Extra-large city comprehensive traffic data intelligent research and strategy system

Abstract

The invention provides an ultra-large city comprehensive traffic data and intelligent research and strategy system, which establishes a multi-dimensional traffic research and strategy task through a traffic research and strategy task module, and processes the traffic research and strategy task through a traffic characteristic diagnosis module to obtain a traffic characteristic index vector rank and a candidate traffic root cause index set. The traffic traceability analysis module performs further processing to obtain main traffic root cause indexes and measure indexes, and the traffic strategy generation module outputs traffic characteristic indexes according to the main traffic root cause indexes and the measure indexes to optimize a pre-generation strategy; the traffic simulation deduction module optimizes a simulation task item distribution model generated by a pre-generation strategy according to traffic characteristic indexes to obtain a pre-selection result, and the traffic strategy evaluation module and the traffic strategy release module finally perform evaluation and release, thereby solving a problem that a single factor outputs a single execution scheme with relatively high precision. Therefore, when single execution schemes are overlapped and fused into a set of optimal executable solutions for the comprehensive traffic of the super-large city, scheme execution conflicts are easy to occur.

Description

Technical Field

[0001] This invention relates to the fields of intelligent transportation and decision support systems, specifically to a digital and intelligent decision-making system for integrated transportation in megacities. Background Technology

[0002] With the development of urban transportation informatization, digitalization and intelligentization, the main research and application trend in various fields of urban transportation is to organically integrate new technologies, methods and tools such as advanced equipment, Internet of Things, big data, digital twins, parallel simulation, large language models, and AI intelligent agents into the comprehensive research, tracing, comparison and judgment of traffic laws, mechanisms, situations, events, trends and influencing factors. This is to better solve the problems of reducing management and service costs, improving service quality, and providing digital and intelligent tools and test fields for the entire process of urban transportation planning, construction, operation, maintenance and research.

[0003] Existing approaches often employ single new technologies, methods, and tools to study, trace, compare, and judge traffic patterns, mechanisms, situations, events, trends, and influencing factors. While this approach can solve certain traffic problems to some extent, it is prone to failure issues such as global errors and conflicting implementation when implementing continuous, overlapping, and integrated solutions for the best feasible complete solution for comprehensive transportation in megacities.

[0004] In conclusion, a more comprehensive technical solution needs to be designed, and an optimal and feasible complete solution for integrated transportation in megacities needs to be designed in a holistic manner to achieve innovative upgrades in the informatization, digitalization, and intelligentization of urban transportation. Summary of the Invention

[0005] The purpose of this invention is to design a comprehensive intelligent data analysis and decision-making system for megacities, containing a seven-node execution chain to automate the entire process of customized traffic tasks, analyze and judge the situation, and generate precise strategies. It shares a digital foundation and intelligent computing engine infrastructure, and deploys current data monitoring or business application systems in parallel. It acquires business requirements and publishes strategies, jointly achieving system-level online intelligent output of the best executable overall analysis, judgment, and decision support solutions for various sub-sectors of comprehensive transportation in megacities, including road traffic, rail transit, hub traffic, and static traffic, covering specific tasks such as planning, design, operation, renovation, and service. This addresses the problem that existing technologies only provide isolated solutions for single business functions, and when multiple solutions are superimposed and connected to form a complete, optimal, executable solution for the overall comprehensive transportation of megacities, conflicts and contradictions easily arise in the execution of solutions at different stages.

[0006] The present invention provides a comprehensive intelligent transportation system for megacities, comprising:

[0007] The Traffic Research and Planning Task Module is used to establish preset and specified traffic research and planning tasks for existing urban traffic entities in the traffic geographic information system or 2D / 3D traffic spatiotemporal models, and / or future digital research and planning tasks using the traffic planning blueprints, project design drawings, or traffic transformation drawings of roads, bridges, and hub stations to build GIS / BIM digital models. In the preset, customized, and future digital research and planning tasks, the following five vector members are used: the overall outer contour of the traffic object, the internal traffic sub-objects and their boundaries in 2D / 3D coordinate vectors, the start and end time of the research and planning task and the minimum analysis step size period vectors, the research and planning objectives such as the traffic flow increase ratio or the accident rate reduction vectors, the available spatiotemporal data of the data resources, the data of IoT sensing devices, and the shared access data vectors, and the constraint conditions such as the requirement that the traffic organization method cannot be modified and the signal control cycle vectors. These five vector members together constitute the traffic research and planning task object matrix.

