CN116739385B

CN116739385B - Urban traffic evaluation method and device based on spatial superposition analysis

Info

Publication number: CN116739385B
Application number: CN202310996262.7A
Authority: CN
Inventors: 董义阳; 刘志武; 梁犁丽; 徐志; 吕振豫; 杨恒; 翟然; 王鹏翔; 刘琨; 殷兆凯
Original assignee: Beijing Gezhouba Electric Power Rest House; China Three Gorges Corp
Current assignee: Beijing Gezhouba Electric Power Rest House; China Three Gorges Corp
Priority date: 2023-08-09
Filing date: 2023-08-09
Publication date: 2024-01-30
Anticipated expiration: 2043-08-09
Also published as: CN116739385A

Abstract

The invention relates to the technical field of urban flood disasters and road traffic safety, and discloses a method and a device for evaluating urban traffic based on space superposition analysis, wherein the method comprises the steps of performing space data superposition on site monitoring rainfall data, urban surface data and urban underground pipeline data to generate urban grid-by-grid space superposition data, training a preset deep learning model, and generating an urban flood ponding depth prediction model; acquiring weather forecast data in a target urban area, and determining a grid-by-grid ponding depth prediction result by utilizing an urban flood ponding depth prediction model; collecting urban population distribution data in the current time period, inputting the urban population distribution data into an urban traffic flow prediction model, and generating a grid-by-grid traffic flow prediction result; and carrying out space superposition analysis on the grid-by-grid ponding depth prediction result and the grid-by-grid traffic flow prediction result, and evaluating urban traffic in the urban grid. The method improves the prediction precision of flood disasters so as to realize accurate evaluation of urban traffic.

Description

Urban traffic evaluation method and device based on spatial superposition analysis

Technical Field

The invention relates to the technical field of urban flood disasters and road traffic safety, in particular to an urban traffic evaluation method and device based on space superposition analysis.

Background

The sudden, occurrence time and range uncertainty of flood disasters are influenced by factors such as urban ground surfaces, terrains, underground and rainfall types, so that flood disasters are prevented and controlled with great difficulty. The traditional urban storm ponding prediction method is mainly used for short prediction of areas where urban waterlogging disasters possibly occur based on real-time information of water level monitoring stations in cities. The correlation between factors limited by factors of influence or factors of influence is difficult to determine, so that flood disaster prediction is not accurate enough. Therefore, how to improve the prediction precision of flood disasters so as to realize accurate evaluation of urban traffic becomes a technical problem to be solved urgently.

Disclosure of Invention

In view of the above, the invention provides a space superposition analysis-based urban traffic evaluation method and a space superposition analysis-based urban traffic evaluation device, which aim to solve the technical problem of how to improve the prediction precision of flood disasters so as to realize accurate evaluation of urban traffic.

In a first aspect, the present invention provides a method for evaluating urban traffic based on spatial superposition analysis, including: acquiring site monitoring rainfall data, urban surface data and urban underground pipeline data in a target urban area, and performing spatial data superposition on the site monitoring rainfall data, the urban surface data and the urban underground pipeline data in the target urban area to generate urban grid-by-grid spatial superposition data; training a preset deep learning model based on the grid-by-grid space superposition data of the cities to generate a city flood ponding depth prediction model; acquiring weather forecast data in a target urban area, and determining grid-by-grid ponding depth prediction results by using a city flood ponding depth prediction model based on the weather forecast data; acquiring urban population distribution data in a current time period, inputting the urban population distribution data in the current time period into an urban traffic flow prediction model, and generating a grid-by-grid traffic flow prediction result; and carrying out space superposition analysis on the grid-by-grid ponding depth prediction result and the grid-by-grid traffic flow prediction result, and evaluating urban traffic in the urban grid based on the space superposition analysis result to generate an urban traffic evaluation result.

According to the urban traffic evaluation method based on the space superposition analysis, based on the space superposition analysis thought and the urban space region discretization thought, all factors such as sky, earth surface, underground and the like which influence the urban flood ponding depth are superposed on the same urban grid, and urban grid-by-grid space superposition data are generated; training based on grid-by-grid space superposition data of the generated city to generate a city flood ponding depth prediction model, further based on weather forecast data, determining a grid-by-grid ponding depth prediction result by using the city flood ponding depth prediction model, and combining the weather forecast data to realize prediction of ponding depths of different areas of the city and improve flood disaster prediction precision; and finally, generating grid-by-grid traffic flow prediction results through population distribution data and the urban traffic flow prediction model to predict traffic flow spatial distribution of different time nodes in the future of urban traffic, and carrying out spatial superposition analysis on the urban grid-by-grid ponding depth prediction results so as to realize accurate evaluation of urban traffic.

In an alternative embodiment, spatial data superposition is performed on site monitoring rainfall data, urban surface data and urban underground pipeline data in a target urban area to generate urban grid-by-grid spatial superposition data, and the method comprises the following steps: performing space discretization on the target city area to generate a plurality of city grids; according to the space topological relation, site monitoring rainfall data, urban surface data and urban underground pipeline data are divided into a plurality of urban grids respectively, and urban grid-by-grid space superposition data are generated.

According to the urban traffic evaluation method based on the space superposition analysis, based on the space superposition analysis thought and the urban space region discretization thought, all factors such as sky, earth surface, underground and the like which influence the urban flood ponding depth are superposed on the same grid, and each unit grid represents the regional position, regional attribute and precipitation information of the city so as to accurately predict the specific region where the urban flood disaster occurs and comprehensively and finely predict the flood disaster distribution conditions of different regions of the city.

In an optional implementation manner, training the preset deep learning model based on the grid-by-grid space superposition data of the city to generate a city flood ponding depth prediction model comprises the following steps: inputting the urban grid-by-grid space superposition data into a preset flood ponding deep learning model, and generating an output result of the preset flood ponding deep learning model; the method comprises the steps of constructing a preset flood ponding deep learning model at each urban grid; obtaining urban flood ponding distribution data, comparing an output result of a preset flood ponding deep learning model with the urban flood ponding distribution data, and generating an urban flood ponding deep prediction model when the comparison result meets preset conditions.

According to the urban traffic evaluation method based on space superposition analysis, the urban flood ponding depth prediction model is constructed, so that the study of the correlation among multiple influence factors is realized, the uncertainty of the correlation among the multiple influence factors is avoided, the urban flood ponding depth prediction is inaccurate, and the accuracy of the urban flood ponding depth prediction is improved.

In an alternative embodiment, determining a grid-by-grid water depth prediction result using a city flood water depth prediction model based on weather forecast data includes: inputting a weather forecast data set of the urban target area into a site precipitation prediction model, and outputting site precipitation prediction data; and inputting site precipitation prediction data into a city flood ponding depth prediction model, and outputting a grid-by-grid ponding depth prediction result.

According to the urban traffic evaluation method based on space superposition analysis, through combining grid ponding depth prediction with weather forecast data, urban flood space distribution in different time periods such as hour by hour, day by day, week by week in the future is predicted, and the transportation can be deployed in advance, so that property and personal safety of residents are guaranteed to the greatest extent, and urban development is prevented from being subject to heavy creation.

In an alternative embodiment, the urban flood water depth prediction model is a long-term memory deep learning model.

According to the urban traffic evaluation method based on space superposition analysis, the long-term and short-term memory deep learning model remembers the result trained in the previous period, the model is adjusted based on the input data of the current time sequence, the model simulation result is continuously corrected by continuously increasing the input data, the information of a history unit can be better distributed, and the capability of capturing the long-term dependency relationship in the time sequence is provided.

In an alternative embodiment, the spatial superposition analysis is performed on the grid-by-grid ponding depth prediction result and the grid-by-grid traffic flow prediction result, and based on the spatial superposition analysis result, urban traffic in the urban grid is evaluated, and an urban traffic evaluation result is generated, including: comparing the grid-by-grid ponding depth prediction result with a preset depth threshold value to generate a ponding depth grade; comparing the grid-by-grid traffic flow prediction result with a preset flow threshold value to generate a traffic flow grade; and determining urban traffic risk levels in the urban grids based on the ponding depth levels and the traffic flow levels, and taking the urban traffic risk levels as urban traffic evaluation results.

