Community fire rescue path optimization method and system

Abstract

The invention relates to the technical field of fire safety management, and particularly discloses a community fire rescue path optimization method and system, and the method comprises the steps: building a community map database through employing a geographic information system, collecting the traffic flow data of each road segment through employing a traffic camera, and then, carrying out the collection of the traffic flow data; performing time sequence analysis on the traffic flow data of each road section by using a data processing technology based on deep learning so as to capture a traffic flow time sequence dynamic change trend of each road section, and further performing global association coding on traffic flow time sequence change characteristics of each road section according to a communication relationship among the road sections of the community so as to obtain traffic flow time sequence data; therefore, the overall traffic road condition mode of the community is revealed, and the optimization recommendation of the rescue route is realized on the basis. Through the mode, real-time optimization of a community fire rescue path can be realized, the fire response speed is improved, and the loss caused by a fire disaster is reduced.

Description

Technical Field

[0001] The present application relates to the field of fire safety management technology, and more specifically, to a community fire rescue path optimization method and system. Background Art

[0002] In the context of rapid urbanization, communities are the basic building blocks of cities, and the effectiveness of their fire safety management is directly related to the safety of residents' lives and property. In emergency situations such as fires, rapid and accurate fire rescue response is the key to reducing casualties and property losses. However, traditional fire rescue path planning methods face many challenges, especially in complex and changing community environments and real-time traffic conditions.

[0003] Traditional firefighting route planning mainly relies on static maps and preset routes. Although these methods can guide firefighting and rescue operations to a certain extent, they have obvious limitations. Static maps cannot reflect the real-time traffic conditions in the community, such as sudden situations such as road closures, traffic accidents or traffic congestion, which will seriously affect the speed and efficiency of rescue vehicles. In addition, preset routes are often based on historical experience and fixed patterns, lacking flexibility and adaptability to specific fire and traffic conditions.

[0004] Therefore, an optimized community fire rescue path optimization method and system are expected. Summary of the invention

[0005] In order to solve the above technical problems, the present application is proposed. The embodiment of the present application provides a community fire rescue route optimization method and system, which uses a geographic information system to build a community map database, and uses traffic cameras to collect traffic flow data of each road section, and then uses deep learning-based data processing technology to perform time series analysis on the traffic flow data of each road section to capture the time series dynamic change trend of the traffic flow of each road section, and further based on the connectivity between the various road sections in the community, the time series change characteristics of the traffic flow of each road section are globally associated and encoded to reveal the overall traffic road condition pattern of the community, so as to achieve optimized recommendation of rescue routes on this basis. In this way, real-time optimization of community fire rescue routes can be achieved, the fire response speed can be improved, and the losses caused by fire can be reduced.

[0006] Accordingly, according to one aspect of the present application, a community fire rescue path optimization method is provided, which includes:

[0007] Use geographic information systems to build community map databases;

[0008] Use traffic cameras to collect the time queue of traffic flow on each road section;

[0009] Extracting the time series features of the traffic flow of each road section respectively from the time queue of the traffic flow to obtain a set of the time series features of the traffic flow of the road section;

[0010] Performing distinguishing enhancement on the set of time series features of the road section traffic flow based on the essential feature description to obtain a set of distinguishable enhanced time series features of the road section traffic flow;

[0011] Based on the connectivity relationship between the road sections in the community map database, a set of distinguishable enhanced time series features of the road section traffic flow is globally context-associated encoded to obtain a traffic condition time series encoding feature;

[0012] Based on the traffic condition temporal coding characteristics, an optimized rescue recommended route is generated.

[0013] According to another aspect of the present application, a community fire rescue path optimization system is provided, which includes:

[0014] A community map database construction module is used to construct a community map database using a geographic information system;

[0015] The vehicle flow collection module is used to collect the time queue of vehicle flow in each road section using traffic cameras;

[0016] A vehicle flow time series feature extraction module is used to extract vehicle flow time series features from the time queues of vehicle flow on each road section to obtain a set of road section vehicle flow time series features;

[0017] A feature distinguishing enhancement module, used for performing distinguishing enhancement on the set of time series features of the road section traffic flow based on the essential feature description to obtain a set of distinguishable enhanced time series features of the road section traffic flow;

[0018] A traffic condition global association coding module, for performing global context association coding on a set of distinguishable enhanced time series features of the traffic flow of the road section based on the connectivity relationship between the road sections in the community map database to obtain a traffic condition time series coding feature;

[0019] The rescue route optimization generation module is used to generate an optimized rescue recommended route based on the traffic road condition temporal coding characteristics.

[0020] Compared with the prior art, the community fire rescue route optimization method and system provided by this application uses a geographic information system to build a community map database, and uses traffic cameras to collect traffic flow data of each road section. Then, the traffic flow data of each road section is analyzed in time series using data processing technology based on deep learning to capture the dynamic change trend of the traffic flow time series of each road section, and further, based on the connectivity between the various road sections in the community, the traffic flow time series change characteristics of each road section are globally associated and coded to reveal the overall traffic road condition pattern of the community, thereby realizing the optimization recommendation of the rescue route on this basis. In this way, the real-time optimization of the community fire rescue route can be achieved, the fire response speed can be improved, and the losses caused by the fire can be reduced. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] By describing the embodiments of the present application in more detail in conjunction with the accompanying drawings, the above and other purposes, features and advantages of the present application will become more apparent. The accompanying drawings are used to provide a further understanding of the embodiments of the present application and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the present application and do not constitute a limitation of the present application. In the accompanying drawings, the same reference numerals generally represent the same components or steps.

[0022] Figure 1 This is a flowchart of a community fire rescue path optimization method according to an embodiment of the present application.

[0023] Figure 2 This is a data flow diagram of a community fire rescue path optimization method according to an embodiment of the present application.

