High-speed traffic knowledge graph system, method and device and storage medium

Abstract

The invention provides a high-speed traffic knowledge graph system, method and device and a storage medium, and the method comprises a graph construction module which is used for constructing a graph set comprising a domain knowledge graph, a real-time data graph and a business knowledge graph, and storing the graph set in a background, and the business knowledge graph is constructed in an integrated manner; the graph maintenance module is used for updating the knowledge graph in the graph set, and the real-time data graph is updated in real time in a fragment index mode; the intelligent analysis module is used for analyzing the traffic condition by adopting a machine learning model based on the real-time data graph stored in the background and obtaining an analysis result; and the intelligent question and answer module is used for generating corresponding answers of the questions of the user based on the graph set, extracting intentions of the questions of the user after receiving the questions of the user at the front end, judging whether the questions of the user comprise feature entities or not according to the intentions, and automatically generating prompt texts if the questions of the user do not comprise the feature entities. The map suitable for the expressway is established.

Description

Technical Field

[0001] The present invention relates to the field of traffic knowledge graph technology, and in particular to a high-speed traffic knowledge graph system, method, device and storage medium. Background Art

[0002] The traffic knowledge graph aims to represent and store traffic-related signals in a structured manner. As an emerging technical means, it shows great potential in improving the level of traffic management, facilitating more intelligent query, analysis and prediction.

[0003] In the related technologies, although knowledge graphs have made significant progress in the field of urban transportation, their application in the field of highway transportation is still in the initial exploration stage. There are significant differences between highway traffic patterns and urban traffic patterns. Therefore, when conducting flow analysis or abnormal vehicle processing, it is necessary to consider the characteristics of highways. For example, due to the limited roads on highways, the spread of traffic congestion and the chain reactions caused by accidents are more prominent on highways, but in existing technologies, analysis is often limited to the current section of the road. In the process of constructing knowledge graphs, it is usually divided into two methods: bottom-up and top-down. In practical applications, it has been proposed to use a hybrid strategy to make up for the shortcomings of a single method, but it is unclear how to combine them. In addition, highway traffic conditions change rapidly, and vehicles enter or exit the highway at every moment. The real-time data graph has always been a dynamically changing process. The traditional knowledge graph query change mechanism may be difficult to meet this timeliness requirement.

[0004] Based on the above analysis of the development status of this technology field, the existing technology lacks a method for constructing and maintaining highway traffic knowledge graphs, designing application functions that fit highway scenarios, using a clear process to adopt a combination of top-down and bottom-up methods when constructing business knowledge graphs, and using sharded indexing to maintain real-time data graphs with higher real-time requirements. Summary of the invention

[0005] The purpose of the present invention is to provide a high-speed traffic knowledge graph system, method, device and storage medium, aiming to solve the above-mentioned problems in the prior art.

[0006] According to a first aspect of an embodiment of the present invention, a high-speed transportation knowledge graph system is provided, including:

[0007] A graph construction module is used to construct a graph collection including a domain knowledge graph, a real-time data graph, and a business knowledge graph, and store the graph collection in the background. The business knowledge graph is constructed in an integrated manner.

[0008] The graph maintenance module is used to update the knowledge graph in the graph collection, where the real-time data graph is updated in real time using a shard index method;

[0009] Intelligent analysis module, which is used to analyze traffic conditions and obtain analysis results based on real-time data maps stored in the background using machine learning models;

[0010] The intelligent question-and-answer module is used to generate corresponding answers to user questions based on the graph set. After receiving the user's question from the front end, it extracts the user's intention of the question and determines whether the user's question includes feature entities based on the intention. If not, a prompt text is automatically generated.

[0011] According to a second aspect of an embodiment of the present invention, a high-speed transportation knowledge graph method is provided, comprising:

[0012] A graph collection including a domain knowledge graph, a real-time data graph and a business knowledge graph is constructed through a graph construction module, and the graph collection is stored in the background, wherein the business knowledge graph is constructed in an integrated manner;

[0013] The knowledge graph in the graph collection is updated through the graph maintenance module, where the real-time data graph is updated in real time using shard indexing;

[0014] Through the intelligent analysis module, based on the real-time data map stored in the background, the machine learning model is used to analyze the traffic situation and obtain the analysis results;

[0015] The intelligent question-and-answer module generates corresponding answers to user questions based on the graph collection. After receiving the user's question at the front end, the user's intention is extracted, and it is determined whether the user's question includes feature entities based on the intention. If not, prompt text is automatically generated.