According to the urban traffic evaluation method based on space superposition analysis, the danger levels of different areas of the city are obtained based on the ponding depth level and the traffic flow level, and the road traffic of the city is deployed and commanded in advance based on the predicted danger levels, so that the property and personal safety of residents are guaranteed to the greatest extent, and the urban development is prevented from being subjected to heavy creation.

In an alternative embodiment, the method further comprises: and (3) carrying out real-time deployment on urban road transportation based on the spatial superposition analysis result, and carrying out danger early warning.

According to the urban traffic evaluation method based on space superposition analysis, which is provided by the embodiment, the property and personal safety of residents are ensured to the greatest extent, and urban development is prevented from being subjected to heavy creation.

In a second aspect, the present invention provides an urban traffic evaluation device based on spatial superposition analysis, including: the first superposition module is used for acquiring site monitoring rainfall data, urban surface data and urban underground pipeline data in the target urban area, and carrying out spatial data superposition on the site monitoring rainfall data, the urban surface data and the urban underground pipeline data in the target urban area to generate urban grid-by-grid spatial superposition data; the first training module is used for training a preset deep learning model based on the grid-by-grid space superposition data of the cities to generate a city flood ponding depth prediction model; the first acquisition module is used for acquiring weather forecast data in a target urban area, and determining grid-by-grid ponding depth prediction results by utilizing a city flood ponding depth prediction model based on the weather forecast data; the first acquisition module is used for acquiring urban population distribution data in the current time period, inputting the urban population distribution data in the current time period into the urban traffic flow prediction model and generating a grid-by-grid traffic flow prediction result; the first evaluation module is used for carrying out space superposition analysis on the grid-by-grid ponding depth prediction result and the grid-by-grid traffic flow prediction result, evaluating urban traffic in the urban grid based on the space superposition analysis result and generating an urban traffic evaluation result.

In a third aspect, the present invention provides a computer device comprising: the urban traffic evaluation method based on the spatial superposition analysis comprises a memory and a processor, wherein the memory and the processor are in communication connection, the memory stores computer instructions, and the processor executes the computer instructions so as to execute the urban traffic evaluation method based on the spatial superposition analysis according to the first aspect or any corresponding implementation mode.

In a fourth aspect, the present invention provides a computer readable storage medium, on which computer instructions are stored, the computer instructions being configured to cause a computer to perform the urban traffic assessment method according to the first aspect or any one of the embodiments corresponding thereto.

Drawings

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings that are needed in the description of the embodiments or the prior art will be briefly described, and it is obvious that the drawings in the description below are some embodiments of the present invention, and other drawings can be obtained according to the drawings without inventive effort for a person skilled in the art.

FIG. 1 is a schematic flow chart of an urban traffic evaluation method based on spatial superposition analysis according to an embodiment of the invention;

FIG. 2 is a schematic diagram of a back propagation neural network model according to an embodiment of the present invention;

FIG. 3 is a schematic flow chart (II) of an urban traffic evaluation method based on spatial superposition analysis according to an embodiment of the invention;

FIG. 4 is a schematic flow chart (III) of an urban traffic evaluation method based on spatial superposition analysis according to an embodiment of the invention;

FIG. 5 is a block diagram of a long-term memory deep learning model unit according to an embodiment of the present invention;

FIG. 6 is a schematic flow chart (IV) of an urban traffic evaluation method based on spatial superposition analysis according to an embodiment of the invention;

FIG. 7 is a schematic diagram of a back propagation neural network model according to an embodiment of the present invention (II);

FIG. 8 is a schematic flow chart (V) of an urban traffic evaluation method based on spatial superposition analysis according to an embodiment of the invention;

FIG. 9 is a schematic flow chart (six) of an urban traffic evaluation method based on spatial superposition analysis according to an embodiment of the invention;

FIG. 10 is a block diagram of an urban traffic evaluation device based on spatial superposition analysis according to an embodiment of the present invention;

Fig. 11 is a schematic diagram of a hardware structure of a computer device according to an embodiment of the present invention.

Detailed Description

For the purpose of making the objects, technical solutions and advantages of the embodiments of the present invention more clear, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention, and it is apparent that the described embodiments are some embodiments of the present invention, but not all embodiments of the present invention; all other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.

The urban traffic evaluation method based on the spatial superposition analysis can be applied to urban flood disaster prediction electronic equipment and road traffic evaluation electronic equipment; the electronic device may include, but is not limited to, a notebook, desktop, mobile terminal, such as a cell phone, tablet, etc.; of course, the urban traffic evaluation method based on spatial superposition analysis provided in the present specification may also be applied to an application program running in the above-mentioned electronic device.

The embodiment of the invention provides an urban traffic evaluation method based on space superposition analysis, which improves the prediction precision of flood disasters by a research method of space element superposition analysis so as to realize accurate evaluation of urban traffic.

According to an embodiment of the present invention, there is provided an embodiment of an urban traffic evaluation method based on spatial superposition analysis, it being noted that the steps illustrated in the flowchart of the drawings may be performed in a computer system such as a set of computer-executable instructions, and that, although a logical order is illustrated in the flowchart, in some cases, the steps illustrated or described may be performed in an order different from that herein.

In this embodiment, an urban traffic evaluation method based on spatial superposition analysis is provided, which may be used in the above notebook, desktop computer, mobile terminal, such as mobile phone, tablet computer, etc., fig. 1 is a flowchart of an urban traffic evaluation method based on spatial superposition analysis according to an embodiment of the present invention, as shown in fig. 1, the flowchart includes the following steps:

step S101, site monitoring rainfall data, urban surface data and urban underground pipeline data in a target urban area are obtained, spatial data superposition is carried out on the site monitoring rainfall data, the urban surface data and the urban underground pipeline data in the target urban area, and urban grid-by-grid spatial superposition data are generated.

Specifically, as shown in table 1, the urban flood disaster influence data includes rainfall data, urban surface data, and urban underground piping data.

The rainfall data is a determining element of flood disasters, if no rainfall exists, urban flood is difficult to occur (except for a coastal city), wherein different rainfall distributions of urban rainfall (based on ground monitoring stations or radar rainfall monitoring) represent different rainfall space forms (strong rain, rainy time, rain center and rainfall running track) generated above the city, and the different rainfall space distributions directly determine the occurrence of the flood disasters of one city.

In addition to rainfall data, important factors to be considered are different under-ground conditions (roads, woodlands, grasslands, house lands) and ground topography change conditions (ground gradient trend, topography) of the urban ground surface; the influence of the urban surface factors is ignored in the traditional urban water accumulation prediction, so that the accuracy of the traditional research method for predicting the occurrence probability of the urban flood disaster based on rainfall data is relatively high, but the specific area of the urban flood disaster is generally difficult to accurately predict due to the influence of the urban surface factors is ignored, and the specific area of the urban flood disaster is accurately predicted by considering the urban surface data space distribution condition.

Urban underground pipeline data comprise an underground drainage pipeline outlet, a drainage pipeline caliber and a sewage drainage capacity, and the corresponding capacity of different areas of a city when flood disasters occur can be described, so that the possibility of occurrence of the flood disasters in different areas of the city can be further finely predicted.

And (3) analyzing urban flood by considering spatial distribution of urban sky, earth surface and underground main urban flood influence factors and by considering all spatial distribution scenes to carry out superposition analysis.

Step S102, training a preset deep learning model based on the grid-by-grid space superposition data of the cities to generate a city flood ponding depth prediction model.

Step S103, weather forecast data in the target urban area are obtained, and based on the weather forecast data, grid-by-grid ponding depth prediction results are determined by utilizing the urban flood ponding depth prediction model.

Step S104, urban population distribution data in the current time period is collected, the urban population distribution data in the current time period is input into an urban traffic flow prediction model, and a grid-by-grid traffic flow prediction result is generated.