[0024] Figure 3 This is a flowchart of step S140 in the community fire rescue path optimization method according to an embodiment of the present application.

[0025] Figure 4 This is a flowchart of step S150 in the community fire rescue path optimization method according to an embodiment of the present application.

[0026] Figure 5 This is a block diagram of a community fire rescue path optimization system according to an embodiment of the present application. DETAILED DESCRIPTION

[0027] Below, the exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited to the exemplary embodiments described here.

[0028] As mentioned in the above background technology, traditional static maps cannot reflect the real-time traffic conditions of the community, such as emergencies such as road closures, traffic accidents or traffic congestion, and lack flexibility and adaptability for specific fire and traffic conditions when planning routes. Based on this, this application proposes a community fire rescue route optimization method, which uses a geographic information system to build a community map database, and uses traffic cameras to collect traffic flow data of each road section. Then, the traffic flow data of each road section is analyzed in time series using data processing technology based on deep learning to capture the dynamic change trend of the traffic flow time series of each road section, and further based on the connectivity between the various road sections in the community, the traffic flow time series change characteristics of each road section are globally associated and encoded to reveal the overall traffic road condition pattern of the community, thereby realizing the optimization recommendation of rescue routes on this basis. In this way, real-time optimization of community fire rescue routes can be achieved, the fire response speed can be improved, and the losses caused by fire can be reduced.

[0029] Figure 1 This is a flowchart of a community fire rescue path optimization method according to an embodiment of the present application. Figure 2 Schematic diagram of data flow of the community fire rescue path optimization method according to the embodiment of the present application. Figure 1 and Figure 2 As shown, according to the community fire rescue path optimization method of the embodiment of the present application, the steps include: S110, using a geographic information system to build a community map database; S120, using traffic cameras to collect the time queues of traffic flow in each road section; S130, respectively extracting traffic flow time series features of the time queues of traffic flow in each road section to obtain a set of road section traffic flow time series features; S140, performing distinguishing enhancement on the set of road section traffic flow time series features based on the essential feature description to obtain a set of distinguishable enhanced time series features of road section traffic flow; S150, based on the connectivity relationship between the various road sections in the community map database, performing global context association encoding on the set of distinguishable enhanced time series features of the road section traffic flow to obtain traffic road condition time series coding features; S160, generating an optimized rescue recommended route based on the traffic road condition time series coding features.

[0030] In the above-mentioned community fire rescue path optimization method, the step S110 uses a geographic information system to construct a community map database. It should be understood that the road network inside and around the community is complex and diverse, and traditional static maps are difficult to accurately reflect real-time road conditions. As a powerful tool, the Geographic Information System (GIS) can present the geographical elements in the real world in a digital form through the collection, storage, management, analysis and visualization of geographic spatial data, provide accurate location information and spatial analysis capabilities, and provide new technical means for fire rescue path planning. By using the geographic information system to construct a community map database, the coordinate system can be used to locate and describe geographical objects, such as the starting and ending points of the road, intersections, etc., and different types of geographic information (such as road layers, building layers, etc.) can be superimposed and managed through the concept of layers, so as to construct a complete community geographic information model, intuitively present important geographic information such as road layout, building distribution, and location of fire facilities in the community, and provide an accurate geographic spatial framework for subsequent path planning, so that the driving path of the rescue vehicle can be planned and optimized in a real geographical environment, avoiding rescue delays caused by unfamiliarity with the community's geographical conditions.

[0031] In the process of constructing the community map database, obtaining high-resolution satellite images or aerial photography images is an important way to understand the overall picture of the community. The image data provided by commercial satellite companies have different resolution levels, which can be selected according to the size and accuracy requirements of the community. For example, for small communities, sub-meter resolution images may be sufficient to meet the needs, and can clearly distinguish the outlines of buildings, the direction of roads, and major geographic features. For large and complex communities, higher resolution data may be required to obtain more detailed information, such as the location of doors and windows of buildings, small gardens or courtyards, and other details. After acquiring the image data, it needs to be pre-processed, including image correction, cropping, and splicing to ensure the geometric accuracy and integrity of the image, so that it can accurately cover the scope of the target community and match the actual geographic coordinate system.

[0032] Topographic maps are an important part of geographic information, providing detailed information on terrain undulations, contour lines, water systems, landforms, etc. These data can usually be obtained from national surveying and mapping departments or professional geographic data suppliers. The contour data in topographic maps is crucial for analyzing the terrain slope and drainage conditions within the community. When planning rescue routes, understanding the terrain undulations can help determine the vehicle's traffic capacity and potential danger areas, such as avoiding roads with steep slopes to prevent rescue vehicles from encountering difficulties or danger during driving. Water system data can identify the location of water bodies such as rivers, lakes, and ponds, which is of great significance for the use of water sources and avoiding dangerous waters in fire rescue work. Through the digitization of topographic map data, it is converted into a format suitable for use by geographic information systems, such as vector data, so that it can be integrated and analyzed with other data.

[0033] Although satellite images and topographic maps can provide a lot of basic information, field measurements are still indispensable for some specific geographical elements. For example, the precise height of buildings, the specific location and attribute information of fire-fighting facilities (such as the type of fire hydrants, pressure level, water supply capacity, etc.), the actual width and road conditions of roads (such as whether there are potholes, obstacles, etc.), and some special facilities in the community (such as underground parking lot entrances, narrow passages, etc.). Field measurements can use professional measuring instruments, such as total stations, GPS receivers, etc., to ensure high data accuracy. For the measurement of building height, the total station can calculate its accurate height by measuring the three-dimensional coordinates of the top and bottom of the building, and record relevant building structure information, such as the number of floors, building materials, etc. This information is of great reference value for the formulation of rescue strategies and equipment selection during fire fighting. For the measurement of fire-fighting facilities, it is necessary not only to record their location coordinates, but also to record their technical parameters and maintenance in detail to ensure that they can fully play their role in the rescue process.