[0008] The traffic feature diagnosis module is used to construct a comparative and intuitive traffic feature index library based on the data resources of the research and policy task. This library represents traffic efficiency, traffic safety, traffic energy consumption, low-carbon transportation, and service level. It also constructs a traffic root cause index library representing the fundamental causes of problems, such as road alignment, pavement materials, road grade, traffic capacity, traffic organization, weather factors, and policy measures. The module calculates all feature indicators for traffic research and policy tasks based on the data resource vector for a specified traffic object and time period. It outputs the calculated values ​​of each indicator, the positive and negative deviations within the threshold interval, the positive and negative deviation rates within the threshold interval, the positive and negative trends, the intensity of positive and negative changes, and the range and variance row vectors. It also correlates and calculates the sub-object members within the traffic object. The same traffic characteristic indicators of peripheral connected traffic objects are used to obtain a list of characteristic indicator results. The continuous changes and abrupt changes of indicator data in the traffic object space are analyzed to determine the traffic research and policy task and expand the scope of root cause indicator analysis. Based on knowledge or algorithms in the fields of traffic planning, traffic design, traffic engineering and traffic control, the functional causal relationship, logical causal relationship, fuzzy feature relationship and other causal relationship between traffic characteristic indicators and traffic root cause indicators are matched. A one-to-many causal knowledge graph between traffic characteristic indicators and possible traffic root cause indicators is established item by item. A knowledge matrix of traffic characteristic indicators mapping to traffic root cause indicators is generated. The traffic characteristic indicator vector list and the mapping traffic root cause indicator matrix are output according to the weight and / or positive and negative change range.

[0009] The traffic source tracing analysis module is used to construct a library of operable and executable traffic measures indicators representing traffic construction, traffic organization, traffic control, information services, and traffic guidance based on the traffic objects, research objectives, and constraint vectors of the research and policy tasks. It takes a list of traffic characteristic indicator vectors as input and performs single-feature principal component analysis, multi-factor weight analysis, factor influence analysis, and correlation analysis on the mapped traffic root cause indicator matrix. It analyzes the direct primary root causes, direct secondary root causes, indirect primary root causes, and indirect secondary root causes of a single characteristic indicator. It optimizes the weights of each root cause in the causal knowledge graph according to three levels: functional relationship, numerical connection, or logical association, and sorts them to obtain a sorted causal graph. The entire list of traffic characteristic indicators is then used to filter high-frequency root causes, merge similar root causes, and organize root causes. The data is arranged in a sequence and further optimized to obtain a root cause graph of traffic feature indicators and weighted traffic root cause indicators. The root causes are sorted according to direct, indirect, primary and secondary root causes. The root cause indicators are matched with the measure indicator library to obtain a one-to-many root cause measure graph. For each root cause indicator, a dynamic programming function for measure selection is established by introducing comprehensive constraint factors such as research and policy task constraints, optimal effect, lowest economic cost, shortest time, lowest complexity, best operability, and lowest correlation. The weight of the measure constraint factors is adjusted according to the research and policy objectives and constraints, and the optimal measure indicator vector is output. The knowledge graph of traffic root cause indicators and optimal traffic measure indicators is further sorted. Thus, a two-level knowledge graph data chain is formed from feature indicators to root cause indicators and from root cause indicators to measure indicators. The data is then output to the traffic strategy generation node module.

[0010] The traffic strategy generation module is used to generate a traffic strategy workflow based on the input of the sorted traffic root cause indicators and the optimized traffic measure indicators. With the goal of optimizing the research and development, and according to the upstream and downstream connection relationship and process design of the measure indicators, it fully arranges, extracts, merges and organizes the traffic measure indicators to generate the traffic measure indicator process, i.e., the strategy workflow. In a measure indicator process, the Cartesian product is used to adjust the comprehensive constraint factor and repeatedly call the optimized dynamic programming function to meet the goal of optimizing the traffic characteristic indicators in the research and development task, and maximizing the solution of direct primary root causes, direct secondary root causes and indirect primary root causes. The module generates and outputs the optimized preliminary traffic strategy set.

