interpolation target point and the th elevation node; represents the x - coordinate of the interpolation target point; represents the y - coordinate of the interpolation target point; Calculate the interpolation result according to the weight coefficients and the kernel function , and update it to the terrain grid data to obtain the corrected terrain grid data. The specific formula is: ; Among them, represents the total number of measured elevation nodes.

[0008] Step 3: Use the land use data and soil data as underlying surface data to conduct surface water permeability analysis and construct a surface model; the land use data includes pervious surfaces and impervious surfaces; the pervious surfaces include land surfaces, grass surfaces, gravel surfaces, etc.; the impervious surfaces include non-depression storage impervious surfaces and depression storage impervious surfaces; the non-depression storage impervious surfaces include hardened surfaces such as building surfaces and road surfaces; the depression storage impervious surfaces include hardened surfaces with water storage structures, etc.; the soil data includes water holding capacity, permeability, hydraulic conductivity, etc.

[0009] Step 4: In the terrain grid data, according to the drainage network census data, conduct generalization processing on the pipe network and drainage pipe well nodes to construct a generalized drainage network model.

[0010] Step 5: In the corrected terrain grid data, according to the river data, road data, and building data, construct a basic catchment area according to the division criteria; use the basic catchment area as the division unit, and according to the main pipe direction of the drainage pipeline and the distribution of drainage pipe well nodes in the generalized drainage network model, divide the sub-catchment areas by the Thiessen polygon method to be more in line with the real scenario.

[0011] In a feasible implementation manner, the division criteria include: the overall river basin cannot be truncated; the main road cannot be truncated; the contiguous buildings cannot be truncated.

[0012] In a feasible implementation manner, the specific method for dividing the sub-catchment areas includes: Step b1: Obtain the positions of all generalized nodes (drainage pipe well nodes) within the basic catchment area as the generation points of the Thiessen polygon; Step b2: For any two adjacent generation points, calculate the midpoint of the line connecting the two and the slope of its perpendicular bisector. The specific formulas include: ; Among them, represents the midpoint; represents the x-coordinate of the th generation point; represents the x-coordinate of the th generation point; represents the y-coordinate of the th generation point; Among them, represents the slope of the perpendicular bisector; Among them, the equation of the perpendicular bisector is: ; Among them, represents the y - coordinate of represents the x - coordinate of Step b3: Calculate the intersection points of each perpendicular bisector with other perpendicular bisectors, connect these intersection points as the vertices of the Thiessen polygon, generate the Thiessen polygon area, and obtain several sub - catchments, so as to ensure that each sub - catchment has a set of water level gauges and flow meters.

[0013] Step 6: After integrating the surface model and the drainage network generalization model, based on the precipitation data, conduct sub - catchment runoff calculation to construct an urban waterlogging prediction model.

[0014] In a feasible implementation manner, the sub - catchment runoff calculation includes infiltration calculation, runoff generation calculation, and outflow calculation; The infiltration calculation refers to calculating the water flow data that penetrates into the pervious surface and reaches the unsaturated soil after precipitation reaches the sub - catchment. The formula includes: The infiltration volume calculation formula is specifically: ; Among them, represents the infiltration volume; represents the time step; represents the instantaneous infiltration rate, and the specific formula is: ; Among them, represents the initial infiltration rate (unit: mm / h); represents the stable infiltration rate; is the infiltration attenuation coefficient (unit: 1 / h); represents the precipitation duration (unit: s); The runoff generation calculation refers to calculating the remaining water flow data after deducting losses such as vegetation interception, evaporation, and infiltration from precipitation. The formula includes: ; Among them, represents the runoff generation amount; represents the precipitation intensity (unit: mm / h); represents the maximum drainage depth of the drainage pipe well node; represents the infiltration amount; The outflow calculation includes taking the sub - catchment as a non - linear reservoir, combining the runoff generation calculation, and calculating the outlet flow of the drainage pipe well node in the sub - catchment through the Manning equation and the continuity equation. The specific formula includes: The continuity equation of the non - linear reservoir model, and the specific formula is: ; Wherein, represents the area of the sub - catchment; represents the total water volume of the sub - catchment; represents the depression storage depth of the sub - catchment (i.e., the low - lying area on the ground, such as the pervious ground and the impervious ground with depression storage, which can store the precipitation depth); represents the outlet flow of the drain pipe well node in the sub - catchment, which can correspond to the flow value collected by the flowmeter, and the specific formula is: ; Wherein, represents the overland flow width of the sub - catchment; represents the Manning roughness coefficient; represents the surface detention depth of the sub - catchment; represents the slope of the sub - catchment; When is greater than , the precipitation turns into surface runoff; When is less than , the precipitation is stored in the depression.

[0015] Step 7: Based on the Internet of Things perception data of the drainage pipe network, use the genetic algorithm to perform intelligent calibration on the urban waterlogging prediction model to obtain the optimal parameters of the urban waterlogging prediction model.