Specifically, historical population distribution data and historical road traffic flow data of a preset historical time period of a target city area are obtained; inputting historical population distribution data of a preset historical time period into a preset road traffic flow prediction model, and obtaining an output result of the preset road traffic flow prediction model; comparing the output result of the preset road traffic flow prediction model with the historical road traffic flow data of the preset historical time period, and generating the urban road traffic flow prediction model when the comparison result meets the preset condition.

Further, the real-time road traffic flow of a city is related to population distribution (population distribution of residential buildings), time period in a day, motor vehicle number of a city, non-motor vehicle number and the like, but for a city, the road traffic flow of the city has close relation with the date in the year and the moment of the day, if the winter air temperature in one year is reduced, the number of non-motor vehicles is reduced, the number of resident travel selected transportation means is increased, on the contrary, the corresponding number is gradually increased along with warming, and the number of motor vehicles in front of holidays such as five, eleven and the like is suddenly increased all the day, because a plurality of people select driving travel; the traffic flow at the two times will increase sharply, and according to the different population distribution rules of the city, the traffic flow in different areas of the corresponding city has a certain rule; by comprehensively considering traffic flow of different areas of a prediction city, a Back-Propagation (BP) neural network model can be adopted, as shown in fig. 2, a target city is divided into a plurality of city grids based on a grid subdivision form, a BP neural network is constructed on each city subdivision grid, the annual date and time (month, day and hour) (X1) of the city and the population number (X2) of residents in the grid are taken as input information, the corresponding traffic flow on the grid is taken as output (Y), and the BP neural network is trained, so that the traffic flow prediction result at any moment in one year can be obtained, and the model is trained based on the traffic flow information of recent years (such as the near 5 years), so as to generate the urban road traffic flow prediction model.

Step S105, carrying out space superposition analysis on the grid-by-grid ponding depth prediction result and the grid-by-grid traffic flow prediction result, and evaluating urban traffic in the urban grid based on the space superposition analysis result to generate an urban traffic evaluation result.

Specifically, real-time deployment is carried out on urban road traffic migration based on a space superposition analysis result, and danger early warning is carried out; maximally guaranteeing property and personal safety of residents and preventing urban development from being subject to heavy creation.

Further, based on the constructed urban flood ponding depth prediction model and urban traffic flow prediction model of different grids, in order to timely adjust urban traffic, spatial superposition analysis is needed to be further carried out on the grid-by-grid ponding depth prediction result and the grid-by-grid traffic flow prediction result, spatial superposition analysis is needed to be further carried out on the same urban area, the ponding depth prediction result is larger than a certain dangerous threshold value, and the predicted traffic flow in the area reaches a certain value, so that the urban area is a dangerous area, and traffic dangers of different areas can be warned based on superposition relations between ponding depth and urban road traffic flow.

Further, based on the urban flood water depth prediction model, the area possibly causing great harm is predicted in time, the area where the flood disaster occurs in the city of 'hour-day-week' can be predicted, traffic in the 'unit grid' is evacuated, and the prediction result is fed back to the traffic control department so as to block road sections for a plurality of hours in advance, and vehicles and personnel are transferred.

Further, because the change and the prediction of weather are still more uncertain, if heavy storm is about to happen, real-time interaction is carried out with weather data, so that the improved result of the model is continuously corrected, and the traffic departments are linked in real time to correct the traffic command strategy, so that the optimal road traffic condition can be ensured to be adopted when each urban storm happens, the casualties and the loss of people and property under urban flood disasters can be avoided to the greatest extent, and the timely adjustment, optimization and layout of urban traffic are realized.

Further, based on the dangerous grades of the space superposition analysis result, responding sequentially from high to low, informing the current area of personnel withdrawal at the highest-level extremely dangerous area at the highest speed, rapidly transferring the traffic flow, and carrying out traffic control to prevent personnel from entering, so as to realize dangerous grade degradation; the lowest-level non-danger can not be managed, but places with large traffic flow and places with deep water accumulation can be hidden in general, so that if traffic migration command has residual force, the management can be distributed and managed in non-key areas in the same way, and emergency conditions can be dealt with more timely.

In this embodiment, an urban traffic evaluation method based on spatial superposition analysis is provided, which may be used in the above notebook, desktop computer, mobile terminal, such as mobile phone, tablet computer, etc., and fig. 3 is a flowchart of an urban traffic evaluation method based on spatial superposition analysis according to an embodiment of the present invention, as shown in fig. 3, where the flowchart includes the following steps:

Step S301, site monitoring rainfall data, urban surface data and urban underground pipeline data in a target urban area are obtained, spatial data superposition is carried out on the site monitoring rainfall data, the urban surface data and the urban underground pipeline data in the target urban area, and urban grid-by-grid spatial superposition data are generated.

Specifically, the step S301 includes:

step S3011, performing spatial discretization processing on the target city area, and generating a plurality of city grids.

Specifically, rainfall data are time and space double change factors, urban surface data and urban underground pipeline data are space change factors, and for a certain field of rainfall, whether a city can generate a flood disaster, the position where the flood disaster occurs and the degree of the flood disaster depend on the space distribution of the rainfall intensity, the urban surface space characteristic distribution and the urban underground drainage characteristic space distribution; in order to consider spatial superposition of all the influencing elements, the embodiment adopts a mesh subdivision form, and the mesh is divided into rectangular meshes, such as squares (10 m x 10m, for example), so that the city is spatially discretized, and a foundation is provided for carrying out regional prediction on flood risks in different areas of the city by processing the city spatially discretization.

Step S3012, dividing site monitoring rainfall data, urban surface data and urban underground pipeline data into a plurality of urban grids according to the spatial topological relation, and generating urban grid-by-grid spatial superposition data.

Specifically, based on each divided rectangular grid, the information such as precipitation intensity data, precipitation duration data, urban land surface land utilization type, urban gradient, urban underground drainage pipeline width, drainage capacity, water holding capacity and the like is all described on all grids of the city, all influence factor attributes are given to the grids, and all element space topologies determining whether flood disasters occur on the grids are superimposed on the grids, so that space superposition processing of all the influence elements is realized, each unit grid represents the regional position, regional attribute and precipitation information of the city, and each grid contains information after the influence element space superposition in table 1.

Step S302, training a preset deep learning model based on grid-by-grid space superposition data of cities to generate a city flood ponding depth prediction model; please refer to step S102 in the embodiment shown in fig. 1 in detail, which is not described herein.

Step S303, obtaining weather forecast data in a target urban area, and determining grid-by-grid ponding depth prediction results by utilizing a city flood ponding depth prediction model based on the weather forecast data; please refer to step S103 in the embodiment shown in fig. 1 in detail, which is not described herein.

Step S304, urban population distribution data in the current time period is collected, the urban population distribution data in the current time period is input into an urban traffic flow prediction model, and a grid-by-grid traffic flow prediction result is generated; please refer to step S104 in the embodiment shown in fig. 1 in detail, which is not described herein.

Step S305, carrying out space superposition analysis on the grid-by-grid ponding depth prediction result and the grid-by-grid traffic flow prediction result, and evaluating urban traffic in the urban grid based on the space superposition analysis result to generate an urban traffic evaluation result; please refer to step S105 in the embodiment shown in fig. 1 in detail, which is not described herein.

In this embodiment, an urban traffic evaluation method based on spatial superposition analysis is provided, which may be used in the above notebook, desktop computer, mobile terminal, such as mobile phone, tablet computer, etc., and fig. 4 is a flowchart of an urban traffic evaluation method based on spatial superposition analysis according to an embodiment of the present invention, as shown in fig. 4, where the flowchart includes the following steps:

step S401, station monitoring rainfall data, urban surface data and urban underground pipeline data in a target urban area are obtained, spatial data superposition is carried out on the station monitoring rainfall data, the urban surface data and the urban underground pipeline data in the target urban area, and urban grid-by-grid spatial superposition data are generated; please refer to step S101 in the embodiment shown in fig. 1 in detail, which is not described herein.