[0034] Roads are a key element of community transportation, and the accuracy of their data directly affects the planning of rescue routes. In addition to obtaining the basic direction and shape of roads through satellite images and topographic maps, other important information needs to be supplemented through field surveys. This includes the name of the road, road grade (such as main roads, secondary roads, branch roads, etc.), number of lanes, traffic control information (such as one-way streets, restricted areas, etc.), intersection types (crossroads, T-junctions, roundabouts, etc.) and road ancillary facilities (such as street lights, traffic signs, markings, etc.). The use of a vehicle-mounted mobile measurement system can efficiently collect the linear characteristics and ancillary facilities of the road. Through laser scanning and camera shooting, the three-dimensional spatial information and high-definition images of the road can be obtained to further enrich the details of the road data. For the collection of intersection information, special attention should be paid to recording its turning restrictions and traffic flow conditions, which is crucial for the traffic decision of rescue vehicles at intersections. At the same time, the slope, curvature and other parameters of the road are measured to evaluate the speed and handling performance of the rescue vehicle during driving, so as to ensure that the selected rescue route is safe and feasible.

[0035] At the same time, building data is also an important part of the community map database and plays a key role in fire rescue work. When collecting building data, in addition to the appearance and location information of the building, it is also necessary to have an in-depth understanding of its internal structure and functional layout. Through field visits and cooperation with property management departments, information on the use of buildings (such as residential, commercial, office, industrial, etc.) is obtained. Buildings with different uses have different fire risks and rescue needs. For residential buildings, it is necessary to understand the number of households, residential population density, location and number of evacuation passages, etc.; for commercial buildings, it is necessary to understand their business area, internal passage layout, fire protection facilities, and distribution of densely populated areas. Using a combination of architectural drawings and on-site surveys, draw the internal plan of the building, mark key locations such as stairwells, elevator shafts, fire passages, and fire control rooms, and provide detailed spatial information for personnel evacuation and fire rescue operations when a fire occurs. At the same time, record information such as the construction year of the building and the type of structure (such as brick-concrete structure, frame structure, steel structure, etc.). These factors will affect the fire resistance and collapse risk of the building, and need to be considered in the formulation of the rescue plan.

[0036] In addition to the above main data types, it is also necessary to collect other geographic element information in the community, such as the location and area of ​​public spaces such as parks, green spaces, and squares, which can be used as temporary evacuation and assembly sites when a fire occurs; the direction and distribution of power, communication, gas and other pipelines to avoid damage to these infrastructures caused by rescue operations, and to be able to cut off relevant pipelines in time when necessary to prevent secondary disasters; the location and related information of special places such as schools, hospitals, and nursing homes, which are densely populated and difficult to rescue, and need to be given special attention in the rescue plan. By collecting these diverse geographic element data, a comprehensive and detailed community geographic information dataset can be constructed, providing rich data resources for subsequent database construction and analysis.

[0037] Since the various types of data collected have different formats, precisions and coordinate systems, a series of data processing and integration operations are required to enable them to work together effectively in the geographic information system to form an organic whole. First of all, data from different sources may be stored in different file formats, such as satellite images may be in the common TIFF format, topographic map data may be in the DWG format, and field measurement data may be in the CSV or XLS format, while geographic information systems usually support specific vector and raster data formats, such as Shapefile, GeoJSON, etc. Therefore, it is necessary to use professional data conversion tools to convert various data formats into formats compatible with the geographic information system to ensure that the data can be smoothly imported into the system for subsequent processing. During the conversion process, attention should be paid to the accuracy and integrity of the data to avoid data loss or deformation due to format conversion. For example, when converting a building plan in CAD format to Shapefile format, it is necessary to carefully check whether the attribute information of the graphic elements is fully retained and whether the connectivity of the lines is correct to ensure the accuracy and availability of the building data in the geographic information system.

[0038] Since the data comes from different data sources, their coordinate systems may be inconsistent, which will cause the data to be unable to be accurately superimposed and analyzed in the geographic information system. Common coordinate systems include geodetic coordinate systems (such as WGS84, Beijing 54, Xi'an 80, etc.) and plane rectangular coordinate systems (such as local coordinate systems). When building a community map database, all data need to be unified into a specific coordinate system. Usually, a coordinate system that matches local practical applications is selected, such as a local city coordinate system, to ensure the accuracy and consistency of the spatial position of the data. Through coordinate conversion algorithms and tools, data in different coordinate systems are converted to the target coordinate system. This process involves complex mathematical calculations, including projection transformation, affine transformation, etc. It is necessary to select a suitable conversion method based on the specific situation and accuracy requirements of the data, and verify the accuracy of the converted results to ensure that the data can accurately reflect its actual geographical location in the new coordinate system.

[0039] In addition, considering that there may be various errors, missing values ​​and noise in the collected data, cleaning and preprocessing operations are required to improve the quality and availability of the data. For erroneous data, such as coordinate points that obviously do not conform to the actual geographical features, incorrect road connection relationships, etc., they need to be corrected through manual inspection and data editing tools. For missing values, a variety of methods can be used to handle them. For example, for missing building height information, it can be estimated and supplemented based on the height of similar buildings in the surrounding area; for missing road attribute information, it can be improved through field review or reference to other relevant materials. Noise data, such as salt and pepper noise in satellite images and outliers in measurement data, can be removed by filtering algorithms to smooth the data curve and improve the accuracy and reliability of the data. In the process of data cleaning, strict data quality control standards and processes should be established to ensure that each data item is carefully checked and verified, to ensure the integrity and accuracy of the data, and to provide a reliable basis for subsequent analysis and application.