[0016] According to a third aspect of an embodiment of the present invention, there is provided an electronic device, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the computer program is executed by the processor, the steps of the high-speed traffic knowledge graph method provided in the second aspect of the present disclosure are implemented.

[0017] According to a fourth aspect of an embodiment of the present invention, a computer-readable storage medium is provided, on which a program for implementing information transmission is stored. When the program is executed by a processor, the steps of the high-speed traffic knowledge graph method provided in the second aspect of the present disclosure are implemented.

[0018] The technical solution provided by the embodiment of the present invention includes the following beneficial effects: a graph collection suitable for the field of highway traffic is established as a whole, which can monitor the highway traffic status in real time and respond quickly, and support intelligent analysis and intelligent question and answer functions; the business knowledge graph is constructed in an integrated manner, which comprehensively considers the dynamic changes at the strategic level and the actual situation; the real-time data graph is updated in real time using a shard index method, which can assist the real-time data graph to achieve timely updates.

[0019] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate one or more embodiments of this specification or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this specification. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0021] Figure 1 is a schematic diagram of a high-speed transportation knowledge graph system according to an embodiment of the present invention;

[0022] Figure 2 is a schematic diagram of a shard index update according to an embodiment of the present invention;

[0023] Figure 3 is a schematic diagram of a high-speed traffic map structure according to an embodiment of the present invention;

[0024] Figure 4 is a schematic diagram of a system architecture of an embodiment of the present invention;

[0025] Figure 5 is a flow chart of a high-speed transportation knowledge graph method according to an embodiment of the present invention;

[0026] Figure 6 is a schematic diagram of an electronic device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0027] In order to enable those skilled in the art to better understand the technical solutions in one or more embodiments of this specification, the following will be combined with the drawings in one or more embodiments of this specification to clearly and completely describe the technical solutions in one or more embodiments of this specification. Obviously, the described embodiments are only part of the embodiments of this specification, not all of the embodiments. Based on one or more embodiments of this specification, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of this document.

[0028] System Example

[0029] According to an embodiment of the present invention, a high-speed transportation knowledge graph system is provided. Figure 1 is a flow chart of the high-speed transportation knowledge graph system according to an embodiment of the present invention. Figure 1 As shown, the high-speed transportation knowledge graph system according to an embodiment of the present invention specifically includes:

[0030] The graph construction module 10 is used to construct a graph set including a domain knowledge graph, a real-time data graph and a business knowledge graph, and store the graph set in the background, wherein the business knowledge graph is constructed in an integrated manner, specifically including:

[0031] The domain knowledge graph construction module is used to construct in a top-down manner, obtain the rules and common concepts in the field of high-speed transportation as a knowledge system, establish a first graph framework based on the knowledge system, fill in specific data into the first graph framework, and obtain a domain knowledge graph;

[0032] Since the domain knowledge graph focuses more on standardized and consistent domain knowledge, it is constructed in a top-down manner. The top-down approach starts from high-level concepts or a global perspective and then fills in specific data.

[0033] A real-time data graph construction module is used to construct in a bottom-up manner to obtain real-time data generated on high-speed roads, including toll station entrance and exit flow data, gantry vehicle flow data, emergency command data, asset information data, chartered vehicle order data, and hazardous cargo order data, to obtain a real-time data graph;

[0034] Toll station entrance and exit flow data refers to the time and license plate number of vehicles entering and exiting the toll station. Gantry traffic flow data refers to information including when vehicles pass through specific locations. Emergency command data refers to the design of traffic accidents, weather forecasts and other related events and their handling. Asset information data refers to the status and maintenance history of infrastructure such as road facilities, bridges and tunnels. Charter order data refers to details of car rental services for group travel or special needs. Dangerous goods transport manifest data refers to relevant regulations and implementation of dangerous goods transportation.

[0035] Since the real-time data map includes a large amount of detailed data on vehicles and engineering on the highway, a bottom-up approach is chosen to build it. The bottom-up approach usually processes basic data first and then integrates them into larger units.