Step S402, training a preset deep learning model based on the grid-by-grid space superposition data of the cities to generate a city flood ponding depth prediction model.

Specifically, the step S402 includes:

step S4021, inputting the superposition data of the grid-by-grid space of the city into a preset flood water deep learning model, and generating an output result of the preset flood water deep learning model; and constructing a preset flood ponding deep learning model at each urban grid.

Specifically, the preset flood water deep learning model may be a Long Short-Term Memory deep learning model (LSTM).

Step S4022, obtaining urban flood ponding distribution data, comparing an output result of a preset flood ponding deep learning model with the urban flood ponding distribution data, and generating an urban flood ponding depth prediction model when the comparison result meets a preset condition.

Specifically, the urban flood ponding depth prediction model is a long-period memory deep learning model; in order to predict urban flood disasters, an LSTM deep learning model is adopted, spatial data of sky, land and underground are overlapped to each grid unit of a city, scene data of the urban flood disasters generated in the past of the unit are extracted, an LSTM model is built on each built grid of the urban unit, built three-in-one spatial information (sky, ground and underground) of each built grid of the unit is used as input data of the LSTM deep learning model of each grid of the unit, and meanwhile, corresponding historical monitored urban flood water distribution data (urban water distribution area and water depth data) are used as calibration data of each grid of the unit, so that an LSTM model is trained, and a prediction simulation result is obtained; the embodiment selects the LSTM deep learning neural network model because the model is a time sequence model, namely, input data is sequentially input and trained according to the time occurrence sequence, and the LSTM model remembers the previous time period #, the time sequence of the input data is the same as the time sequence of the input data t-1) training results based on the current time sequencetMoment) adjusts the model, and the model simulation result is continuously corrected by continuously increasing the input data, so that a history list can be better distributedMeta information, and has the ability to capture long-term dependencies in a time series.

Further, the LSTM model structure is shown in FIG. 5, in which the LSTM comprises a plurality of time-cycled cells, each cycled cell having an additional memory stateAnd a plurality of gates controlling information flow, including an input gate, a memory gate, a forget gate, and an output gate; in FIG. 5 +.>Indicating implicit status,/->Representation oftInput of time of day->、/>、/>、/>Respectively representtForget gate, input gate, output gate and cell state at the moment; the correlation calculation method of LSTM is as follows:

wherein:representing a memory update vector, ">，/>Respectively representing the information states of the previous unit and the unit;representing Sigmoid function (an activation function for neural networks),/a>Representing a hyperbolic cosine function; /> 、 Representing a weight matrix; />Representing the bias vector; />，/>Respectively representing the input of the last unit and the output of the unit; />Representing vector scalar product,/->Representation->And outputting the time.

It should be noted that, for a city, the city surface characteristic condition and the underground drainage characteristic condition generally have relatively small changes or relatively small changes in a short period, so that the model can be continuously trained only by continuously correcting the rainfall data set again during simulation; however, when considering the long-term situation, the earth surface and underground of a city may have larger changes, and for this situation, it is difficult to describe the traditional neural network and other circulation models, because it is difficult to consider the time sequence changes, but the LSTM model can completely consider the changes, and can change the earth surface characteristic data or underground characteristic data based on the dataset of the current moment, and the model can adjust the model trained by continuing the previous training content and combining with the new input data, so as to realize a prediction model which is continuously 'self-adjusted' and self-learned along with time sequence input, and accurately predict the urban flood.

Through continuous model training, an LSTM urban flood ponding model of the urban grid by grid can be obtained, the output result of the model is that under the input of precipitation, the flooding ponding depth H generated on the grid unit is not ponding, if 0 is not ponding, and if 0 is larger than 0, the ponding depth is generated, the output result of a preset flood ponding deep learning model is compared with urban flood ponding distribution data, and when the comparison result meets the preset condition, an urban flood ponding depth prediction model is generated; the preset condition may be that the loss value is smaller than a preset threshold, and a person skilled in the art can set the magnitude of the preset threshold according to actual needs, which is not limited herein; the urban flood ponding depth prediction model is based on the fact that each grid of the city can be predicted to output flood ponding depth, and further flood ponding distribution conditions of the whole city can be obtained.

Step S403, obtaining weather forecast data in a target urban area, and determining grid-by-grid ponding depth prediction results by utilizing a city flood ponding depth prediction model based on the weather forecast data; please refer to step S103 in the embodiment shown in fig. 1 in detail, which is not described herein.

Step S404, urban population distribution data in the current time period is collected, the urban population distribution data in the current time period is input into an urban traffic flow prediction model, and a grid-by-grid traffic flow prediction result is generated; please refer to step S104 in the embodiment shown in fig. 1 in detail, which is not described herein.

Step S405, carrying out space superposition analysis on the grid-by-grid ponding depth prediction result and the grid-by-grid traffic flow prediction result, and evaluating urban traffic in the urban grid based on the space superposition analysis result to generate an urban traffic evaluation result; please refer to step S105 in the embodiment shown in fig. 1 in detail, which is not described herein.

In this embodiment, an urban traffic evaluation method based on spatial superposition analysis is provided, which may be used in the above notebook, desktop computer, mobile terminal, such as mobile phone, tablet computer, etc., and fig. 6 is a flowchart of an urban traffic evaluation method based on spatial superposition analysis according to an embodiment of the present invention, as shown in fig. 6, where the flowchart includes the following steps:

step S601, acquiring site monitoring rainfall data, urban surface data and urban underground pipeline data in a target urban area, and performing spatial data superposition on the site monitoring rainfall data, the urban surface data and the urban underground pipeline data in the target urban area to generate urban grid-by-grid spatial superposition data; please refer to step S101 in the embodiment shown in fig. 1 in detail, which is not described herein.

Step S602, training a preset deep learning model based on grid-by-grid space superposition data of cities to generate a city flood ponding depth prediction model; please refer to step S102 in the embodiment shown in fig. 1 in detail, which is not described herein.

Step S603, weather forecast data in the target urban area are obtained, and based on the weather forecast data, grid-by-grid ponding depth prediction results are determined by using a city flood ponding depth prediction model.

Specifically, the step S603 includes:

and step S6031, inputting the weather forecast data set of the urban target area into a site precipitation prediction model, and outputting site precipitation prediction data.

Specifically, when the urban flood water depth prediction model is built, the method is applied to predicting whether the future urban flood disaster occurs or not in the next step, the built urban flood water depth prediction model inputs precipitation monitoring data of different sites distributed in the city, and the site precipitation prediction data is needed to be accessed for early prediction, and the site precipitation prediction data is relatively related to weather forecast data, so that the method is used for realizing the future site monitoring data based on the future weather forecast data, and site rainfall data prediction is carried out by constructing the relationship between the site precipitation prediction data and the site precipitation prediction data in a related manner; taking site precipitation prediction model as BP neural network model as an example, wherein the grid-by-grid BP neural network model is shown in fig. 7, wherein the input data is meteorological data (F), and the output grid-by-grid rainfall data is (P).

Further, historical meteorological data and site historical monitoring rainfall data in a target city area are obtained, the historical meteorological data are input into a site rainfall prediction model to be trained, and an output result of the site rainfall prediction model to be trained is obtained; comparing the output result of the site precipitation prediction model to be trained with the site historical monitoring rainfall data, and generating the site precipitation prediction model when the comparison result meets the preset condition.

And step S6032, inputting site precipitation prediction data into a city flood ponding depth prediction model, and outputting a grid-by-grid ponding depth prediction result.

Specifically, based on a site precipitation prediction model and weather forecast data, future urban storm ponding situations can be predicted, and a grid-by-grid ponding depth prediction result is output; the predicted scene of the future urban flood disaster is in one-to-one correspondence with the hour-by-hour, day-by-day and week-by-week scale data of the weather forecast; in this embodiment, the urban flood forecast period is consistent with the weather forecast period.