[0040] In order to facilitate the management and analysis of a large amount of geographic information in the community map database, it is necessary to reasonably stratify and encode the data. Different types of geographic elements are divided into different layers according to their characteristics and uses, such as road layers, building layers, water system layers, fire protection facilities layers, etc. Each layer has an independent attribute table, which is used to store the relevant attribute information of the geographic elements of the layer, such as road name, width, grade and other attributes in the road layer, and building use, height, structure type and other attributes in the building layer. At the same time, each geographic element is given a unique code, and the coding rules should be systematic and logical, so that the type, location and other relevant information of the element can be easily identified and queried. For example, for roads, the road grade and sequence numbering method can be adopted. The main road is coded as "Z01", "Z02", etc., the secondary road is coded as "C01", "C02", etc., and the branch road is coded as "L01", "L02", etc. Through coding, specific road information can be quickly located and queried, improving the efficiency of data management and analysis.

[0041] In geographic information systems, establishing topological relationships between geographic elements is of great significance for data analysis and application. Topological relationships can accurately describe spatial relationships such as adjacency, inclusion, and intersection between geographic elements, such as the cross-connection relationship between roads and the inclusion relationship between buildings and plots. Through the topological construction algorithm, the processed and integrated data is topologically analyzed and processed to create topological elements such as topological polygons, topological lines, and topological points, and establish topological relationships between them. When constructing the road network topology, it is important to ensure that the connection points of the roads are accurate and the topological nodes at the intersections can correctly reflect the connectivity of the roads, which is crucial for subsequent path analysis and rescue route planning. At the same time, the establishment of topological relationships can also be used for data integrity checking and error repair. For example, through topological checking, problems such as isolated road segments or unclosed polygons can be discovered and corrected in a timely manner to ensure the logical consistency and accuracy of the data.

[0042] After data collection, processing and integration, we can start to build the community map database and establish a corresponding database management mechanism to ensure the efficient operation of the database and the safe storage and update of data. First of all, for database selection and design, select a suitable database management system (DBMS) according to the characteristics and application requirements of the community map data. Common geographic information database management systems include relational databases (such as Oracle Spatial, PostgreSQL with PostGIS extension) and non-relational databases (such as MongoDB, HBase, etc.). For community map databases, considering that they need to store and manage a large amount of structured geographic data and perform complex spatial query and analysis operations, relational databases with powerful spatial data processing capabilities are usually selected, and the spatial data engine of the geographic information system (such as ArcSDE, PostGIS, etc.) is combined for data storage and management. In the database design stage, according to the data stratification and coding system, a reasonable database table structure is designed, including feature class tables, attribute tables, topological relationship tables, etc., and the association relationships and primary key and foreign key constraints between each table are determined to ensure the integrity and consistency of the data. At the same time, considering the scalability and performance optimization of the data, a reasonable index design is carried out for the database, such as establishing spatial indexes for the spatial fields of geographic elements and establishing ordinary indexes for commonly used query fields to improve the efficiency of data query and retrieval.

[0043] Next, after the database design is completed, data is stored and updated. First, the processed and integrated data is imported into the geographic information database according to the designed database structure. This process can be achieved using the import tool provided by the database management system or the corresponding functional module of the geographic information system software. In the process of data storage, the accuracy and completeness of the data must be ensured, and the quality of the stored data must be checked and verified to promptly discover and correct possible problems. At the same time, a data update mechanism is established. Since community geographic information will change over time, such as new buildings, road reconstruction, and fire protection facility updates, the database needs to be updated regularly to ensure the timeliness and practicality of the data. Data updates can be achieved through regular data collection and comparison, using new satellite images, field measurement data, and existing database data for difference analysis, extracting the changed geographic elements, and updating them to the database. During the update process, attention should be paid to maintaining data consistency and backing up historical data for data backtracking and analysis.

[0044] In addition, it is crucial to ensure the security of the community map database, which involves the confidentiality, integrity and availability of data. Therefore, a series of security measures can be taken, such as setting user permissions and access control, granting corresponding database access rights according to the roles and responsibilities of different users, and ensuring that only authorized users can read and write to the database; performing data backup and recovery, regularly performing full and incremental backups of the database, storing the backup data in a safe and reliable medium, and establishing a data recovery strategy so that data can be restored in a timely manner when the database fails or data is lost; performing database performance monitoring and optimization, using the performance monitoring tools provided by the database management system to monitor the operating status of the database in real time, including CPU usage, memory usage, disk I / O and other indicators, to promptly discover and resolve possible performance bottlenecks, such as improving the operating efficiency of the database by optimizing query statements, adjusting database parameters, and increasing hardware resources, to ensure that the database can stably and efficiently provide data support for community fire rescue path optimization and other related applications.

[0045] In the above-mentioned community fire rescue path optimization method, the step S120 uses traffic cameras to collect the time queues of traffic flow in each section of the road. It should be understood that traffic flow data is a key indicator reflecting the road congestion situation, and changes in traffic flow in different time periods will have an important impact on the selection of rescue routes. Based on this, in the technical solution of the present application, the time queues of traffic flow in each section of the road are collected by traffic cameras to understand the traffic conditions of the road in real time, thereby providing dynamic traffic information for subsequent data analysis and rescue path optimization, helping to optimize the rescue path, avoid congested sections, and improve rescue efficiency. Here, the traffic camera is an effective real-time monitoring tool, which uses video monitoring technology to detect and count vehicles passing through a specific section of the road, and can record and transmit traffic flow data 24 hours a day.

[0046] In practical applications, traffic cameras can be installed on the main roads and key intersections of the community to ensure full coverage of the main traffic sections in the community. The camera should be installed in a place where the vehicle's driving conditions can be clearly captured to avoid obstruction and interference. The video data collected by the camera is transmitted to the data processing center in real time through wired or wireless transmission. In the data processing center, the video stream is processed using specialized video analysis software. Through advanced image recognition algorithms, the system can automatically analyze the vehicle information in the video stream, extract the traffic flow data of each section at fixed time intervals (such as every minute, etc.), and store it in chronological order to form a time queue for traffic flow.