[0016] In a feasible implementation manner, the said Step 7 includes: Step 71: Define the parameters and population size of the initial population in the genetic algorithm; the parameters include the maximum infiltration rate, the minimum infiltration rate, the Manning roughness coefficient, the infiltration attenuation coefficient, etc.; Step 72: Align the Internet of Things perception data of the drainage pipe network and the precipitation data according to the time sequence as the calibration result data; Step 73: Input the parameters of the current population into the urban waterlogging prediction model, obtain the simulation values of each population at each time sequence, and extract the Internet of Things perception data of the drainage pipe network corresponding to the time sequence as the monitoring values for calculating the objective function; the objective function is used to calculate the degree of coincidence between the simulation values and the monitoring values of the urban waterlogging prediction model; Step 74: Write the objective function values of each individual in the current population into the individual fitness vector respectively, sort them in descending order, and select the individuals in the top preset proportion as the basic data, so as to select excellent individuals as the basis; Step 75: Perform single - point crossover processing on the current population, and then perform mutation processing on the individuals in the population to obtain the offspring population; Step 76: Based on the parameters of the offspring population, iteratively execute Step 73 until the iteration end condition is reached, and obtain the optimal parameters of the urban waterlogging prediction model.

[0017] Preferably, the objective function is the Nash-Sutcliffe efficiency coefficient , and the specific formula is: ; where, represents the monitored value at the th time node; represents the simulated value at the th time node; represents the average value of all monitored values; represents the total number of monitored time nodes; In this way, by using the Internet of Things perception data of the drainage pipe network, after comparing the monitored values with the simulated values (i.e., predicted values), the parameters of the urban waterlogging prediction model can be calibrated in a traceable manner.

[0018] Preferably, the crossover probability in the single-point crossover process is 0.65; the mutation probability in the mutation process is 0.05.

[0019] Preferably, the iteration end condition is that the maximum value of the objective function tends to be stable.

[0020] Step 8: Collect the current precipitation data of the city, input it into the urban waterlogging prediction model, and output the waterlogging prediction result.

[0021] In a feasible implementation manner, the waterlogging prediction result includes the outlet flow of each drainage pipe well node, etc.

[0022] In a second aspect, based on the same inventive concept, the present application also provides an intelligent waterlogging prediction system based on Internet of Things perception of the drainage pipe network, including a data acquisition module, a data processing module, and a result generation module; The data acquisition module is used to collect topographic data, land use data, soil data, river data, road data, building data, precipitation data, etc. of the city; the topographic data includes drainage pipe network census data and Internet of Things perception data of the drainage pipe network, etc.; the drainage pipe network census data includes manually measured elevation values of pipeline manholes, etc.; the Internet of Things perception data of the drainage pipe network includes water level values and flow values of each drainage pipe well node, etc.; the precipitation data includes precipitation intensity, etc.; The data processing module includes a topographic grid unit, a surface model unit, a drainage pipe network generalization model unit, a sub-catchment unit, an urban waterlogging prediction model unit, and a prediction unit; The terrain grid unit preprocesses the elevation values of the manually measured pipeline manholes, filters out data noise points exceeding the elevation threshold, and obtains terrain grid data; based on the terrain data, interpolation processing is performed through radial basis functions to obtain corrected terrain grid data; The surface model unit uses land use data and soil data as underlying surface data to perform surface water permeability analysis and construct a surface model; the land use data includes pervious surfaces and impervious surfaces; the pervious surfaces include land surfaces, grass surfaces, gravel surfaces, etc.; the impervious surfaces include non-depression storage impervious surfaces and depression storage impervious surfaces; the non-depression storage impervious surfaces include hardened surfaces such as building surfaces and road surfaces; the depression storage impervious surfaces include hardened surfaces with water storage structures, etc.; the soil data includes water holding capacity, permeability, hydraulic conductivity, etc.; The drainage network generalization model unit performs network and drainage pipe well node generalization processing in the terrain grid data according to the drainage network census data to construct a drainage network generalization model; The sub-catchment unit constructs a basic catchment in the corrected terrain grid data according to river data, road data, and building data according to the division criteria; taking the basic catchment as the division unit, according to the main pipe direction of the drainage pipeline and the distribution of drainage pipe well nodes in the drainage network generalization model, the sub-catchments are divided by the Thiessen polygon method; The urban waterlogging prediction model unit integrates the surface model and the drainage network generalization model, and based on precipitation data, performs sub-catchment runoff calculation to construct an urban waterlogging prediction model; based on the drainage network Internet of Things perception data, the genetic algorithm is used to perform intelligent calibration on the urban waterlogging prediction model to obtain the optimal parameters of the urban waterlogging prediction model; The prediction unit is used to input the current precipitation data of the city into the urban waterlogging prediction model and output the waterlogging prediction result; The result generation module is used to send out the waterlogging prediction result.

[0023] In a third aspect, based on the same inventive concept, the present application also provides an intelligent waterlogging prediction device based on drainage network Internet of Things perception, including a processor, a memory, and a bus. The memory stores instructions and data read by the processor. The processor is used to call the instructions and data in the memory to execute the above-mentioned intelligent waterlogging prediction method based on drainage network Internet of Things perception. The bus is connected between each functional component for transmitting information.