Step S604, collecting urban population distribution data in the current time period, inputting the urban population distribution data in the current time period into an urban traffic flow prediction model, and generating a grid-by-grid traffic flow prediction result; please refer to step S104 in the embodiment shown in fig. 1 in detail, which is not described herein.

Step S605, carrying out space superposition analysis on the grid-by-grid ponding depth prediction result and the grid-by-grid traffic flow prediction result, and evaluating urban traffic in the urban grid based on the space superposition analysis result to generate an urban traffic evaluation result; please refer to step S105 in the embodiment shown in fig. 1 in detail, which is not described herein.

In this embodiment, an urban traffic evaluation method based on spatial superposition analysis is provided, which may be used in the above notebook, desktop computer, mobile terminal, such as mobile phone, tablet computer, etc., and fig. 8 is a flowchart of an urban traffic evaluation method based on spatial superposition analysis according to an embodiment of the present invention, as shown in fig. 8, where the flowchart includes the following steps:

step S801, station monitoring rainfall data, urban surface data and urban underground pipeline data in a target urban area are obtained, spatial data superposition is carried out on the station monitoring rainfall data, the urban surface data and the urban underground pipeline data in the target urban area, and urban grid-by-grid spatial superposition data are generated; please refer to step S101 in the embodiment shown in fig. 1 in detail, which is not described herein.

Step S802, training a preset deep learning model based on grid-by-grid space superposition data of cities to generate a city flood ponding depth prediction model; please refer to step S102 in the embodiment shown in fig. 1 in detail, which is not described herein.

Step S803, weather forecast data in a target urban area are obtained, and a grid-by-grid ponding depth prediction result is determined by utilizing a city flood ponding depth prediction model based on the weather forecast data; please refer to step S103 in the embodiment shown in fig. 1 in detail, which is not described herein.

Step S804, collecting urban population distribution data in the current time period, inputting the urban population distribution data in the current time period into an urban traffic flow prediction model, and generating a grid-by-grid traffic flow prediction result; please refer to step S104 in the embodiment shown in fig. 1 in detail, which is not described herein.

And S805, performing spatial superposition analysis on the grid-by-grid accumulated water depth prediction result and the grid-by-grid traffic flow prediction result, and evaluating urban traffic in the urban grid based on the spatial superposition analysis result to generate an urban traffic evaluation result.

Specifically, the step S805 includes:

step S8051, comparing the grid-by-grid ponding depth prediction result with a preset depth threshold value to generate a ponding depth grade.

Specifically, in this embodiment, the unit of the grid ponding depth prediction result and the preset depth threshold may be set uniformly as: millimeter; the preset depth threshold may be set to one or more, such as the preset depth threshold is set to 0, 30, 60; when the grid ponding depth prediction result H is H=0, the current urban grid is indicated to be ponding-free; when the grid ponding depth prediction result H is 0< H <30, the current urban grid is indicated to be low ponding; when the grid ponding depth prediction result H is 30< H <60, the current urban grid is the medium ponding; when the grid ponding depth prediction result H is 60< H, the current urban grid is indicated to be high ponding; the preset depth threshold and the water accumulation depth level can be set by a person skilled in the art according to practical situations, and the method is not limited herein.

Step S8052, comparing the grid-by-grid traffic flow prediction result with a preset flow threshold value to generate a traffic flow grade.

Specifically, in this embodiment, the units of the mesh traffic flow prediction result and the preset flow threshold may be set uniformly as: vehicle/hour; the preset flow threshold may be set to one or more, such as the preset flow threshold is set to 0, 50, 100; when the grid traffic flow prediction result Q is Q=0, the current urban grid is no flow; when the grid traffic flow prediction result Q is 0< Q <50, the current urban grid is indicated to be low flow; when the grid traffic flow prediction result Q is 50< Q <100, the current urban grid is the medium flow; when the grid traffic flow prediction result Q is 100< Q, the current urban grid is indicated to be high flow; the preset traffic threshold and traffic flow level may be set by those skilled in the art according to actual conditions, and are not limited herein.

Step S8053, determining urban traffic risk level in the urban grid based on the ponding depth level and the traffic flow level, and taking the urban traffic risk level as an urban traffic evaluation result.

Specifically, as shown in table 2, the risk levels are divided into six levels, from no risk to extreme risk, different risk levels can be reflected on grids of the city according to different colors, further more visual display is given to management staff, and the higher the risk level is, the higher the level to be processed is; in addition, the flow grade and the water accumulation grade can be specifically classified according to the actual conditions of different cities, and only the classification of the grades is shown in the embodiment.

Note that: "none" means no danger; "lower" means lower risk; the meaning of "more dangerous" and "dangerous" is unchanged; "higher" means higher risk; "extreme" means extremely dangerous; the danger levels are six levels, and the corresponding color is the danger prompt color; wherein "(hidden)" means that although the area is not dangerous, if the sudden traffic flow increases or the city water increases, the danger level of the area will be rapidly increased, wherein the danger is still hidden and vigilance needs to be kept.

As shown in fig. 9, a method for evaluating urban traffic based on spatial superposition analysis is described below by way of a specific embodiment.

Example 1:

in order to realize early warning and forecasting of flood disasters of cities from an early hour scale to a peri-day scale and early forecasting of real-time traffic flow of road traffic, so that traffic operation is deployed in advance and travel is arranged, the method for evaluating urban traffic based on space superposition analysis in the embodiment comprises the following steps:

(1) Data preparation and space element superposition analysis: in order to comprehensively predict urban flood, a research method of space element superposition analysis is provided to provide a basis for subsequent model construction and flood prediction, and mainly comprises data element preparation, urban grid segmentation and influence element space superposition.

(1) Data element preparation.

The occurrence of urban flood disasters is related to many factors, and most of traditional urban flood disasters are studied by focusing the eyes on deep analysis of scene storm, but potential contribution of other influencing factors to the occurrence of flood disasters is ignored, and if the influence factors are considered from the space, the influence factors can be divided into three aspects of sky, earth surface and underground:

1) The (sky) site monitoring rainfall data is a determining element of a flood disaster, if no rainfall exists, urban flood is difficult to occur (except for a coastal city), wherein different rainfall distributions of urban rainfall (based on ground monitoring sites or radar rainfall monitoring) represent different rainfall space forms (rainfall, rainfall duration, rainfall rain center and rainfall center migration track) generated above the city, and the different rainfall space distributions directly determine the occurrence of the flood disaster of one city.

2) The urban ground surface topography and coverage data, besides the rainfall data monitored by the (sky) site, important factors to be considered are different ground surface conditions (roads, woodland, grasslands and house land) and ground surface topography change conditions (ground gradient trend and topography), because the traditional urban water accumulation prediction focuses on rainfall elements, but many researches neglect the influence of urban ground surface factors, the traditional research method has relatively high precision on the possibility of predicting the occurrence of flood disasters based on the rainfall factors, but because the influence of the urban ground surface factors is neglected, the specific area of the urban flood disasters is generally difficult to accurately predict, so the embodiment also considers the urban ground surface spatial distribution condition, thereby providing a foundation for realizing accurate prediction of urban flood disaster areas; the urban surface topography and coverage data comprise elevation data, urban surface coverage data, urban topography, gradient data and urban road data.

3) The urban underground drainage pipeline data comprises urban pipeline distribution data, urban pipeline caliber data, urban pipeline drainage data, urban drainage port distribution, sewage drainage capacity and the like, and can be used for describing the corresponding capacity of different areas of a city when flood disasters occur, so that the possibility of occurrence of the flood disasters in different areas of the city can be further finely predicted.

The urban flood is analyzed by considering the spatial distribution of the influence factors of the urban sky, the earth surface and the underground main urban flood and by considering all the spatial distribution scenes for superposition analysis; the specific data are shown in Table 1.