[0036] A business knowledge graph construction module is used to construct using an integrated top-down and bottom-up approach, obtain high-speed transportation-related businesses and generate clustering results of the businesses using a clustering algorithm, calculate the average revenue contribution within each category based on the clustering results, select a category as a core business cluster based on the average revenue contribution, build a second graph framework based on the core business cluster, fill in the second graph framework and expand the graph using a bottom-up approach to obtain a business knowledge graph;

[0037] In the embodiment of the present invention, after determining the core business cluster, it is necessary to define relevant key concepts. If "highway monitoring" is defined as a core business in the cluster, it is necessary to define concepts such as "vehicle", "gantry", "toll station", etc.

[0038] That is to say, the integration method first carries out a top-down process, and after filling, it expands and refines the map from the bottom up.

[0039] In the embodiment of the present invention, a knowledge graph is constructed using data collected by sensors or cameras. The data processing process includes extraction, integration, processing, conversion and loading. The extraction of various types of data includes full extraction, incremental extraction and real-time extraction. Entity recognition and relationship extraction use natural language processing technology to identify entities and their attributes, and extract the relationships between entities, and construct the entities and relationships into a knowledge graph. In the embodiment of the present invention, the graph database is Neo4j.

[0040] The graph maintenance module 12 is used to update the knowledge graph in the graph collection, wherein the real-time data graph is updated in real time using a shard index method, specifically including:

[0041] Direct update module, used to update domain knowledge graph and business knowledge graph;

[0042] Since the domain knowledge graph and the business knowledge graph are updated less frequently, the update method commonly used in the field to ensure accuracy can be used. The embodiment of the present invention does not limit the update method of the knowledge graph in the direct update model;

[0043] The shard index update module is used to update the real-time data map, divide the real-time data map into data slices and store them in the corresponding nodes, maintain the copies of the data slices and the node storage locations in the standard retrieval library, first update the copies and then replace the original data slices at the corresponding node locations with the updated copies;

[0044] In the distributed index update module, a large-scale real-time data graph is divided into multiple data slices and stored in corresponding nodes respectively. Each node stores part of the graph and a corresponding series of data. The use of sharding can disperse the load and make the system easier to scale horizontally.

[0045] The node distribution can directly form a complete real-time data map. The source data is stored in the distribution node. Creating a copy can avoid operational errors of direct update, ensure security, and quickly locate the location of the required update. In the embodiment of the present invention, preferably, the unique identifier that can identify the object is used as the query object in the standard search library, and the updated copy is replaced after quality inspection. The original data piece is written to check whether the updated part causes new conflicts. Figure 2 Schematic diagram of updating a shard index according to an embodiment of the present invention. Figure 2 As shown, the distributed index architecture is shown.

[0046] The intelligent analysis module 14 is used to analyze the traffic situation and obtain the analysis results by using a machine learning model based on the real-time data map stored in the background, which specifically includes:

[0047] The traffic analysis module is used to obtain real-time data maps and construct road segment features, obtain a pre-constructed highway traffic map structure, wherein the nodes of the highway traffic map structure represent road segments, and the edges of the highway traffic map structure represent the existence of connection relationships, select dependent road segments corresponding to the road segment to be analyzed in the highway traffic map structure, integrate the road segment features of the road segment to be analyzed and the dependent road segments as the total features, and input the total features into a recursive neural network to obtain the traffic prediction result of the road segment to be analyzed;

[0048] In the field of traffic, edges are often used as road segments. However, in the embodiment of the present invention, it is important to select dependent road segments that are more relevant to the road segment to be analyzed when performing traffic analysis. It is more reasonable to use connection relationships as edges. Figure 3 Schematic diagram of the high-speed traffic map structure of an embodiment of the present invention. Figure 3 As shown, the process of problem abstraction is shown, ABCDE are all road sections, and other basic information of the graph structure is not restricted in the embodiment of the present invention;

[0049] The traffic analysis module is specifically used for:

[0050] Obtain toll station entrance and exit flow data, gantry traffic flow data, emergency command data, and asset information data in the real-time data map, and extract additional time features, gantry positions, section lengths, and average flow rates in the same historical time period as basic section features;

[0051] Fourier transform is used to extract the periodic features of traffic data, and historical events of road sections at different time scales are extracted. In the embodiment of the present invention, the time scale includes hours, days, weeks and months, that is, events before the time scale interval based on the current time node are extracted, that is, events one hour ago, one day ago, one week ago and one month ago, and the periodic features and historical events are spliced ​​into auxiliary features of the road section;