[0011] The traffic simulation and extrapolation module is used to conduct full-process digital experiments based on the input set of preliminary traffic strategies. It establishes multimodal simulation tasks for each preliminary traffic strategy, analyzes traffic characteristic index data after strategy execution in a digital twin environment, and creates simulation tasks with traffic research and development tasks as the primary primary label and traffic strategies as secondary sub-labels. The module allows users to edit input traffic strategies, start the simulation system, calculate and analyze traffic characteristic indicators, and output the optimal or optimal feasible solution for the research and development objectives. Simple strategies for a single traffic research and development task can be output in a single simulation using a single simulation software. Complex strategies for a single traffic research and development task, such as the continuous process analysis of traffic flow, parking, and passenger flow at traffic hubs, require a combination of passenger flow simulation software and traffic software, with simulations divided into segments according to passenger flow to vehicle flow and vehicle flow to passenger flow. The output of the first-segment simulation serves as the input for the second-segment simulation, which in turn provides the analysis results of traffic characteristic indicators. This leads to multimodal simulation and extrapolation tasks, including single-task single-simulation systems, single-task multi-simulation systems, multi-strategy multi-indicator single-simulation systems, and multi-strategy multi-indicator multi-simulation systems. The module obtains accurate traffic characteristic index data after quantitative analysis by simulation tools and outputs it to the traffic strategy evaluation module.

[0012] The traffic strategy evaluation module is used to analyze and recommend strategies based on accurate traffic feature index data. After importing real data to train the strategy, the simulation tool analysis output data is compared with the real data item by item. After optimization, a quasi-release strategy is obtained. The weights of traffic feature index items are adjusted multiple times and the traffic simulation simulation module is repeatedly executed to obtain the optimal solution or optimal feasible solution of the comprehensive impact factor and recommend the strategy. The release strategy is used and adjusted in actual combat to obtain the execution strategy. The system continuously accumulates and evaluates three types of strategy sets: pre-selected strategy data, release strategy data, and execution strategy data.

[0013] The traffic strategy release module is used to generate a strategy release report based on the recommended strategy data. It calls upon the traffic research and policy task vector set, feature indicators in traffic feature diagnosis, root cause indicators in traffic source tracing analysis, measure indicators in traffic strategy generation, simulation tools and simulation process in traffic simulation deduction, and quantitative analysis and adjustment process in traffic strategy evaluation. The traffic research and policy system outputs the released strategy to the business system through the digital platform.

[0014] Preferably, the traffic objects in the traffic research and planning task have their two-dimensional or three-dimensional GIS boundaries confirmed by the traffic system ontology geographic information system. They contain traffic sub-objects and external connecting objects. The time period includes historical time period selection, current real-time, and future predefined time. The research and planning objectives include levels such as improvement, problem solving, and predictive optimization, used to determine the operational boundaries and exit conditions of the research and planning system. The data resources include traffic object ontology data, perception data, process data, etc., that can be collected, monitored, and processed in the physical traffic environment, as well as data provided by information equipment that can be expected to be built, and simulation data obtained from the simulation and deduction module.

[0015] Preferably, the traffic business system and the traffic simulation module complete data collaboration and strategy release feedback through a digital and intelligent foundation to form the local integrated model.

[0016] Preferably, the traffic business system and the traffic simulation and deduction module are integrated in different locations through information and digital infrastructure to achieve data interconnection and complementary capabilities, forming a type 1 cross-regional integration model.

[0017] Preferably, the two local integration models are further integrated to form a remote integration model type 2.

[0018] This invention proposes a digital and intelligent research and policy system for integrated transportation in megacities, comprising seven sequential node modules. The system establishes multi-dimensional vector traffic research and policy tasks through a traffic research and policy task module. A traffic feature diagnosis module processes these tasks to obtain a ranking of traffic feature index vectors and a set of candidate traffic root cause indicators. A traffic source tracing analysis module further processes these tasks to obtain key traffic root cause indicators and corresponding measures. A traffic strategy generation module outputs a pre-generated strategy based on the key traffic root cause indicators and corresponding measures. A traffic simulation and deduction module generates a simulation task allocation model based on the pre-generated strategy and obtains pre-selected results. Finally, a traffic strategy evaluation module and a traffic strategy release module evaluate and release the strategy. This system addresses the problem that existing technologies only provide isolated solutions for single business functions. When multiple solutions are superimposed and connected throughout the process to form a complete and executable solution for the overall integrated transportation of megacities, contradictions and conflicts easily arise in the execution of solutions at different stages. Attached Figure Description

[0019] Figure 1 A flowchart of the research and development algorithm provided in an embodiment of the present invention;

[0020] Figure 2 This is a schematic diagram of the structure of the digital intelligence research and strategy system provided in an embodiment of the present invention;

[0021] Figure 3 A schematic diagram of the structure of a comprehensive intelligent transportation system for megacities provided in an embodiment of the present invention;

[0022] Figure 4 This is a schematic diagram of a local integration model, a remote integration model 1, or a remote integration model 2 in a comprehensive transportation digital research and development system for megacities, provided as an embodiment of the present invention.