(2) Urban meshing

The influence factors of urban flood disasters are described, wherein the rainfall data are time and space double change factors, the urban surface factors and the urban underground factors are space change factors, and for a certain rainfall, whether the urban flood disasters occur, the positions of the occurrence flood disasters and the degree of the flood disasters depend on the space distribution of the rainfall intensity, the urban surface space characteristic distribution and the urban underground drainage characteristic space distribution; in order to take all the influencing elements into consideration in space superposition, a grid subdivision form is adopted, the simulated city boundary is subjected to grid division according to a certain size to obtain square grids (10 m is 10 m), the square grids are numbered, the numbered square grids are mapped to different areas of the city, space discretization of the city is realized, and a foundation is provided for carrying out regional prediction on flood risks of different areas of the city through carrying out space discretization treatment on the city.

(3) Influence element spatial overlay analysis

Dividing the preparation data into each grid cell according to the spatial topological relation: based on each square grid divided in the step (2), the precipitation intensity data, the precipitation duration data, the urban land surface land utilization type, the urban gradient, the urban underground drainage pipeline width, the drainage capacity, the water containing capacity and the like are all described on all grids of the city, all influence factor attributes are given to the grids, and all element space topologies for determining whether flood disasters occur on the grids are superimposed on the grids, so that space superposition processing of all the influence elements of the space is realized, and each unit grid represents the regional position, the regional attribute and the precipitation information of the city; specifically, firstly, the simulated city boundary is subjected to grid division according to a certain size; secondly, mapping the grid numbers to different areas of the city; finally, dividing the prepared data into all the grid cells according to the space topological relation; wherein each grid contains information after spatial superposition of the influencing elements in table 1.

(2) And (3) constructing a city flood model: based on the content described in (1), namely, constructing a grid-by-grid space superposition data format of cities, in order to be able to predict the occurrence of urban flood disasters, an LSTM deep learning model is adopted, scene data of urban flood disasters generated in the past by the space data of sky, land and underground are superimposed on each grid unit of the cities, scene data of the urban flood disasters generated in the past by the unit are extracted, an LSTM model is constructed on each constructed grid of the urban units, and constructed unit-by-unit space three-in-one information (sky, land and underground water information, namely, in FIG. 9) ，/>，/>，……，/>，) As the LSTM deep learning model for each cell grid (i.e., +.in FIG. 9)>，/>，/>，……，/>，/>) And then LSTM depth of each cell gridThe learning model outputs the depth of the urban grid-by-grid flood ponding (i.e. +.f. in FIG. 9)>，/>，/>，……，/>，/>) Meanwhile, urban flood disaster data (namely urban flood space distribution data and urban flood ponding depth monitoring data) are used as calibration data of unit-by-unit grids, so that an LSTM model is trained, and a prediction simulation result is obtained.

Through continuous model training, an LSTM urban flood ponding model of the urban grid by grid can be obtained, the output result of the model is that under the input of precipitation, the flooding ponding depth H generated on the grid unit is not ponded, and if the model is larger than 0, the model indicates that the ponding depth H is generated; and because each grid of the city outputs the depth of the flood ponding, the flood ponding distribution condition of the whole city can be obtained, and thus, the flood disaster ponding prediction model of the city is built.

(3) Building an urban flood early warning and forecasting model: based on the step (2), constructing a flood ponding model with predictable urban space distribution; when the flood ponding model of the city is built, the model is applied to the prediction of whether the flood disaster of the city occurs and the area where the flood disaster occurs in the future in the next step; in order to realize the function, a BP neural network model is also required to be constructed to simulate and predict future precipitation situations, because the constructed LSTM model inputs precipitation monitoring data of different sites distributed in the city, and forecast data of weather is required to be accessed in advance, in order to calculate future site monitoring data based on the future weather data, a relation between the weather forecast data and precipitation space monitoring data is also required to be further constructed, and the model is a constructed second learning model, and because the weather data and the site monitoring data are relatively related, the model construction can be realized only by adopting a common BP neural network, and the future city storm water situations can be forecast based on the weather forecast; the predicted situation of the future urban flood disaster can be in one-to-one correspondence based on the scale data of weather prediction, such as hour by hour, day by day, week and the like, namely the urban flood prediction period is always consistent with the weather prediction period; the BP neural network model by grid is shown in fig. 7, wherein the input data is meteorological data (F), and the precipitation data by grid is (P).

(4) Predicting urban road traffic flow and pre-judging urban traffic in advance:

the predicted urban ponding depth distribution has great influence on urban road traffic in areas with water accumulation depth, and is difficult to distinguish because the areas with water accumulation depth or traffic flow is less; therefore, in order to predict the influence of urban road traffic, real-time prediction data of road traffic flow are also required to be obtained, whether the water accumulation area affects the road traffic can be judged by carrying out space superposition analysis on the water accumulation data predicted by the city and the prediction data of the urban road traffic flow, and the urban road traffic is timely evacuated based on the prediction result; the method mainly comprises the following three aspects:

(1) urban road traffic flow prediction model construction

The real-time road traffic flow of a city is related to population distribution (population distribution of residential buildings), time period in a day, motor vehicle number of a city, non-motor vehicle number and the like, but for a city, the road traffic flow of the city has close relation with the date in an year and the moment of the day, if the temperature in winter in one year is reduced, the number of the non-motor vehicles is reduced, the number of resident travel selecting and sitting vehicles is increased, and conversely, the corresponding number is gradually increased along with warming, and if the number of motor vehicles is increased suddenly all the day before holidays in five, eleven and the like in one year, because a plurality of people can select driving travel; the traffic flow at the two times will increase sharply, and according to the different population distribution rules of the city, the traffic flow in different areas of the corresponding city has a certain rule; by comprehensively considering traffic flow of different areas of a predicted city and adopting a BP neural network model, as shown in fig. 2, based on grids divided in (2) in (1), a BP neural network is constructed on each grid divided in the city, the annual date and time (month, day and hour) of the city (X1) and the population number of residents in the grid (X2) are taken as input information, the corresponding traffic flow on the grid is taken as output (Y), the BP neural network is trained, the traffic flow prediction result at any moment in one year can be obtained, and the model can be trained based on the traffic flow information in recent years (such as the near 5 years).

(2) Urban road traffic flow and urban ponding prediction space superposition analysis

Based on the construction model of urban road traffic flow in different areas constructed in the step (1), in order to timely adjust urban traffic, further space superposition analysis is needed, namely, the flood prediction ponding result of a city and the traffic flow result predicted by the urban road are further subjected to space superposition analysis, on the same urban area, the ponding depth prediction result is larger than a certain dangerous threshold value, and the predicted traffic flow in the area reaches a certain value, the urban area is a dangerous area, the traffic dangers of the different areas can be warned based on the superposition relation between the ponding depth and the urban road traffic flow, the dangerous grade is divided into six grades as shown in the table 2, from no danger to extreme danger, different dangerous grades can be reflected on grids of the city according to different colors, further more visual display is provided for managers, and the higher the dangerous grade indicates that the grade to be processed is higher; in addition, the flow grade and the water accumulation grade can be specifically classified according to the actual conditions of different cities.

(3) Timely deploying urban road traffic migration based on the spatial superposition analysis result: based on the areas of the city of the predictable future "hour-day-week" in (3) where flood disasters occur in particular; timely predicting the region possibly causing great harm based on the prediction model, evacuating traffic in the 'unit grid', feeding back the prediction result to a traffic command department so as to block road sections for several hours in advance and transfer vehicles and personnel; however, because the change and prediction of weather are still more uncertain, if heavy storm is about to happen, real-time interaction is carried out with weather data, so that the improved result of the model is continuously corrected, and the traffic departments are linked in real time to correct the traffic command strategy, so that the optimal road traffic condition can be adopted when each urban storm happens, the casualties and people and property loss under urban flood disasters can be avoided to the greatest extent, and the timely adjustment, optimization and layout of urban traffic are realized; in addition, based on the danger level of spatial superposition analysis in the step (2), responding sequentially from high to low, informing the current area of personnel removal at the highest speed for the highest-level extremely dangerous area, rapidly transferring traffic flow, and performing traffic control to prevent personnel from entering, so as to realize danger level degradation; the lowest-level non-danger can not be managed, but places with large traffic flow and places with deep water accumulation can be hidden in general, so that if traffic migration command has residual force, the management can be distributed and managed in non-key areas in the same way, and emergency conditions can be dealt with more timely.