[0024] In a feasible implementation manner, the device further includes water level gauges and flow meters arranged at each drainage pipe well node, which are respectively used to obtain water level values and flow values.

[0025] Adopting the above technical solutions, the present invention has the following beneficial effects: An intelligent prediction method, system and device for urban waterlogging based on the Internet of Things perception of drainage pipe networks provided by the present invention combines the general survey data of drainage pipe networks, uses radial basis function (RBF) interpolation processing to correct the data of conventional terrain areas, and improves the accuracy and freshness of terrain data in key facility areas such as drainage pipe networks; taking the Internet of Things perception data of drainage pipe networks and historical precipitation data as input data, the intelligent calibration of the parameters of the urban waterlogging prediction model is carried out through the genetic algorithm, which improves the accuracy of urban waterlogging prediction and makes the prediction results more in line with the actual situation. Description of the Drawings

[0026] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following will briefly introduce the drawings required for the description of the specific embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0027] Figure 1 It is a flow chart of an intelligent prediction method for urban waterlogging based on the Internet of Things perception of drainage pipe networks provided by an embodiment of the present invention; Figure 2 It is a schematic diagram of the construction and optimization principle of the urban waterlogging prediction model provided by an embodiment of the present invention; Figure 3 It is a flow chart of the interpolation processing method provided by an embodiment of the present invention; Figure 4 It is a flow chart of the method for dividing sub-catchments provided by an embodiment of the present invention; Figure 5 For Figure 1 It is the specific flow chart of step 7 in Figure 6 It is a system diagram of an intelligent prediction method for urban waterlogging based on the Internet of Things perception of drainage pipe networks provided by an embodiment of the present invention. Detailed Embodiments

[0028] The following will clearly and completely describe the technical solutions of the present invention with reference to the drawings. Obviously, the described embodiments are some embodiments of the present invention, rather than all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts fall within the protection scope of the present invention.

[0029] In the description of the present invention, it should be noted that the orientation or positional relationship indicated by the terms "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer", etc. is based on the orientation or positional relationship shown in the drawings. It is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be construed as a limitation to the present invention. In addition, the terms "first", "second", "third" are only used for descriptive purposes and cannot be construed as indicating or implying relative importance.

[0030] In the description of the present invention, it should be noted that unless otherwise clearly specified and defined, the terms "installed", "connected", "connected" should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be directly connected or indirectly connected through an intermediate medium, and it can be the communication inside two elements. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific situations.

[0031] The following further explains and illustrates the present invention in combination with specific embodiments.

[0032] It should also be noted that the following specific embodiments or specific implementation manners are a series of optimized setting manners listed by the present invention to further explain the specific invention content, and these setting manners can be combined with each other or used in association with each other.

[0033] Embodiment 1: As Figure 1-2 shown, an intelligent prediction method for urban waterlogging based on the Internet of Things perception of drainage networks provided in this embodiment includes the following steps: Step 1, collect topographic data, land use data, soil data, river data, road data, building data, precipitation data, etc. of the city; the topographic data includes drainage network census data and drainage network Internet of Things perception data, etc.; the drainage network census data includes elevation values of manually measured pipeline manholes, etc.; the drainage network Internet of Things perception data includes water level values and flow values of each drainage pipeline manhole node, etc.; the precipitation data includes precipitation intensity, etc.

[0034] Furthermore, the drainage network census data, the land use data, the soil data, the river data, the road data, and the building data are obtained by accessing the public resource data server (such as the official website servers of the natural resources planning department, the municipal department, etc.); the precipitation data is obtained by accessing the meteorological resource data server (such as the meteorological observatory server).

[0035] Furthermore, the IoT perception data of the drainage pipe network is obtained by setting IoT-based water level gauges and flow meters at each drainage pipe well node.

[0036] Step 2: In the geographic information software, preprocess the elevation values of the manually measured pipeline wells, screen out the data noise points exceeding the elevation threshold, and obtain the terrain grid data to ensure that the elevation values conform to the actual situation; based on the terrain grid data, perform interpolation processing through the radial basis function (RBF) to obtain the corrected terrain grid data, thereby making the terrain grid data more accurate.