[0047] In the above-mentioned community fire rescue path optimization method, the step S130 extracts the time series characteristics of the traffic flow of each section of the time queue to obtain a set of time series characteristics of the traffic flow of the section. In a specific example of the present application, the step S130 includes: inputting the time queues of the traffic flow of each section of the road into the traffic flow time series encoder based on the RNN-LSTM hybrid model to obtain a set of time series characteristics of the traffic flow of the section. That is, considering that the traffic flow data has obvious time series characteristics, its change trend is affected by various factors (such as time, date, weather, etc.), showing a complex nonlinear relationship. Therefore, in order to accurately capture the time series change trend of the traffic flow of each section, the present application further adopts the RNN-LSTM hybrid model with significant advantages in processing time series data to construct a traffic flow time series encoder, and processes the time queues of the traffic flow of each section of the road to obtain a set of time series characteristics of the traffic flow of the section. Those skilled in the art should know that the recurrent neural network (RNN) can memorize and utilize historical information in sequence data by introducing a loop structure, but it has the problem of gradient vanishing or gradient explosion when processing long sequence data. As a variant of RNN, the long short-term memory network (LSTM) effectively solves the long dependency problem of RNN by designing a special gating structure (input gate, forget gate, output gate). In the traffic flow time series encoder, the RNN-LSTM hybrid model takes the time queue of the traffic flow as the input sequence, and processes the traffic flow data of each time step in turn. By using the sequence memory ability of the RNN model and the gating mechanism of the LSTM model to learn the dynamic change law and long-term dependency in the traffic flow data, it can effectively extract the key time series information of the traffic flow of the road section, such as trend characteristics, periodic characteristics, volatility characteristics, etc., and comprehensively reflect the time series change pattern of the traffic flow of each road section, thereby providing more accurate information for subsequent traffic condition analysis and rescue path optimization.

[0048] In the above-mentioned community fire rescue path optimization method, the step S140 performs a distinguishing enhancement on the set of time series characteristics of the traffic flow of the road section based on the essential feature description to obtain a set of distinguishable enhanced time series characteristics of the traffic flow of the road section. In particular, considering that in actual traffic conditions, the time series characteristics of the traffic flow of different sections may have similarities and ambiguities, which will bring difficulties to the subsequent traffic condition analysis and rescue path selection. Therefore, in order to more accurately identify and distinguish the traffic conditions of each section, the present application further enhances and distinguishes the time series characteristics of the traffic flow of each section, so that the traffic flow characteristics of different sections are more obvious and easy to distinguish, thereby improving the accuracy and reliability of the rescue path optimization.

[0049] Figure 3FIG. 1 is a flow chart of step S140 in the community fire rescue path optimization method according to an embodiment of the present application. Figure 3 As shown, the step S140 includes: S141, performing field mapping on each road section traffic flow time series feature in the set of the road section traffic flow time series features to obtain a set of road section traffic flow time series semantic field features; S142, performing essential feature extraction on the set of road section traffic flow time series semantic field features to obtain essential features of the road section traffic flow time series semantic field; S143, based on the essential features of the road section traffic flow time series semantic field, performing feature differentiation and enhancement processing on each road section traffic flow time series feature in the set of the road section traffic flow time series features to obtain a set of distinguishable and enhanced time series features of the road section traffic flow.

[0050] Specifically, the step S141 is expressed by the formula:

[0051] h i =W1I i W2

[0052] Among them, I i represents the time series feature of the traffic flow of the ith road section in the set of the time series features of the traffic flow of the road section, W1 and W2 represent the first linear mapping matrix and the second linear mapping matrix respectively, and h i Indicates the I i The corresponding road section traffic flow temporal semantic field characteristics.

[0053] That is, first, a field mapping operation is performed on the time series characteristics of traffic flow on each road section to map them into a unified field space, thereby unifying the feature dimensions so as to better reflect the intrinsic correlation between the time series characteristics of traffic flow on each road section in the unified field space.

[0054] Specifically, the step S142 includes: calculating the field depth factor of each road section traffic flow temporal semantic field feature in the set of road section traffic flow temporal semantic field features to obtain a set of road section traffic flow temporal semantic field depth factors; based on the set of road section traffic flow temporal semantic field depth factors, calculating the field essential features of the set of road section traffic flow temporal semantic field features to obtain the road section traffic flow temporal semantic field essential features. More specifically, calculating the field essential features of the set of road section traffic flow temporal semantic field features includes: normalizing the set of road section traffic flow temporal semantic field depth factors, and using them as a set of weight coefficients to perform weighted aggregation on the set of road section traffic flow temporal semantic field features to obtain the road section traffic flow temporal semantic field essential features.

[0055] Accordingly, the step S142 is expressed by the formula:

[0056]

[0057] Among them, ‖·‖ represents the L2 norm of the vector, log2(·) represents the logarithmic function with base 2, and e i Indicates the I i The corresponding road traffic flow time series semantic field depth factor, exp(·) represents the natural exponential function with e as the base, a i Indicates the I i The corresponding normalized road traffic flow time series semantic field depth factor, v c It represents the essential characteristics of the time series semantic field of road traffic flow.

[0058] Here, by further calculating the field depth factor of the semantic field characteristics of the time series of traffic flow on each road section after mapping, we can measure its information density and feature importance in the field space, and get the set of field depth factors of the semantic field characteristics of the time series of traffic flow on the road section. Based on this, we use the field effect normalization mechanism to convert it into a weight coefficient, and perform weighted aggregation on the semantic field characteristics of the time series of traffic flow on each road section by position, so as to extract the feature representation that comprehensively reflects the overall universal law of the time series characteristics of traffic flow on each road section, that is, the essential characteristics of the semantic field of the time series of traffic flow on the road section.