[0052] Splicing the basic features of the road section and the auxiliary features of the road section into the road section features;

[0053] Use a pre-trained graph neural network to generate an embedding vector for each node, calculate the cosine similarity between the node represented by the section to be analyzed and the node within a preset number of hops based on the embedding vector, filter out dependent sections based on the cosine similarity, that is, filter out several sections with higher cosine similarity rankings as dependent sections, combine the section features of the section to be analyzed and all dependent sections as total features, and use a recursive neural network to obtain a traffic prediction result. In the embodiment of the present invention, the recursive neural network is a long short-term memory network LSTM, because the LSTM model can predict the results of different time steps, which helps to provide more sufficient information reserves during intelligent question answering;

[0054] If a section of a highway is congested, the other routes available to the driver are fewer than those in urban sections. In urban traffic analysis, only directly related sections are often focused on, while in the embodiments of the present invention, sections that are not directly related can be focused on, and the preset hop count setting can limit the distance between the section to be analyzed and the dependent section to a certain extent.

[0055] In the embodiment of the present invention, the abnormal identification of vehicles is mainly divided into two categories. The first is based on the real-time data analysis generated by the vehicle itself, corresponding to the first abnormal identification module. The second is to identify the lane change and intersection of the vehicle. For high-speed traffic, the lane change and intersection of vehicles are more stringent than those in the city, corresponding to the second abnormal identification module.

[0056] The first abnormality identification module is used to classify all vehicles in the current high-speed traffic into vehicle classes according to the size of the vehicle, obtain the real-time data map to build the clustering results of the vehicle classes, and regard the vehicles with abnormal clustering results as the first abnormal vehicles;

[0057] In the embodiment of the present invention, vehicles can be divided into small vehicles, medium vehicles and large vehicles according to their volume scale, and corresponding cluster graphs are constructed for each type of vehicle, thereby obtaining a small vehicle cluster graph, a medium vehicle cluster graph and a large vehicle cluster graph; vehicles of similar model and volume have many commonalities in form patterns, so in each cluster graph, a vehicle sample that is far away from the cluster can be regarded as the first abnormal vehicle;

[0058] The second abnormality recognition module is used to use the target detection model to identify vehicles changing lanes and interspersing, and regard the vehicles whose lane changes and interspersing times exceed a preset value as the second abnormal vehicles.

[0059] The second abnormality identification module is specifically used for:

[0060] Obtain high-speed traffic video and divide it into frames, reduce the resolution of the frame as a pre-processed image. For the embodiment of the present invention, the content to be recognized does not require the clarity of the image, and reducing the resolution can improve the computing efficiency. In the pre-processed image, the distance threshold range where the lane line is located is the area to be recognized;

[0061] A lightweight target recognition model is used to identify the lane changes of vehicles in the area to be identified. When the number of lane changes of a vehicle exceeds the threshold, a Kalman filter is used to track the lane changes of the vehicle and a list of all filter tracking is maintained, in which the number of lane changes of the vehicle is recorded.

[0062] It should be noted that the number threshold is not equal to the preset value. The number threshold is the critical value that needs to be detected by the Kalman filter, and the preset value is the critical value for determining it as the second abnormal vehicle. Preferably, for chartered order data and hazardous cargo order data, they can be tracked directly through the Kalman filter, or a smaller number threshold and preset value can be set to focus on them.

[0063] The intelligent question-answering module 16 is used to generate corresponding answers to user questions based on the graph set, extract the intention of the user's question after receiving the user's question at the front end, and determine whether the user's question includes a feature entity based on the intention. If not, a prompt text is automatically generated, specifically including:

[0064] The intention analysis module is used to identify the intention of users’ questions using the naive Bayes classifier. The user’s questions are input through the front-end interface.

[0065] In the embodiment of the present invention, the user asks a question "How is the traffic flow on section A?", and the naive Bayes classifier classifies the user's question as "traffic flow prediction";

[0066] The automatic prompt module is used to extract the content entities in the user's question and obtain the characteristic topics defined in the preset rule matching table. If there is a content entity belonging to the characteristic topic, the user's question includes the characteristic entity, and the corresponding answer is directly searched in the atlas collection. Otherwise, a prompt text is generated to prompt the user to supplement the characteristic entity, and the answer is retrieved after the user has completed the supplement.