Based on a space superposition analysis idea and an urban space region discretization idea, the embodiment superimposes all factors such as sky, earth surface, underground and the like which influence the depth of urban flood ponding on the same grid; further, an LSTM model is constructed on each unit grid to predict the water accumulation depth of different areas of the city, and then a BP neural network model is combined as a pre-processing link to predict urban flood space distribution in different time periods of hour by hour, day by day and the like in the future in combination with weather forecast; finally, predicting the flow space distribution of different time nodes in the future of the urban traffic by constructing the BP neural network, carrying out space superposition analysis with the depth of the urban flood ponding to obtain the dangerous grades of different areas of the urban, and carrying out advanced deployment command on the urban road traffic based on the predicted dangerous grades.

The embodiment also provides an urban traffic evaluation device based on spatial superposition analysis, which is used for realizing the embodiment and the preferred implementation mode, and is not described in detail; as used below, the term "module" may be a combination of software and/or hardware that implements a predetermined function; while the means described in the following embodiments are preferably implemented in software, implementation in hardware, or a combination of software and hardware, is also possible and contemplated.

The embodiment provides an urban traffic evaluation device based on spatial superposition analysis, as shown in fig. 10, including:

the first superposition module 1001 is configured to obtain site monitoring rainfall data, urban surface data and urban underground pipeline data in a target urban area, and spatially superimpose the site monitoring rainfall data, the urban surface data and the urban underground pipeline data in the target urban area to generate grid-by-grid spatial superposition data;

the first training module 1002 is configured to train the preset deep learning model based on the grid-by-grid spatial superposition data of the city, and generate a city flood ponding depth prediction model;

a first obtaining module 1003, configured to obtain weather forecast data in the target urban area, and determine a grid-by-grid ponding depth prediction result by using a city flood ponding depth prediction model based on the weather forecast data;

the first collection module 1004 is configured to collect urban population distribution data in a current time period, input the urban population distribution data in the current time period into an urban traffic flow prediction model, and generate a grid-by-grid traffic flow prediction result;

the first evaluation module 1005 is configured to perform spatial superposition analysis on the grid-by-grid ponding depth prediction result and the grid-by-grid traffic flow prediction result, and evaluate urban traffic in the urban grid based on the spatial superposition analysis result, so as to generate an urban traffic evaluation result.

In some alternative embodiments, the first overlay module 1001 includes:

the discretization processing submodule is used for carrying out space discretization processing on the target city area to generate a plurality of city grids;

and the superposition sub-module is used for dividing the site monitoring rainfall data, the urban surface data and the urban underground pipeline data into a plurality of urban grids according to the space topological relation to generate urban grid-by-grid space superposition data.

In some alternative embodiments, the first training module 1002 includes:

the first input submodule is used for inputting the grid-by-grid space superposition data of the cities into a preset flood water deep learning model and generating an output result of the preset flood water deep learning model; the method comprises the steps of constructing a preset flood ponding deep learning model at each urban grid;

the first comparison sub-module is used for acquiring urban flood ponding distribution data, comparing an output result of a preset flood ponding deep learning model with the urban flood ponding distribution data, and generating an urban flood ponding deep prediction model when the comparison result meets a preset condition.

In some alternative embodiments, the first acquisition module 1003 includes:

The second input submodule is used for inputting the weather forecast data set of the urban target area into the site precipitation prediction model and outputting site precipitation prediction data;

and the third input sub-module is used for inputting site precipitation prediction data into the urban flood ponding depth prediction model and outputting a grid-by-grid ponding depth prediction result.

In some optional embodiments, the first training module specifically includes that the urban flood ponding depth prediction model is a long-term memory deep learning model.

In some alternative embodiments, the first evaluation module 1005 includes:

the second comparison submodule is used for comparing the grid-by-grid ponding depth prediction result with a preset depth threshold value to generate a ponding depth grade;

the third comparison sub-module is used for comparing the grid-by-grid traffic flow prediction result with a preset flow threshold value to generate a traffic flow grade;

and the determination submodule is used for determining the urban traffic risk level in the urban grid based on the ponding depth level and the traffic flow level, and taking the urban traffic risk level as an urban traffic evaluation result.

In some alternative embodiments, further comprising:

the deployment module is used for carrying out real-time deployment on urban road traffic migration based on the spatial superposition analysis result and carrying out danger early warning.

Further functional descriptions of the above respective modules and units are the same as those of the above corresponding embodiments, and are not repeated here.

The urban traffic evaluation device based on spatial superposition analysis in this embodiment is presented in the form of functional units, where the units refer to ASIC (Application Specific Integrated Circuit ) circuits, processors and memories executing one or more software or fixed programs, and/or other devices that can provide the above functions.

The embodiment of the invention also provides computer equipment, which is provided with the urban traffic evaluation device based on the spatial superposition analysis shown in the figure 10.

Referring to fig. 11, fig. 11 is a schematic structural diagram of a computer device according to an alternative embodiment of the present invention, as shown in fig. 11, the computer device includes: one or more processors 10, memory 20, and interfaces for connecting the components, including high-speed interfaces and low-speed interfaces; the various components are communicatively coupled to each other using different buses and may be mounted on a common motherboard or in other manners as desired. The processor may process instructions executing within the computer device, including instructions stored in or on memory to display graphical information of the GUI on an external input/output device, such as a display device coupled to the interface. In some alternative embodiments, multiple processors and/or multiple buses may be used, if desired, along with multiple memories and multiple memories. Also, multiple computer devices may be connected, each providing a portion of the necessary operations (e.g., as a server array, a set of blade servers, or a multiprocessor system). One processor 10 is illustrated in fig. 11.

The processor 10 may be a central processor, a network processor, or a combination thereof. The processor 10 may further include a hardware chip, among others. The hardware chip may be an application specific integrated circuit, a programmable logic device, or a combination thereof. The programmable logic device may be a complex programmable logic device, a field programmable gate array, a general-purpose array logic, or any combination thereof.

Wherein the memory 20 stores instructions executable by the at least one processor 10 to cause the at least one processor 10 to perform the methods shown in implementing the above embodiments.

The memory 20 may include a storage program area that may store an operating system, at least one application program required for functions, and a storage data area; the storage data area may store data created according to the use of the computer device, etc. In addition, the memory 20 may include high-speed random access memory, and may also include non-transitory memory, such as at least one magnetic disk storage device, flash memory device, or other non-transitory solid-state storage device. In some alternative embodiments, memory 20 may optionally include memory located remotely from processor 10, which may be connected to the computer device via a network. Examples of such networks include, but are not limited to, the internet, intranets, local area networks, mobile communication networks, and combinations thereof.

Memory 20 may include volatile memory, such as random access memory; the memory may also include non-volatile memory, such as flash memory, hard disk, or solid state disk; the memory 20 may also comprise a combination of the above types of memories.

The computer device further comprises input means 30 and output means 40. The processor 10, memory 20, input device 30, and output device 40 may be connected by a bus or other means, for example in fig. 11.

The input device 30 may receive input numeric or character information and generate key signal inputs related to user settings and function control of the computer apparatus, such as a touch screen, a keypad, a mouse, a trackpad, a touchpad, a pointer stick, one or more mouse buttons, a trackball, a joystick, and the like. The output means 40 may include a display device, auxiliary lighting means (e.g., LEDs), tactile feedback means (e.g., vibration motors), and the like. Such display devices include, but are not limited to, liquid crystal displays, light emitting diodes, displays and plasma displays. In some alternative implementations, the display device may be a touch screen.