[0037] Furthermore, as Figure 3 shown, the interpolation processing includes the following steps: Step a1: Take the Gaussian kernel function as the RBF kernel function, and the specific formula is: ; where represents the Euclidean distance between two elevation nodes in the terrain grid data; represents the width parameter of the kernel function, which can be set as the average value of the Euclidean distances between all elevation nodes in the terrain grid data; Step a2: According to the elevation node positions and elevation values, construct an interpolation matrix; For example, given the coordinates corresponding to the measured elevation nodes in the terrain grid data, the elements of the interpolation matrix can be expressed as: ; where represents the element in the interpolation matrix between the th elevation node and the th elevation node; represents the x-coordinate of the th elevation node; represents the x-coordinate of the th elevation node; represents the y-coordinate of the th elevation node; represents the y-coordinate of the th elevation node; Calculate the weight coefficients through a system of linear equations, and the specific formula is: ; where represents the weight coefficient; represents the vector of known elevation values; Step a3: Taking each measured elevation node as the center, expand the grid indices of a preset number (e.g., 40) in eight directions respectively to obtain all interpolation target points; the eight directions are north, northeast, east, southeast, south, southwest, west, and northwest; calculate the Euclidean distance between the interpolation target point and the elevation node, and the specific formula is: ; where, represents the Euclidean distance between the interpolation target point and the th elevation node; represents the x - coordinate of the interpolation target point; represents the y - coordinate of the interpolation target point; Calculate the interpolation result according to the weight coefficient and the kernel function , update it to the terrain grid data to obtain the corrected terrain grid data, and the specific formula is: ; where, represents the total number of measured elevation nodes.

[0038] Step 3: Taking the land use data and soil data as the underlying surface data, conduct surface water permeability analysis and construct a surface model; the land use data includes permeable surfaces and impermeable surfaces; the permeable surfaces include land surfaces, grass surfaces, gravel surfaces, etc.; the impermeable surfaces include non - depression storage impermeable surfaces and depression storage impermeable surfaces; the non - depression storage impermeable surfaces include hardened surfaces such as building surfaces and road surfaces; the depression storage impermeable surfaces include hardened surfaces with water storage structures, etc.; the soil data includes water - holding capacity, permeability, hydraulic conductivity, etc.

[0039] Step 4: In the terrain grid data, according to the drainage network census data, conduct generalization processing of the pipe network and drainage pipe well nodes to construct a generalized drainage network model.

[0040] Step 5: In the corrected terrain grid data, according to the river data, road data, and building data, construct a basic catchment area according to the division criteria; taking the basic catchment area as the division unit, according to the main pipe direction of the drainage pipeline and the distribution of drainage pipe well nodes in the generalized drainage network model, divide the sub - catchment areas by the Thiessen polygon method to be more in line with the real scenario.

[0041] Furthermore, the division criteria include: the overall river basin cannot be truncated; the main roads cannot be truncated; the contiguous buildings cannot be truncated.

[0042] Furthermore, as Figure 4 shown, the specific method for dividing the sub - catchment areas includes: Step b1: Obtain the positions of all generalized nodes (drain pipe well nodes) within the basic catchment area as the generation points for Thiessen polygons. Step b2: For any two adjacent generation points, calculate the midpoint of the line connecting them and the slope of its perpendicular bisector, thereby forming several triangles with perpendicular bisectors on their sides. The specific formulas are as follows: ; Among them, represents the midpoint; represents the x - coordinate of the th generation point; represents the x - coordinate of the th generation point; represents the y - coordinate of the ; Among them, represents the slope of the perpendicular bisector; Among them, the equation of the perpendicular bisector is: ; Among them, represents the y - coordinate of ; represents the x - coordinate of Step b3: Calculate the intersection points of each perpendicular bisector with other perpendicular bisectors (within each triangle), and connect these intersection points (i.e., the circumcenters of each triangle) as the vertices of the Thiessen polygon to generate a Thiessen polygon area, obtaining several sub - catchment areas, so as to ensure that each sub - catchment area has a set of water level gauges and flow meters.

[0043] Furthermore, in a city built against mountains, due to the often large slopes of the terrain, which significantly affect the water flow direction, therefore, multiply the coordinates of the circumcenter by the adjustment coefficient for proportional offset of the circumcenter coordinates towards the low - lying areas; The adjustment coefficient is obtained by calculating the elevation weights of the two vertices with the highest elevation values compared to the vertex with the lowest elevation value among the three vertices (i.e., the generation points) of the triangle to which the circumcenter belongs, and then taking the average. The specific formulas are as follows: ; Among them, represents the Euclidean distance between the th generation point and the th generation point; Indicates the elevation difference between the th generation point and the th generation point;

[0044] Step 6. After integrating the surface model and the drainage network generalization model, based on precipitation data, conduct sub-catchment runoff calculation to construct an urban waterlogging prediction model.