[0023] Figure 5 This is a schematic diagram of the traffic simulation and deduction module type in the traffic digital research and policy system provided in an embodiment of the present invention; Detailed Implementation

[0024] The present invention will be further illustrated below with reference to specific embodiments. It should be understood that these embodiments are for illustrative purposes only and are not intended to limit the scope of the invention. Furthermore, it should be understood that after reading the teachings of this invention, those skilled in the art can make various alterations or modifications to the invention, and these equivalent forms also fall within the scope defined by the appended claims.

[0025] In this invention, the terms are explained as follows:

[0026] 1. For example Figure 1 As shown, the research and strategy system is a scientific thinking algorithm process that, for existing problems, improvement needs, or quality enhancements of single or complex objects, establishes research and strategy tasks by integrating expert experience, data analysis algorithm models, and other means. The system aims to complete the research and strategy tasks for the target objects by carrying out a complete process of task objectives, problem tracing, current situation analysis, simulation and deduction, selection and implementation, quantitative evaluation, scheme and strategy judgment and strategy comparison, and recommends scientific optimization of complete and executable strategies.

[0027] 2. The research and strategy task creation module, which is the first module of the research and strategy system, creates research and strategy tasks that include task objects and their corresponding task objectives and data resources. Specifically, the physical objects and time-space boundaries, or event flows, that are used to conduct research and strategy are defined, collectively referred to as task objects. Task objectives are established around these task objects, such as identifying problems, making improvements, and optimizing adjustments. The data resources are defined as the internal and external resources that are available and usable within and around the task objects.

[0028] 3. Feature diagnosis module as the second module of the research and strategy system: Through threshold matching and parameter comparison, a feature indicator system is established around the task objectives in the research and strategy task to reflect the superficial problems in a certain aspect, rank the severity of the superficial problems, identify the main links of the correlation, and clarify the situation or trend characteristics. The result is a feature indicator diagnosis set that gives the task objectives through weighting method, principal component analysis and other methods, including ranking and proportion.

[0029] 4. The source analysis module, the third module of the research and policy system, establishes a one-to-many root cause indicator system based on the input set of characteristic indicator diagnoses. This system reflects the origins, essence, and characteristics of the characteristic indicators. It analyzes and comprehensively traces the inherent causal or correlational relationships between characteristic indicators and root cause indicators, reflecting the ranking and primary causes of root cause indicators corresponding to the severity of the characteristic indicators. In the research and policy system, root cause indicators should possess one or more fundamental characteristics that are preventable, changeable, avoidable, or eliminateable. Indicators that cannot be scientifically defined, quantified, or for which measures cannot be taken cannot be considered root cause indicators.

[0030] 5. The strategy generation module, the fourth module of the research and development system, takes a one-to-many root cause indicator system as input, transforms the task objective into a characteristic indicator adjustment objective, and establishes a root cause indicator adjustment strategy with the goal of optimizing the characteristic indicators. The root cause indicators and optimization strategies have a one-to-many causal relationship. Based on constraints such as best effect, best cost, best timeliness, lowest complexity, and strongest operability, it reverse-engineers multiple executable strategies that can achieve or approximate the adjustment objective from the problem from effect to cause and from cause to strategy. Strategies that are obviously unreasonable and unexecutable are directly eliminated.

[0031] 6. The simulation and deduction module, the fifth module of the strategy development system, takes the set of executable strategies as the prior set. Based on the singleness, continuity, concurrency, and comprehensiveness of the strategies, it conducts virtual experiments of simulation and strategy deduction item by item. Obviously unreasonable strategies are directly eliminated, ultimately obtaining the optimal characteristic index data and analysis results for the set of executable strategies. The simulation outputs the experimental data range of characteristic indicators; the strategy deduction, based on various forms such as online systems, offline systems, and human-machine collaborative systems, conducts comprehensive research and analysis on the potential effects, risks, management, and costs of implementing a certain strategy using Delphi method, principal component analysis, and other methods.