[0059] Specifically, the step S143 includes: inputting the semantic field essential features of the road section traffic flow time series and the road section traffic flow time series features into a semantic difference feature extraction network based on a sigmoid activation function to obtain the road section traffic flow time series essential difference features; calculating the position point multiplication between the road section traffic flow time series essential difference features and the road section traffic flow time series features to obtain the distinguishable enhanced time series features of the road section traffic flow, which is expressed by the formula:

[0060] O i =I i ⊙[Sigmoid(M m v c +H m I i +b m )]

[0061] Among them, M m and H m are the two learnable weight parameter matrices of the semantic difference feature extraction network, and b m represents the bias vector, Sigmoid(·) represents the sigmoid activation function, ⊙ represents the multiplication by position, O i Indicates the I i The corresponding road section traffic flow can distinguish and enhance the temporal characteristics.

[0062] That is, taking the essential characteristics of the semantic field of the time series of the traffic flow of the road section as a benchmark, a neural network model is further introduced to learn the difference information in the time series characteristics of the traffic flow of each road section relative to the essential characteristics of the semantic field of the time series of the traffic flow of the road section, and through element-by-element distinguishable reinforcement processing, the unique information in the time series of the traffic flow of each road section is targetedly amplified to obtain a set of distinguishable and reinforced time series characteristics of the traffic flow of the road section, thereby ensuring that in the subsequent rescue path optimization process, the traffic conditions of different sections can be accurately identified and distinguished, thereby improving the accuracy and response speed of path selection.

[0063] In the above-mentioned community fire rescue route optimization method, the step S150, based on the connectivity relationship between the various road sections in the community map database, performs global context association coding on the set of distinguishable enhanced time series features of the road section traffic flow to obtain the traffic road condition time series coding features. Figure 4 FIG. 1 is a flow chart of step S150 in the community fire rescue path optimization method according to an embodiment of the present application. Figure 4 As shown, the step S150 includes: S151, extracting the connectivity relationship between the various road sections from the community map database, and constructing a road section topology matrix based on the connectivity relationship between the various road sections; S152, inputting the set of distinguishable enhanced temporal features of the road section traffic flow and the road section topology matrix into a traffic condition context encoder based on a graph convolutional neural network model to obtain a traffic condition temporal coding feature matrix as the traffic condition temporal coding feature.

[0064] It should be understandable that considering that the roads within the community are interconnected, a complex traffic network is formed. Therefore, while analyzing the time series change characteristics of traffic flow, it is also necessary to consider the global correlation effects between the various road sections in the community. Based on this, the present application further constructs a road section topology matrix based on the connectivity relationship between the various road sections in the community map database, and expresses the connectivity relationship between the road sections in the form of a mathematical matrix, so as to facilitate the subsequent global analysis and path optimization of traffic flow, and improve the computational efficiency and accuracy of rescue path planning. Specifically, the rows and columns of the road section topology matrix correspond to the various road sections in the community, respectively. If there is a direct connectivity relationship between two road sections, the corresponding position in the matrix is ​​marked as 1, otherwise it is marked as 0. Through this matrix representation, the topological structure of the entire community road network can be concisely and intuitively reflected, providing a clear basic data structure for subsequent traffic flow analysis and rescue path optimization.

[0065] After the above processing, the set of distinguishable enhanced time series features of the road traffic volume reflects the dynamic traffic conditions of the road, while the road section topology matrix describes the static structural relationship of the road network. In order to fully and accurately grasp the overall traffic conditions of the community, it is necessary to further organically integrate these two types of information to extract more comprehensive and context-aware traffic conditions. Based on this, the present application adopts a graph convolutional neural network model to process the set of distinguishable enhanced time series features of the road section traffic volume and the road section topology matrix. The graph convolutional neural network (GCN) model has unique advantages in processing graph structure data (such as road networks). Through the convolution operation defined on the graph structure, it can effectively aggregate the feature information of neighbor nodes. In the present application, the neighbor nodes of each road section are first determined according to the road section topology matrix, and the traffic flow time series features of each road section are regarded as the initial features of the nodes in the graph, and the connectivity between the road sections is regarded as the edge. In the graph convolution process, each node continuously updates its own feature representation by interacting with neighbor nodes and transmitting information, thereby effectively integrating local traffic flow information and global road network structure information. After multiple layers of convolution operations, the final result is a traffic condition time-series coding feature matrix that contains the comprehensive traffic condition characteristics of each road section and its surrounding sections. It effectively reflects the traffic status of each road section and its mutual influence with adjacent sections in the context of the entire community traffic network, thereby providing more comprehensive and accurate traffic condition information for rescue path optimization.

[0066] In the above-mentioned community fire rescue path optimization method, the step S160 generates an optimized rescue recommended route based on the traffic road condition time-series coding feature. In a specific example of the present application, the step S160 includes: inputting the traffic road condition time-series coding feature matrix into the rescue route optimization engine based on the large model to obtain the optimized rescue recommended route. It should be understood that the rescue route optimization engine based on the large model is built based on deep learning technology, has powerful computing power and complex problem solving capabilities, and uses the path search algorithm in the large model (such as an improved version of the A* algorithm, Dijkstra algorithm, etc.), and combines the traffic state information contained in the traffic road condition time-series coding feature matrix to quickly search and evaluate a large number of potential rescue paths in a complex community traffic network, so as to assign corresponding cost values ​​(such as time cost, distance cost, etc.) to each path according to factors such as traffic congestion and road capacity of different sections, and finally find the rescue recommended route with the lowest total cost by continuously exploring and comparing the costs of different paths, so as to ensure that the rescue vehicle can reach the accident site at the fastest speed and in the shortest time, thereby improving the success rate and timeliness of the rescue.