[0067] The content entities identified by named entities are "Road Section A" and "Traffic Flow". The feature theme defined in the preset rule matching table for "Traffic Flow Forecast" is "Time". That is, time is the key point for traffic prediction. There is no time-related content in the content entity, so a prompt text "Suggestion to supplement time information" is generated for the user end;

[0068] After the user's question statement is optimized, a search request is initiated to the backend to obtain results.

[0069] In addition, the high-speed transportation knowledge graph system also includes sorting retrieval,

[0070] The above technical solutions of the embodiments of the present invention are illustrated with reference to the following drawings.

[0071] Figure 4 Schematic diagram of the system architecture of an embodiment of the present invention. Figure 4 As shown, the overall architecture of the present invention is demonstrated, including a platform service layer, a data support layer and a basic layer. The platform service layer is responsible for user interaction and interface display, including knowledge graph construction, intelligent question and answer and other functions; the data support layer is used to store data and related processing; the basic layer is used for the deployment of software and hardware services, and a complete high-speed transportation knowledge graph system architecture is constructed. The system adopts containerized deployment.

[0072] To sum up, in response to the existing problems, the highway traffic knowledge graph system invented this time has established a graph collection suitable for the highway traffic field as a whole, which can monitor the highway traffic status in real time and respond quickly, and support intelligent analysis and intelligent question and answer functions; in traffic analysis, the road sections are innovatively abstracted as nodes, which can capture the influence dependencies of directly or indirectly connected sections, so as to more completely consider the chain effect of highway traffic congestion; in vehicle abnormality identification, only the distance threshold range of the lane line is defined as the area to be identified, which reduces the complexity of image recognition calculation; in the intelligent question and answer module, prompt text can be automatically generated to assist users in optimizing question statements and improving the question and answer effect; the business knowledge graph is constructed in an integrated manner, which comprehensively considers the dynamic changes at the strategic level and the actual situation; the real-time data graph is updated in real time using a shard index method, which can assist the real-time data graph to achieve timely updates, and the data is divided and distributed and stored on different nodes, which can also disperse the load and make the system easier to expand horizontally.

[0073] Method Embodiment

[0074] According to an embodiment of the present invention, a high-speed transportation knowledge graph method is provided. Figure 5 is a schematic diagram of a high-speed transportation knowledge graph method according to an embodiment of the present invention. Figure 5 As shown, the high-speed traffic knowledge graph method according to an embodiment of the present invention specifically includes:

[0075] In step S510, a graph set including a domain knowledge graph, a real-time data graph and a business knowledge graph is constructed by a graph construction module, and the graph set is stored in the background, wherein the business knowledge graph is constructed in an integrated manner, specifically including:

[0076] The domain knowledge graph construction module is constructed in a top-down manner to obtain the rules and common concepts in the field of high-speed transportation as a knowledge system. The first graph framework is established based on the knowledge system, and specific data is filled into the first graph framework to obtain the domain knowledge graph.

[0077] The real-time data map construction module is constructed in a bottom-up manner to obtain the real-time data generated on the high-speed traffic road, including the flow data of toll station entrances and exits, the flow data of gantry vehicles, emergency command data, asset information data, chartered vehicle order data and dangerous goods order data, and obtain the real-time data map;

[0078] The business knowledge graph construction module is constructed using an integrated top-down and bottom-up approach, high-speed transportation-related businesses are obtained and clustering results of the businesses are generated using a clustering algorithm. The average revenue contribution within each category is calculated based on the clustering results, and a category is selected as the core business cluster based on the average revenue contribution. The second graph framework is constructed based on the core business cluster, and the second graph framework is filled in and the graph is expanded in a bottom-up approach to obtain a business knowledge graph.

[0079] In step S520, the knowledge graph in the graph collection is updated by the graph maintenance module, wherein the real-time data graph is updated in real time using a shard index method, specifically including:

[0080] Update the domain knowledge graph and business knowledge graph through direct update modules;

[0081] The real-time data graph is updated through the shard index update module, and the real-time data graph is divided into data slices and stored in the corresponding nodes. The copies of the data slices and the node storage locations are maintained in the standard retrieval library. The copies are updated first and then the updated copies replace the original data slices at the corresponding node positions.