[0045] Furthermore, the sub-catchment runoff calculation includes infiltration calculation, runoff generation calculation, and outflow calculation; The infiltration calculation refers to calculating the water flow data that infiltrates into the permeable surface and reaches the unsaturated soil after precipitation reaches the sub-catchment. The formula includes: The infiltration volume calculation formula is specifically: ; where represents the infiltration volume; represents the time step; represents the instantaneous infiltration rate, and the specific formula is: ; where represents the initial infiltration rate (unit: mm / h); represents the stable infiltration rate; is the infiltration attenuation coefficient (unit: 1 / h); represents the precipitation duration (unit: s); The runoff generation calculation refers to calculating the remaining water flow data after deducting losses such as vegetation interception, evaporation, and infiltration from precipitation. The formula includes: ; where represents the runoff generation volume; represents the precipitation intensity (unit: mm / h); represents the maximum drainage depth of the drainage pipe well node; represents the infiltration volume; The outflow calculation includes taking the sub-catchment as a non-linear reservoir, combining the runoff generation calculation, and calculating the outlet flow of the drainage pipe well node in the sub-catchment through the Manning equation and the continuity equation. The specific formula includes: The continuity equation of the non-linear reservoir model is specifically: ; Among them, represents the area of the sub-catchment; represents the total water volume of the sub-catchment; represents the depth of depression storage in the sub-catchment (i.e., the low-lying area on the ground, such as the pervious ground and the impervious ground with depression storage, which can store the precipitation depth); represents the outlet flow of the drainage pipe well node in the sub-catchment, corresponding to the flow value collected by the flow meter. The specific formula is: ; Among them, represents the overland flow width of the sub-catchment; represents the Manning roughness coefficient; represents the surface detention depth of the sub-catchment; represents the slope of the sub-catchment; When is greater than , the precipitation turns into surface runoff; When is less than , the precipitation is stored in the depression.

[0046] Step 7: Based on the Internet of Things perception data of the drainage network, use the genetic algorithm to intelligently calibrate the urban waterlogging prediction model to obtain the optimal parameters of the urban waterlogging prediction model.

[0047] Furthermore, as Figure 5 shown, the said Step 7 includes: Step 71: Define the parameters and population size of the initial population in the genetic algorithm; the parameters include the maximum infiltration rate, the minimum infiltration rate, the Manning roughness coefficient, and the infiltration attenuation coefficient, etc.; the population size can be 20; Step 72: Align the Internet of Things perception data of the drainage network and the precipitation data in chronological order as the calibration result data; Step 73: Input the parameters of the current population into the urban waterlogging prediction model, obtain the simulation values of each population in chronological order, and extract the Internet of Things perception data of the drainage network corresponding to the chronological order as the monitoring values for calculating the objective function; the objective function is used to calculate the degree of coincidence between the simulation values and the monitoring values of the urban waterlogging prediction model; Step 74: Write the objective function values of each individual in the current population into the individual fitness vector respectively, sort them in descending order, and select the individuals in the front preset proportion (such as 50%) as the basic data, so as to select excellent individuals as the basis; Step 75: Perform single-point crossover processing on the current population, and then perform mutation processing on the individuals in the population to obtain the offspring population; Step 76: Based on the parameters of the offspring population, iteratively execute Step 73 until the iteration end condition is reached, and obtain the optimal parameters of the urban waterlogging prediction model.

[0048] Preferably, the objective function is the Nash-Sutcliffe efficiency coefficient , and the specific formula is: ; Among them, represents the monitoring value at the th time node; represents the simulated value at the th time node; represents the average value of all monitoring values; represents the total number of time nodes for monitoring.

[0049] Preferably, the crossover probability in the single-point crossover process is 0.65; the mutation probability in the mutation process is 0.05.

[0050] Preferably, the iteration end condition is that the maximum value of the objective function tends to be stable.

[0051] Furthermore, the Internet of Things perception data of the drainage pipe network is the data after screening out faulty sensors; the specific method for screening out faulty sensors includes: Step c1: According to the time series, perform data cleaning and normalization processing on the historical data collected by the sensors at each drainage pipe well node to remove noise data for subsequent processing; based on the connection relationship between the drainage pipe well nodes in the drainage pipe network generalization model, determine the upstream sensor and downstream sensor of the target sensor; Step c2: Construct a time series data set for each target sensor, including: using the data of this target sensor, the upstream sensor data, and the downstream sensor data as features; determining the input time step (for example, the past 24 hours) and the output time step (for example, predicting the next 1 hour); converting the data into a three-dimensional array, including the number of samples, the time step, and the number of features; Step c3: Construct an LSTM model, including an input layer, an LSTM layer, a Dropout layer, and an output layer set in sequence; The input layer is used to receive the three-dimensional array; The LSTM layer is used to configure the number of LSTM units; The Dropout layer is used to prevent overfitting; The output layer is used to output the predicted value; Step c4: Divide the time series data set into a training set, a validation set, and a test set in chronological order to avoid data leakage caused by random splitting; set the batch size and the number of training epochs; train the LSTM model using the training set and the validation set; calculate the RMSE metric and the MAE metric of the LSTM model based on the test set and optimize the model parameters; Step c5: Input the current data collected by the target sensor , the upstream sensor , and the downstream sensor into the LSTM model to obtain predicted data; calculate the residual between the current data of the target sensor and the predicted data; then calculate the mean value and the standard deviation of the residuals within the input time step range; then calculate the Pearson correlation coefficient between the current data of the target sensor and the sum of the current data of the upstream sensor and the downstream sensor and respectively, and then make a judgment: If exceeds the residual alarm threshold, and and are both less than the correlation coefficient alarm threshold, then determine that the target sensor is a faulty sensor.

[0052] In this way, faulty sensors can be simply and effectively screened out, eliminating potential data distortion hazards caused by equipment reasons such as sensor failure and false alarms.