[0067] Preferably, the traffic condition time-series coding feature matrix is ​​passed through a rescue route optimization engine based on a large model to obtain an optimized rescue recommended route, which includes:

[0068] Expanding the traffic condition time-series coding feature matrix into a traffic condition time-series coding feature vector;

[0069] Calculate the distance between each pair of eigenvalues ​​of the traffic condition temporal coding feature vector, such as the L2 distance, and take the square root of the distance to obtain the traffic condition temporal coding distance representation matrix, that is,

[0070]

[0071] Wherein, V represents the traffic condition time series encoding feature vector, v i 、v j denote the eigenvalues ​​of the i-th position and the j-th position in the traffic condition temporal coding feature vector, d(·,·) denotes the distance metric function, and D i,j Represents the element value at the (i, j)th position in the traffic condition temporal coding distance representation matrix;

[0072] The traffic condition time series coding autocorrelation matrix of the traffic condition time series coding feature vector is obtained, that is, Wherein, V represents the traffic condition time series encoding feature vector, and V is a row vector, (·) T represents the transpose of a vector, represents matrix multiplication operation, D′ represents the traffic condition temporal coding autocorrelation matrix;

[0073] The traffic condition temporal coding feature vector is matrix-multiplied with the traffic condition temporal coding distance representation matrix to obtain the traffic condition temporal coding primary mapping vector, that is, Wherein, D represents the traffic condition temporal coding distance representation matrix, and V′ represents the traffic condition temporal coding primary mapping vector;

[0074] The traffic condition temporal coding first-level mapping vector is matrix-multiplied with the matrix product of the traffic condition temporal coding distance representation matrix and the traffic condition temporal coding self-correlation matrix to obtain the traffic condition temporal coding multi-level mapping vector Wherein, V″ represents the multi-level mapping vector of traffic condition temporal coding;

[0075] The traffic condition time-series coding multi-level mapping vector is interpolated with the traffic condition time-series coding associated eigenvector composed of the eigenvalues ​​of the traffic condition time-series coding auto-association matrix (if the eigenvalue is insufficient, interpolation or zero filling is performed) to obtain an optimized traffic condition time-series coding feature vector; and the optimized traffic condition time-series coding feature vector is passed through a rescue route optimization engine based on a large model to obtain an optimized rescue recommended route.

[0076] That is, considering that the road section traffic flow can distinguish a set of enhanced temporal features and the road section topology matrix respectively represents the temporal correlation coding features of the traffic flow of each road section and the connectivity relationship topological features of each road section, when performing traffic condition context coding based on the graph convolutional neural network model, the temporal correlation coding features of the traffic flow of each road section have been feature enhanced (that is, the nodes have been feature enhanced), but the road section topology matrix has not been feature enhanced (that is, the edges have not been feature enhanced). This will cause an imbalance and mismatch in the weighting of nodes and edges during graph convolution coding, resulting in the lack of context aggregation coding feature instance decision of the traffic condition temporal coding feature matrix, thereby affecting the accuracy of the optimized rescue recommended route obtained by the rescue route optimization engine based on the large model.

[0077] Therefore, through the linear target mapping representation of the similarity distance representation matrix of the traffic condition time-series coding feature vector after the traffic condition time-series coding feature matrix is ​​expanded, the self-correlation complete similarity instantiation of the traffic condition time-series coding feature matrix is ​​instantiated with a secondary target mapping representation based on a multi-level distribution hierarchy, and the association mismatch negative influence factor is compensated by the association fusion kernel bias to improve the degree of instance determination of the eigenvalues ​​of the traffic condition time-series coding feature matrix under similarity constraints, that is, the degree of significance of the eigenvalues ​​as instances for generating regression decisions, so as to improve the accuracy of the optimized rescue recommended route obtained by the rescue route optimization engine based on the large model of the traffic condition time-series coding feature matrix.

[0078] In summary, the community fire rescue path optimization method according to the embodiment of the present application is explained, which uses a geographic information system to build a community map database, and uses traffic cameras to collect traffic flow data of each road section. Then, the traffic flow data of each road section is analyzed in time series using data processing technology based on deep learning to capture the dynamic change trend of the traffic flow time series of each road section, and further, according to the connectivity between the various road sections in the community, the traffic flow time series change characteristics of each road section are globally associated and coded to reveal the overall traffic road condition pattern of the community, so as to achieve the optimization recommendation of the rescue route on this basis. In this way, the real-time optimization of the community fire rescue path can be achieved, the fire response speed can be improved, and the losses caused by the fire can be reduced.

[0079] Furthermore, the present application also provides a community fire rescue path optimization system.

[0080] Figure 5 FIG. 1 is a block diagram of a community fire rescue path optimization system according to an embodiment of the present application. Figure 5 As shown, according to the community fire rescue path optimization system 100 of the embodiment of the present application, it includes: a community map database construction module 110, which is used to construct a community map database using a geographic information system; a traffic flow collection module 120, which is used to collect the time queue of traffic flow of each road section using a traffic camera; a traffic flow time series feature extraction module 130, which is used to extract traffic flow time series features of the time queue of traffic flow of each road section respectively to obtain a set of road section traffic flow time series features; a feature distinguishing enhancement module 140, which is used to perform distinguishing enhancement on the set of road section traffic flow time series features based on the essential feature description to obtain a set of distinguishable enhanced time series features of road section traffic flow; a traffic condition global association coding module 150, which is used to perform global context association coding on the set of distinguishable enhanced time series features of the road section traffic flow based on the connectivity relationship between the various road sections in the community map database to obtain traffic condition time series coding features; a rescue route optimization generation module 160, which is used to generate an optimized rescue recommended route based on the traffic condition time series coding features.

[0081] Here, those skilled in the art can understand that the specific operations of each module in the above-mentioned community fire rescue path optimization system have been referred to above. Figures 1 to 4 The community fire rescue path optimization method has been introduced in detail, and therefore, its repeated description will be omitted.