[0082] In step S530, the intelligent analysis module uses a machine learning model to analyze the traffic situation based on the real-time data map stored in the background and obtains the analysis results, which specifically include:

[0083] The traffic analysis module is used to obtain real-time data maps and construct road segment features, and obtain a pre-constructed highway traffic map structure, wherein the nodes of the highway traffic map structure represent road segments, and the edges of the highway traffic map structure represent the existence of connection relationships. The dependent road segments corresponding to the road segments to be analyzed are selected in the highway traffic map structure, and the road segment features of the road segments to be analyzed and the dependent road segments are integrated as the total features. The total features are input into the recursive neural network to obtain the traffic prediction results of the road segments to be analyzed;

[0084] The traffic analysis module is specifically used for:

[0085] Obtain toll station entrance and exit flow data, gantry traffic flow data, emergency command data, and asset information data in the real-time data map, and extract additional time features, gantry positions, section lengths, and average flow rates in the same historical time period as basic section features;

[0086] Use Fourier transform to extract the periodic features of traffic data, extract historical events of road sections at different time scales, and splice the periodic features and historical events into auxiliary features of road sections;

[0087] Splicing the basic features of the road section and the auxiliary features of the road section into the road section features;

[0088] Use a pre-trained graph neural network to generate an embedding vector for each node. Based on the embedding vector, calculate the cosine similarity between the node represented by the section to be analyzed and the nodes within a preset number of hops. Filter the dependent sections based on the cosine similarity, and combine the section features of the section to be analyzed and all dependent sections as the total features. Use a recursive neural network to get the traffic prediction results.

[0089] All vehicles in the current high-speed traffic are divided into vehicle classes according to the vehicle volume through the first abnormal identification module, and the clustering results of the vehicle classes are constructed by obtaining the real-time data map, and the vehicles with abnormal clustering results are regarded as the first abnormal vehicles;

[0090] The second abnormal identification module uses the target detection model to identify vehicles changing lanes and inserting, and the vehicle whose lane change frequency exceeds a preset value is regarded as the second abnormal vehicle.

[0091] The second abnormality identification module is specifically used for:

[0092] Obtain high-speed traffic video and divide it into frames, reduce the resolution of the frames as pre-processed images, and the distance threshold range of the lane lines in the pre-processed images is the area to be identified;

[0093] A lightweight target recognition model is used to identify lane changes of vehicles in the area to be identified. When the number of lane changes of a vehicle exceeds the threshold, a Kalman filter is used to track the lane changes of the vehicle, and a list of all filtered tracking is maintained.

[0094] In step S540, the intelligent question-answering module generates the corresponding answer to the user's question based on the graph set, extracts the intention of the user's question after receiving the user's question at the front end, and determines whether the user's question includes the feature entity according to the intention. If not, a prompt text is automatically generated, specifically including:

[0095] The intent analysis module uses the naive Bayes classifier to identify the user's intent when asking questions.

[0096] The content entities in the user's question are extracted through the automatic prompt module to obtain the characteristic theme defined in the preset rule matching table. If there is a content entity belonging to the characteristic theme, the user's question includes the characteristic entity, and the corresponding answer is directly retrieved in the graph collection. Otherwise, a prompt text is generated to prompt the user to supplement the characteristic entity, and the answer is retrieved after the user has completed the supplement.

[0097] To sum up, in response to the existing problems, the highway traffic knowledge graph method invented this time establishes a graph collection suitable for the highway traffic field as a whole, which can monitor the highway traffic status in real time and respond quickly, and support intelligent analysis and intelligent question and answer functions; in traffic analysis, the road sections are innovatively abstracted as nodes, which can capture the influence dependencies of directly or indirectly connected sections, so as to more completely consider the chain effect of highway traffic congestion; in vehicle abnormality identification, only the distance threshold range of the lane line is defined as the area to be identified, which reduces the complexity of image recognition calculation; in the intelligent question and answer module, prompt text can be automatically generated to assist users in optimizing question statements and improving question and answer effects; the business knowledge graph is constructed in an integrated manner, which comprehensively considers the dynamic changes at the strategic level and actual conditions; the real-time data graph is updated in real time using a shard index method, which can assist the real-time data graph in achieving timely updates, and the data can be divided and distributed and stored on different nodes to disperse the load, making the system easier to expand horizontally.