[0053] Furthermore, the residual alarm threshold is ; the correlation coefficient alarm threshold is 0.5.

[0054] Step 8: Collect the current precipitation data of the city, input it into the urban waterlogging prediction model, and output the waterlogging prediction result.

[0055] Furthermore, the waterlogging prediction result includes the outlet flow of each drainage pipe well node, etc.

[0056] Embodiment 2: As Figure 6 shown, this embodiment provides an intelligent waterlogging prediction system based on the Internet of Things perception of the drainage pipe network, including a data collection module, a data processing module, and a result generation module; The data acquisition module is used to collect topographic data, land use data, soil data, river data, road data, building data, precipitation data, etc. of the city; the topographic data includes drainage network census data and drainage network Internet of Things perception data, etc.; the drainage network census data includes manually measured elevation values of pipeline wells, etc.; the drainage network Internet of Things perception data includes water level values and flow values of each drainage pipe well node, etc.; the precipitation data includes precipitation intensity, etc. The data processing module includes a topographic grid unit, a surface model unit, a drainage network generalization model unit, a sub-catchment unit, an urban waterlogging prediction model unit, and a prediction unit; The topographic grid unit preprocesses the manually measured elevation values of pipeline wells, screens out data noise points exceeding the elevation threshold, and obtains topographic grid data; based on the topographic data, interpolation processing is performed through a radial basis function to obtain corrected topographic grid data; The surface model unit uses land use data and soil data as underlying surface data to perform surface water permeability analysis and construct a surface model; the land use data includes permeable surfaces and impermeable surfaces; the permeable surfaces include land surfaces, grass surfaces, gravel surfaces, etc.; the impermeable surfaces include non-pit storage impermeable surfaces and pit storage impermeable surfaces; the non-pit storage impermeable surfaces include hardened surfaces such as building surfaces and road surfaces; the pit storage impermeable surfaces include hardened surfaces with water storage structures, etc.; the soil data includes water holding capacity, permeability, hydraulic conductivity, etc. The drainage network generalization model unit performs network and drainage pipe well node generalization processing in the topographic grid data according to the drainage network census data to construct a drainage network generalization model; The sub-catchment unit constructs a basic catchment in the corrected topographic grid data according to river data, road data, and building data according to the division criteria; taking the basic catchment as the division unit, according to the main pipe direction of the drainage pipeline and the distribution of drainage pipe well nodes in the drainage network generalization model, the sub-catchment is divided by the Thiessen polygon method; The urban waterlogging prediction model unit integrates the surface model and the drainage network generalization model, and based on the precipitation data, performs sub-catchment runoff calculation to construct an urban waterlogging prediction model; based on the drainage network Internet of Things perception data, the urban waterlogging prediction model is intelligently calibrated through a genetic algorithm to obtain the optimal parameters of the urban waterlogging prediction model; The prediction unit is used to input the current drainage network Internet of Things perception data and precipitation data of the city into the urban waterlogging prediction model and output the waterlogging prediction result; The result generation module is used to send out the waterlogging prediction result.

[0057] Embodiment III: This embodiment provides an intelligent prediction device for waterlogging based on the Internet of Things perception of drainage pipe networks, including a processor, a memory, and a bus. The memory stores instructions and data read by the processor, and the processor is used to call the instructions and data in the memory to execute the intelligent prediction method for waterlogging based on the Internet of Things perception of drainage pipe networks as described above. The bus is connected between each functional component for transmitting information.

[0058] Further, the device further includes a water level gauge and a flow meter disposed at each drainage pipe well node, which are respectively used to obtain the water level value and the flow value.

[0059] In another implementation manner of this solution, it can also be implemented in the form of an integrated device. The device may include corresponding modules that execute each or several steps in the above various embodiments. The module may be one or more hardware modules specifically configured to execute the corresponding steps, or implemented by a processor configured to execute the corresponding steps, or stored in a computer-readable medium for implementation by the processor, or implemented through a certain combination.

[0060] The processor executes the various methods and processes described above. For example, the method implementation in this solution can be implemented as a software program, which is tangibly included in a machine-readable medium, such as a memory. In some embodiments, part or all of the software program can be loaded and / or installed via the memory and / or the communication interface. When the software program is loaded into the memory and executed by the processor, one or more steps in the method described above can be executed. Alternatively, in other embodiments, the processor can be configured to execute one of the above methods in any other appropriate manner (for example, by means of firmware).

[0061] The device can be implemented using a bus architecture. The bus architecture can include any number of interconnected buses and bridges, depending on the specific application of the hardware and the overall design constraints. The bus connects various circuits including one or more processors, memories, and / or hardware modules together. The bus can also connect various other circuits such as peripheral devices, voltage regulators, power management circuits, external antennas, etc.

[0062] The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Component (EISA) bus, etc. The bus can be divided into an address bus, a data bus, a control bus, etc.