The embodiments of the present invention also provide a computer readable storage medium, and the method according to the embodiments of the present invention described above may be implemented in hardware, firmware, or as a computer code which may be recorded on a storage medium, or as original stored in a remote storage medium or a non-transitory machine readable storage medium downloaded through a network and to be stored in a local storage medium, so that the method described herein may be stored on such software process on a storage medium using a general purpose computer, a special purpose processor, or programmable or special purpose hardware. The storage medium can be a magnetic disk, an optical disk, a read-only memory, a random access memory, a flash memory, a hard disk, a solid state disk or the like; further, the storage medium may also comprise a combination of memories of the kind described above. It will be appreciated that a computer, processor, microprocessor controller or programmable hardware includes a storage element that can store or receive software or computer code that, when accessed and executed by the computer, processor or hardware, implements the methods illustrated by the above embodiments.

Although embodiments of the present invention have been described in connection with the accompanying drawings, various modifications and variations may be made by those skilled in the art without departing from the spirit and scope of the invention, and such modifications and variations fall within the scope of the invention as defined by the appended claims.

Claims

1. The urban traffic evaluation method based on spatial superposition analysis is characterized by comprising the following steps of:

acquiring site monitoring rainfall data, urban surface data and urban underground pipeline data in a target urban area, and performing spatial data superposition on the site monitoring rainfall data, the urban surface data and the urban underground pipeline data in the target urban area to generate urban grid-by-grid spatial superposition data;

training a preset deep learning model based on the grid-by-grid space superposition data of the city to generate a city flood ponding depth prediction model;

acquiring weather forecast data in a target urban area, and determining a grid-by-grid ponding depth prediction result by utilizing the urban flood ponding depth prediction model based on the weather forecast data;

acquiring urban population distribution data in a current time period, and inputting the urban population distribution data in the current time period into an urban traffic flow prediction model to generate a grid-by-grid traffic flow prediction result;

performing spatial superposition analysis on the grid-by-grid ponding depth prediction result and the grid-by-grid traffic flow prediction result, and evaluating urban traffic in the urban grid based on the spatial superposition analysis result to generate an urban traffic evaluation result;

The step of performing spatial data superposition on the site monitoring rainfall data, the urban surface data and the urban underground pipeline data in the target urban area to generate urban grid-by-grid spatial superposition data comprises the following steps:

performing space discretization on the target city area to generate a plurality of city grids;

dividing the site monitoring rainfall data, the urban surface data and the urban underground pipeline data into a plurality of urban grids according to a space topological relation to generate urban grid-by-grid space superposition data; all the information of precipitation intensity data, precipitation duration data, urban land surface land utilization types, urban gradients, urban underground drainage pipeline widths, drainage capacity and water containing capacity is described on all grids of a city based on each divided urban grid, and all influence factor attributes are given to the grids so as to realize space superposition processing of all influence factors of a space;

the step of determining a grid-by-grid ponding depth prediction result by using the urban flood ponding depth prediction model based on the weather forecast data comprises the following steps:

inputting a weather forecast data set of the urban target area into a site precipitation prediction model, and outputting site precipitation prediction data;

Inputting the site precipitation prediction data into the urban flood ponding depth prediction model, and outputting a grid-by-grid ponding depth prediction result;

training a preset deep learning model based on the grid-by-grid space superposition data of the city to generate a city flood ponding depth prediction model, comprising the following steps:

inputting the grid-by-grid space superposition data of the city into a preset flood ponding deep learning model, and generating an output result of the preset flood ponding deep learning model; the method comprises the steps of constructing a preset flood ponding deep learning model at each urban grid;

and obtaining urban flood ponding distribution data, comparing an output result of the preset flood ponding deep learning model with the urban flood ponding distribution data, and generating the urban flood ponding deep prediction model when the comparison result meets preset conditions.

2. The method of claim 1, wherein the determining a grid-by-grid water depth prediction result using the urban flood water depth prediction model based on the weather forecast data comprises:

And inputting the site precipitation prediction data into the urban flood ponding depth prediction model, and outputting a grid-by-grid ponding depth prediction result.

3. The method of claim 1, wherein the urban flood water depth prediction model is a long-term memory deep learning model.

4. The method of claim 1, wherein the performing spatial overlay analysis on the grid-by-grid water depth prediction result and the grid-by-grid traffic flow prediction result, and evaluating urban traffic in the urban grid based on the spatial overlay analysis result, and generating an urban traffic evaluation result comprises:

comparing the grid-by-grid ponding depth prediction result with a preset depth threshold value to generate a ponding depth grade;

comparing the grid-by-grid traffic flow prediction result with a preset flow threshold value to generate a traffic flow grade;

and determining urban traffic risk levels in the urban grids based on the ponding depth levels and the traffic flow levels, and taking the urban traffic risk levels as urban traffic evaluation results.

5. The method as recited in claim 1, further comprising:

And (3) carrying out real-time deployment on urban road transportation based on the spatial superposition analysis result, and carrying out danger early warning.

6. An urban traffic evaluation device based on spatial superposition analysis, characterized in that the device comprises:

the first superposition module is used for acquiring site monitoring rainfall data, urban surface data and urban underground pipeline data in a target urban area, and carrying out space data superposition on the site monitoring rainfall data, the urban surface data and the urban underground pipeline data in the target urban area to generate urban grid-by-grid space superposition data;

the first training module is used for training a preset deep learning model based on the grid-by-grid space superposition data of the cities to generate a city flood ponding depth prediction model;

the first acquisition module is used for acquiring weather forecast data in a target urban area, and determining a grid-by-grid ponding depth prediction result by utilizing the urban flood ponding depth prediction model based on the weather forecast data;

the first acquisition module is used for acquiring urban population distribution data in a current time period, inputting the urban population distribution data in the current time period into the urban traffic flow prediction model and generating a grid-by-grid traffic flow prediction result;

The first evaluation module is used for carrying out space superposition analysis on the grid-by-grid ponding depth prediction result and the grid-by-grid traffic flow prediction result, evaluating urban traffic in the urban grid based on the space superposition analysis result and generating an urban traffic evaluation result;

the first superposition module includes:

the superposition sub-module is used for dividing the site monitoring rainfall data, the urban surface data and the urban underground pipeline data into a plurality of urban grids according to the space topological relation to generate urban grid-by-grid space superposition data; all the information of precipitation intensity data, precipitation duration data, urban land surface land utilization types, urban gradients, urban underground drainage pipeline widths, drainage capacity and water containing capacity is described on all grids of a city based on each divided urban grid, and all influence factor attributes are given to the grids so as to realize space superposition processing of all influence factors of a space;

the first training module comprises:

The first input submodule is used for inputting the grid-by-grid space superposition data of the city into a preset flood water deep learning model and generating an output result of the preset flood water deep learning model; the method comprises the steps of constructing a preset flood ponding deep learning model at each urban grid;

and the first comparison sub-module is used for acquiring urban flood ponding distribution data, comparing the output result of the preset flood ponding deep learning model with the urban flood ponding distribution data, and generating the urban flood ponding deep prediction model when the comparison result meets the preset condition.

7. The apparatus of claim 6, wherein the first acquisition module comprises:

and the third input sub-module is used for inputting the site precipitation prediction data into the urban flood ponding depth prediction model and outputting a grid-by-grid ponding depth prediction result.

8. The apparatus of claim 6, wherein the first training module specifically comprises the urban flood water depth prediction model as a long-term memory deep learning model.

9. The apparatus of claim 6, wherein the first evaluation module comprises:

and the determining submodule is used for determining urban traffic danger levels in urban grids based on the ponding depth level and the traffic flow level, and taking the urban traffic danger levels as urban traffic evaluation results.

10. The apparatus as recited in claim 6, further comprising:

11. A computer device, comprising:

a memory and a processor, the memory and the processor are in communication connection, the memory stores computer instructions, and the processor executes the computer instructions, thereby executing the urban traffic evaluation method based on spatial superposition analysis according to any one of claims 1 to 5.

12. A computer-readable storage medium, wherein computer instructions for causing a computer to execute the urban traffic evaluation method based on spatial superposition analysis according to any one of claims 1 to 5 are stored on the computer-readable storage medium.