[0098] Electronic device embodiment

[0099] Figure 6 6 is a schematic diagram of an electronic device according to an embodiment of the present invention. The electronic device 600 may include at least one processor 610 and a memory 620. The processor 610 may execute instructions stored in the memory 620. The processor 610 is connected to the memory 620 through a data bus. In addition to the memory 620, the processor 610 may also be connected to an input device 630, an output device 640, and a communication device 650 through a data bus.

[0100] The processor 610 may be any conventional processor, such as a commercially available CPU. The processor may also include a graphics processor (Graphic Process Unit, GPU), a field programmable gate array (Field Programmable Gate Array, FPGA), a system on chip (System on Chip, SOC), an application specific integrated circuit (Application Specific Integrated Circuit, ASIC) or a combination thereof.

[0101] The memory 620 may be implemented by any type of volatile or nonvolatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk.

[0102] In the embodiment of the present disclosure, executable instructions are stored in the memory 620, and the processor 610 can read the executable instructions from the memory 620 and execute the instructions to implement all or part of the steps of the high-speed traffic knowledge graph method of any of the above exemplary embodiments.

[0103] Computer Readable Storage Medium Embodiments

[0104] In addition to the above-mentioned methods and devices, an exemplary embodiment of the present disclosure may also be a computer program product or a computer-readable storage medium storing the computer program product, wherein the computer product includes computer program instructions that can be executed by a processor to implement all or part of the steps described in the high-speed traffic knowledge graph method of any of the above-mentioned exemplary embodiments.

[0105] The computer program product may be written in any combination of one or more programming languages ​​to write program codes for performing the operations of the embodiments of the present application, including object-oriented programming languages ​​such as Java, C++, etc., and also conventional procedural programming languages ​​such as "C" language or similar programming languages ​​and scripting languages ​​(e.g., Python). The program code may be executed entirely on the user computing device, partially on the user computing device, as an independent software package, partially on the user computing device and partially on a remote computing device, or entirely on a remote computing device or server.

[0106] Computer readable storage media can adopt any combination of one or more readable media. The readable medium can be a readable signal medium or a readable storage medium. The readable storage medium can include, for example, but is not limited to, a system, device or device of electricity, magnetism, light, electromagnetic, infrared, or semiconductor, or any combination of the above. More specific examples of readable storage media include: a static random access memory (SRAM) with one or more wires electrically connected, an electrically erasable programmable read-only memory (EEPROM), an erasable programmable read-only memory (EPROM), a programmable read-only memory (PROM), a read-only memory (ROM), a magnetic memory, a flash memory, a magnetic disk or an optical disk, or any suitable combination of the above.

[0107] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A high-speed transportation knowledge graph system, characterized in that: include: A graph construction module is used to construct a graph set including a domain knowledge graph, a real-time data graph and a business knowledge graph, and store the graph set in the background, wherein the business knowledge graph is constructed in an integrated manner; A graph maintenance module, used to update the knowledge graph in the graph collection, wherein the real-time data graph is updated in real time using a shard indexing method; An intelligent analysis module, used to analyze traffic conditions and obtain analysis results using a machine learning model based on the real-time data map stored in the background; The intelligent question-and-answer module is used to generate corresponding answers to user questions based on the graph set, extract the intention of the user's question after receiving the user's question from the front end, and determine whether the user's question includes feature entities based on the intention. If not, a prompt text is automatically generated.

2. The system according to claim 1, characterized in that The graph construction module specifically includes: A domain knowledge graph construction module is used to construct in a top-down manner, obtain the rules and common concepts in the field of high-speed transportation as a knowledge system, establish a first graph framework based on the knowledge system, fill in specific data into the first graph framework, and obtain a domain knowledge graph; A real-time data map construction module is used to construct in a bottom-up manner to obtain real-time data generated on high-speed traffic roads, wherein the real-time data includes toll station entrance and exit flow data, gantry vehicle flow data, emergency command data, asset information data, chartered vehicle order data, and hazardous cargo order data, to obtain a real-time data map; The business knowledge graph construction module is used to construct using an integrated approach including top-down and bottom-up methods, obtain high-speed transportation-related businesses and generate clustering results of the businesses using a clustering algorithm, calculate the average revenue contribution within each category based on the clustering results, select a category as a core business cluster based on the average revenue contribution, build a second graph framework based on the core business cluster, fill in the second graph framework and expand the graph using a bottom-up approach to obtain a business knowledge graph.