[0063] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements on some or all of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A method for intelligent prediction of urban flooding based on drainage network IoT perception, characterized in that: include: Step 1: Collect the city's terrain data, land use data, soil data, river data, road data, building data and precipitation data; the terrain data includes drainage network survey data and drainage network IoT sensing data; the drainage network survey data includes manually measured pipeline well elevation values; the drainage network IoT sensing data includes water level values ​​and flow values ​​of each drainage well node; the precipitation data includes precipitation intensity; Step 2: pre-processing the manually measured pipeline well elevation values, filtering out data noise points exceeding the elevation threshold, and obtaining terrain grid data; interpolating the terrain grid data through radial basis functions to obtain corrected terrain grid data; Step 3: Use land use data and soil data as underlying surface data to analyze surface permeability and construct a surface model; the land use data includes permeable surface and impermeable surface; the impermeable surface includes impermeable surface without depressions and impermeable surface with depressions; the soil data includes water holding capacity, permeability and hydraulic conductivity; Step 4: In the terrain grid data, according to the drainage network census data, generalize the network and drainage well nodes to build a drainage network generalization model; Step 5: In the corrected terrain grid data, based on the river data, road data and building data, according to the division criteria, construct the basic watershed area; Taking the basic catchment area as the division unit, dividing the sub-catchment area by the Thiessen polygon method according to the direction of the main drainage pipeline and the node distribution of the drainage well in the generalized model of the drainage network; Step 6: After integrating the surface model and the drainage network generalization model, the sub-catchment flow calculation is performed based on precipitation data to build an urban waterlogging prediction model; Step 7: Based on the IoT sensing data of the drainage network, the urban waterlogging prediction model is intelligently calibrated through a genetic algorithm to obtain the optimal parameters of the urban waterlogging prediction model; Step 8: Collect the current precipitation data of the city, input it into the urban waterlogging prediction model, and output the waterlogging prediction result.

2. The method according to claim 1, characterized in that The interpolation process includes: Step a1: Use the Gaussian kernel function as the RBF kernel function. The specific formula is: ； in, Represents the Euclidean distance between two elevation nodes in terrain grid data; represents the width parameter of the kernel function, which is set to the average value of the Euclidean distance between all elevation nodes in the terrain grid data; Step a2: construct an interpolation matrix according to the elevation node positions and elevation values; If the coordinates of the measured elevation nodes in the terrain grid data are known, the elements of the interpolation matrix can be expressed as: ； in, Indicates the interpolation matrix at The elevation node and Elements between elevation nodes; Indicates The x-coordinate of the elevation node; Indicates The x-coordinate of the elevation node; Indicates The y coordinate of the elevation node; Indicates The y coordinate of the elevation node; The weight coefficient is calculated through the linear equations. The specific formula is: ； in, represents the weight coefficient; A vector representing known elevation values; Step a3: Taking each measured elevation node as the center, extend a preset number of grid indexes from eight directions to obtain all interpolation target points; the eight directions are north, northeast, east, southeast, south, southwest, west and northwest; calculate the Euclidean distance between the interpolation target point and the elevation node, and the specific formula is: ； in, Represents the interpolation target point and the The Euclidean distance between elevation nodes; Indicates the x-coordinate of the interpolation target point; Indicates the y-coordinate of the interpolation target point; Calculate the interpolation result according to the weight coefficient and kernel function , update to the terrain grid data to obtain the corrected terrain grid data. The specific formula is: ； in, Indicates the total number of measured elevation nodes.

3. The method according to claim 1, characterized in that The division criteria include: the entire river basin cannot be cut off; the main road cannot be cut off; and contiguous buildings cannot be cut off.

4. The method according to claim 1, characterized in that: The specific methods for dividing subcatchments include: Step b1, obtaining all generalized node positions in the basic catchment area as the generating points of Thiessen polygons; Step b2: For any two adjacent generating points, calculate the midpoint of the line connecting the two points and the slope of their perpendicular bisectors. The specific formula includes: ； in, Indicates the midpoint; Indicates The x-coordinate of the generated point; Indicates The x-coordinate of the generated point; Indicates The y coordinate of the generated point; Indicates The y coordinate of the generated point; ； in, represents the slope of the perpendicular bisector; The equation of the perpendicular bisector is: ； in, express The y-coordinate of express The x-coordinate of Step b3: Calculate the intersection points of each perpendicular bisector with other perpendicular bisectors, connect these intersection points as vertices of Thiessen polygons, generate Thiessen polygon areas, and obtain several sub-catchment areas.