[0082] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solution of the present invention can be modified or replaced by equivalents without departing from the spirit and scope of the technical solution of the present invention.

Claims

1. A community fire rescue path optimization method, characterized in that: include: Use geographic information systems to build community map databases; Use traffic cameras to collect the time queue of traffic flow on each road section; Extracting the time series features of the traffic flow of each road section respectively from the time queue of the traffic flow to obtain a set of the time series features of the traffic flow of the road section; Performing distinguishing enhancement on the set of time series features of the road section traffic flow based on the essential feature description to obtain a set of distinguishable enhanced time series features of the road section traffic flow; Based on the connectivity relationship between the road sections in the community map database, a set of distinguishable enhanced time series features of the road section traffic flow is globally context-associated encoded to obtain a traffic condition time series encoding feature; Based on the traffic condition temporal coding characteristics, an optimized rescue recommended route is generated.

2. The community fire rescue path optimization method according to claim 1, characterized in that: Extracting the time series features of the traffic flow of each road section respectively from the time queue of the traffic flow to obtain a set of the time series features of the road section traffic flow, including: The time queues of the traffic flow of each road section are respectively input into the traffic flow time series encoder based on the RNN-LSTM hybrid model to obtain a set of time series features of the traffic flow of the road section.

3. The community fire rescue path optimization method according to claim 2 is characterized in that: The set of the time series features of the road section traffic flow is distinguishably enhanced based on the essential feature description to obtain a set of distinguishable enhanced time series features of the road section traffic flow, including: Performing field mapping on each road section traffic flow time series feature in the set of road section traffic flow time series features to obtain a set of road section traffic flow time series semantic field features; Extracting essential features from the set of time-series semantic field features of the road section traffic flow to obtain the time-series semantic field essential features of the road section traffic flow; Based on the essential characteristics of the semantic field of the road section traffic flow time series, each road section traffic flow time series feature in the set of the road section traffic flow time series features is subjected to feature differentiation and enhancement processing to obtain a set of distinguishable and enhanced time series features of the road section traffic flow.

4. The community fire rescue path optimization method according to claim 3 is characterized in that: Extracting essential features from the set of the time series semantic field features of the road section traffic flow to obtain the essential features of the time series semantic field domain of the road section traffic flow, including: Calculating the field depth factor of each road section traffic flow time series semantic field feature in the set of road section traffic flow time series semantic field features to obtain a set of road section traffic flow time series semantic field depth factors; Based on the set of depth factors of the time series semantic field of the road section traffic flow, the field essential features of the set of time series semantic field features of the road section traffic flow are calculated to obtain the time series semantic field essential features of the road section traffic flow.

5. The community fire rescue path optimization method according to claim 4 is characterized in that: Based on the set of depth factors of the time series semantic field of the road section traffic flow, calculating the field essential features of the set of time series semantic field features of the road section traffic flow to obtain the time series semantic field essential features of the road section traffic flow, including: After normalizing the set of depth factors of the road section traffic flow temporal semantic field, they are used as a set of weight coefficients to perform weighted aggregation on the set of road section traffic flow temporal semantic field features to obtain the essential characteristics of the road section traffic flow temporal semantic field.

6. The community fire rescue path optimization method according to claim 5 is characterized in that: Based on the essential characteristics of the semantic field of the road section traffic flow time series, each road section traffic flow time series feature in the set of the road section traffic flow time series features is subjected to feature differentiation and enhancement processing to obtain a set of distinguishable and enhanced time series features of the road section traffic flow, including: Inputting the semantic field essential features of the road section traffic flow time series and the road section traffic flow time series features into a semantic difference feature extraction network based on a sigmoid activation function to obtain the road section traffic flow time series essential difference features; The essential difference characteristics of the road section traffic flow time series and the time series characteristics of the road section traffic flow are calculated and multiplied by the position points to obtain the distinguishable enhanced time series characteristics of the road section traffic flow.

7. The community fire rescue path optimization method according to claim 6, characterized in that: Based on the connectivity relationship between the road sections in the community map database, a set of distinguishable enhanced time series features of the road section traffic flow is globally context-associated encoded to obtain traffic condition time series coding features, including: Extracting connectivity relationships between the various road sections from the community map database, and constructing a road section topology matrix based on the connectivity relationships between the various road sections; The set of distinguishable enhanced temporal features of the road section traffic flow and the road section topology matrix are input into a traffic condition context encoder based on a graph convolutional neural network model to obtain a traffic condition temporal coding feature matrix as the traffic condition temporal coding feature.

8. The community fire rescue path optimization method according to claim 7, characterized in that: Based on the traffic condition temporal coding features, an optimized rescue recommended route is generated, including: The traffic condition time-series encoding feature matrix is ​​input into a rescue route optimization engine based on a large model to obtain the optimized rescue recommended route.

9. A community fire rescue path optimization system, characterized in that: include: A community map database construction module is used to construct a community map database using a geographic information system; The vehicle flow collection module is used to collect the time queue of vehicle flow in each road section using traffic cameras; A vehicle flow time series feature extraction module is used to extract vehicle flow time series features from the time queues of the vehicle flow of each road section to obtain a set of road section vehicle flow time series features; A feature distinguishing enhancement module, used for performing distinguishing enhancement on the set of time series features of the road section traffic flow based on the essential feature description to obtain a set of distinguishable enhanced time series features of the road section traffic flow; A traffic condition global association coding module, for performing global context association coding on a set of distinguishable enhanced time series features of the traffic flow of the road section based on the connectivity relationship between the road sections in the community map database to obtain a traffic condition time series coding feature; The rescue route optimization generation module is used to generate an optimized rescue recommended route based on the traffic road condition temporal coding characteristics.

Images