3. The system according to claim 1, characterized in that The graph maintenance module specifically includes: A direct update module, used to update the domain knowledge graph and the business knowledge graph; A shard index update module is used to update the real-time data map, divide the real-time data map into data slices and store them in corresponding nodes, maintain the copies of the data slices and the node storage locations in the standard retrieval library, first update the copies and then replace the original data slices at the corresponding node positions with the updated copies.

4. The system according to claim 1, characterized in that The intelligent analysis module specifically includes: A traffic analysis module is used to obtain the real-time data map and construct road segment features, obtain a pre-constructed highway traffic map structure, wherein the nodes of the highway traffic map structure represent road segments, and the edges of the highway traffic map structure represent the existence of a connection relationship, select dependent road segments corresponding to the road segment to be analyzed in the highway traffic map structure, fuse the road segment features of the road segment to be analyzed and the dependent road segment as a total feature, and input the total feature into a recursive neural network to obtain a traffic prediction result of the road segment to be analyzed; A first abnormality identification module is used to classify all vehicles in the current high-speed traffic into vehicle classes according to vehicle volume, obtain the real-time data map to construct clustering results of the vehicle classes, and regard vehicles with abnormal clustering results as first abnormal vehicles; The second abnormality recognition module is used to use the target detection model to identify vehicles changing lanes and interspersing, and regard the vehicle whose lane change and interspersing times exceeds a preset value as a second abnormal vehicle.

5. The system according to claim 4, characterized in that The traffic analysis module is specifically used for: Obtain the toll station entrance and exit flow data, gantry vehicle flow data, emergency command data, and asset information data in the real-time data map, and extract additional time features, gantry positions, section lengths, and average flow rates in the same historical time period as basic section features; Using Fourier transform to extract periodic features of traffic data, extracting historical events of road sections at different time scales, and splicing the periodic features and the historical events into auxiliary features of road sections; splicing the road section basic features and the road section auxiliary features into a road section feature; Use a pre-trained graph neural network to generate an embedding vector for each node, calculate the cosine similarity between the node represented by the section to be analyzed and the nodes within a preset number of hops based on the embedding vector, filter out the dependent sections based on the cosine similarity, concatenate the section features of the section to be analyzed and all dependent sections as the total features, and use a recursive neural network to obtain the traffic prediction result.

6. The system according to claim 4, characterized in that The second abnormality identification module is specifically used for: Acquire a high-speed traffic video and divide it into frames, reduce the resolution of the frames as pre-processed images, and the distance threshold range of the lane lines in the pre-processed images is the area to be identified; A lightweight target recognition model is used to identify lane changes of vehicles in the area to be identified. When the number of lane changes of a vehicle exceeds a threshold, a Kalman filter is used to track the lane changes of the vehicle, and a list of all filtered tracking is maintained.

7. The system according to claim 1, characterized in that The intelligent question-answering module specifically includes: Intent analysis module, used to identify the intent of user questions using naive Bayes classifier; The automatic prompt module is used to extract the content entities in the user's question and obtain the characteristic theme defined in the preset rule matching table. If there is a content entity belonging to the characteristic theme, the user's question includes the characteristic entity, and the corresponding answer is directly retrieved in the knowledge graph collection. Otherwise, a prompt text is generated to prompt the user to supplement the characteristic entity, and the answer is retrieved after the user has completed the supplement.

8. A high-speed transportation knowledge graph method, characterized in that: The high-speed transportation knowledge graph system used in any one of claims 1 to 7 comprises: A graph set including a domain knowledge graph, a real-time data graph and a business knowledge graph is constructed through a graph construction module, and the graph set is stored in the background, wherein the business knowledge graph is constructed in an integrated manner; The knowledge graph in the graph collection is updated through a graph maintenance module, wherein the real-time data graph is updated in real time using a shard indexing method; The intelligent analysis module uses a machine learning model to analyze traffic conditions based on the real-time data map stored in the background and obtain analysis results; The intelligent question-and-answer module generates corresponding answers to user questions based on the graph set, extracts the intention of the user's question after receiving the user's question at the front end, and determines whether the user's question includes feature entities based on the intention. If not, a prompt text is automatically generated.

9. An electronic device, characterized in that: include: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program, when executed by the processor, implements the steps of the high-speed traffic knowledge graph method as described in claim 8.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores an implementation program for information transmission, and when the program is executed by the processor, the steps of the high-speed traffic knowledge graph method as described in claim 8 are implemented.

Images