5. The method according to claim 1, characterized in that Subcatchment runoff calculation includes infiltration calculation, runoff calculation and outflow calculation; The infiltration calculation refers to the calculation of the water flow data after the precipitation reaches the sub-catchment area, penetrates into the permeable surface, and reaches the unsaturated soil. The formula includes: The calculation formula of infiltration volume is as follows: ； in, represents the infiltration volume; represents the time step; It represents the instantaneous infiltration rate. The specific formula is: ； in, represents the initial infiltration rate; represents the steady infiltration rate; is the infiltration attenuation coefficient; Indicates the duration of precipitation; The runoff calculation refers to the calculation of the remaining water flow data after deducting the infiltration loss from the precipitation. The formula includes: ； in, Indicates flow production; Indicates precipitation intensity; Indicates the maximum drainage depth of the drainage pipe well node; Indicates the amount of infiltration; The outflow calculation includes treating the subcatchment as a nonlinear reservoir, combining the flow calculation, and calculating the outlet flow of the drainage pipe well node in the subcatchment through the Manning equation and the continuity equation. The specific formula includes: The continuity equation of the nonlinear reservoir model is as follows: ; in, represents the subcatchment area; represents the total water volume of the subcatchment; It represents the depression storage depth of the subcatchment; It represents the outlet flow of the drainage pipe well node in the subcatchment area, corresponding to the flow value collected by the flow meter. The specific formula is: ； in, represents the overland flow width of the subcatchment; represents the Manning roughness coefficient; It represents the surface water storage depth of the subcatchment; represents the slope of the subcatchment; when Greater than When precipitation turns into surface runoff; when Less than When precipitation falls, it is stored in depressions.

6. The method according to claim 1, characterized in that The step 7 comprises: Step 71, defining the parameters and population size of the initial population in the genetic algorithm; the parameters include the maximum infiltration rate, the minimum infiltration rate, the Manning roughness coefficient and the infiltration attenuation coefficient; Step 72: Align the drainage network IoT sensing data and precipitation data in time sequence as calibration result data; Step 73: Input the parameters of the current population into the urban waterlogging prediction model, obtain the simulated value of each population in the time series, and extract the IoT sensing data of the drainage network corresponding to the time series as the monitoring value for calculating the objective function; the objective function is used to calculate the degree of agreement between the simulated value and the monitoring value of the urban waterlogging prediction model; Step 74, write the objective function value of each individual in the current population into the individual fitness vector respectively, sort them in descending order, and select the individuals with a preset proportion as the basic data; Step 75: Perform single-point crossover processing on the current population, and then perform mutation processing on the individuals in the population to obtain a progeny population; Step 76: Based on the parameters of the offspring population, iteratively execute step 73 until the iteration end condition is reached to obtain the optimal parameters of the urban waterlogging prediction model.

7. The method according to claim 6, characterized in that The objective function is the Nash-Sutcliffe efficiency coefficient , the specific formula is: ； in, Indicates Monitoring value of each time node; Indicates Simulated value of each time node; Indicates the average value of all monitoring values; Indicates the total number of monitored time nodes.

8. The method according to claim 6, characterized in that The crossover probability in the single-point crossover process is 0.65; the mutation probability in the mutation process is 0.

05.

9. An intelligent waterlogging prediction system based on drainage network IoT perception, characterized in that: It includes a data acquisition module, a data processing module and a result generation module; The data acquisition module is used to collect the city's terrain data, land use data, soil data, river data, road data, building data and precipitation data; the terrain data includes drainage network census data and drainage network IoT sensing data; the drainage network census data includes manually measured pipeline well elevation values; the drainage network IoT sensing data includes water level values ​​and flow values ​​of each drainage well node; the precipitation data includes precipitation intensity; The data processing module includes a terrain grid unit, a surface model unit, a drainage network generalization model unit, a sub-catchment area unit, an urban waterlogging prediction model unit and a prediction unit; The terrain grid unit pre-processes the manually measured pipeline well elevation value, screens out data noise points exceeding the elevation threshold, and obtains terrain grid data; based on the terrain data, performs interpolation processing through radial basis function to obtain corrected terrain grid data; The surface model unit uses land use data and soil data as underlying surface data to perform surface permeability analysis and construct a surface model; the land use data includes permeable surface and impermeable surface; the impermeable surface includes impermeable surface without depressions and impermeable surface with depressions; the soil data includes water holding capacity, permeability, and hydraulic conductivity; The drainage network generalization model unit performs generalization processing of the network and the drainage pipe well nodes in the terrain grid data according to the drainage network census data, and constructs a drainage network generalization model; The sub-catchment unit constructs a basic catchment area in the modified terrain grid data according to the river data, road data and building data and the division criteria; Taking the basic catchment area as the division unit, dividing the sub-catchment area by the Thiessen polygon method according to the direction of the main drainage pipeline and the node distribution of the drainage well in the generalized model of the drainage network; The urban waterlogging prediction model unit integrates the surface model and the drainage network generalization model, performs sub-catchment flow calculation based on precipitation data, and constructs an urban waterlogging prediction model; based on the drainage network IoT sensing data, the urban waterlogging prediction model is intelligently calibrated through a genetic algorithm to obtain the optimal parameters of the urban waterlogging prediction model; The prediction unit is used to input the current precipitation data of the city into the urban waterlogging prediction model and output the waterlogging prediction result; The result generation module is used to send out the waterlogging prediction results.

10. An intelligent waterlogging prediction device based on drainage network IoT perception, characterized in that: It includes a processor, a memory and a bus, wherein the memory stores instructions and data read by the processor, the processor is used to call the instructions and data in the memory to execute any method as claimed in claim 1-8, and the bus connects the functional components for transmitting information.

Images