[0032] 7. The strategy evaluation module, the sixth module of the research and development system, inputs root cause indicator adjustment strategies, optimal characteristic indicator data, and analysis results into the strategy evaluation stage to conduct pre-evaluation, quantitative evaluation, comparative evaluation, and post-evaluation to obtain the strategy evaluation results. Specifically, the pre-evaluation involves joint optimization of characteristic indicators and constraints, outputting the pre-release strategy of the research and development system. The quantitative evaluation evaluates the joint characteristic indicators based on measured or survey data and the pre-evaluation strategy scheme according to the pre-release strategy. The comparative evaluation adjusts the characteristic indicator evaluation, compares strategies for introducing new requirements, and compares subsequent implementation strategies with recommended strategies based on the optimal characteristic indicator data and analysis results. The post-evaluation comprehensively evaluates the characteristic indicators of the practical application strategies based on the adjustments made in the joint characteristic indicator evaluation and comparative evaluation, obtaining the strategy evaluation results. Practical application strategies include both prior-selected strategies and ad-hoc strategies.

[0033] 8. The strategy release module, which is the seventh module of the research and development system, inputs the best feature index data, analysis results, and strategy evaluation results as the quasi-release strategy. It verifies the strategy's service targets, strategy objectives, the comparison table of the original and improved feature indices, the basis for strategy selection and parameter comparison, including data processing and analysis of efficiency, cost, cycle, complexity, and difficulties, submits the complete strategy, and generates a strategy release report.

[0034] 9. For example Figure 2 As shown, the data-driven research and strategy system is built upon intelligence and data. Intelligence includes domain intelligence and artificial intelligence. Traditional domain intelligence and smartification, domain knowledge graphs, deep learning, and feature index calculation all fall under the category of intelligence and wisdom. When conducting research and strategy algorithms, small-scale models based on intelligence and domain intelligence technologies play a crucial role. Data includes two major categories: digitization and data. Digitization focuses on spatiotemporal data and object data as the backbone technologies for building future city-level digital spaces, primarily digital twin systems. Data focuses on domain ontology data (object data, master metadata, and graph relationship data), domain big data (spatiotemporal data, perception data, business data, strategy solutions, simulation data, data warehouses, governance data, and algorithm result data, etc.), and related big data systems.

[0035] The above technical solution is applied in the field of urban intelligent transportation. It involves designing the strategy for seven sequential node modules of a research and development system for traffic scenarios, and constructing a traffic digital intelligence research and development system based on a digital foundation and intelligent capability framework. Figure 3 As shown, the transportation digital intelligence research and policy development system constructs a closed-loop process for the entire industry research and policy development task, including research and policy development tasks, feature diagnosis, source analysis, strategy generation, simulation and deduction, strategy evaluation, and strategy release. Based on a digital foundation and intelligent analysis capabilities, it designs and provides a comprehensive transportation digital intelligence research and policy development system for megacities, including:

[0036] The Traffic Research and Planning Task Module is used to establish preset and specified traffic research and planning tasks for the existing physical traffic objects in the city's comprehensive traffic system, or for the future digital research and planning tasks that use the traffic geographic information system or two- or three-dimensional traffic spatiotemporal model to establish GIS / BIM digital models of roads, bridges, hub stations, project design drawings, or traffic transformation drawings.

[0037] The specific execution steps are as follows:

[0038] Step 1, as follows Figure 4 As shown, the system tracks traffic construction, traffic operation, and traffic management data in the business system. It transforms the needs for new construction, reconstruction, management, and optimization of road networks or traffic management requirements into research and development task requirements for traffic object characteristic indicators. The system generates and encodes research and development task objects according to subdivided fields.

[0039] Step 2: In the transportation geographic information system, delineate the overall outer contour of the transportation object, the internal transportation sub-objects, and the two / or three-dimensional coordinates of each boundary, and use them as the spatial vector of the transportation object to fill the first member vector of the research and planning task object.

[0040] Step 3: Determine the start and end times of the research and development task and the minimum analysis step size cycle, including the time vector of the historical cycle, the current implementation or the future cycle, and fill it into the second member vector of the research and development task object module as a member vector.

[0041] Step 4: The research and strategy objectives are to establish a complete analysis and judgment strategy based on the proportion of traffic flow improvement or the reduction of accident rate. The vague description of the objectives in the text is converted into multiple traffic feature index vectors, which are then used as member vectors to fill the third member vector of the research and strategy task object module.

[0042] Step 5: Obtain the data resources of the traffic object in the system, including available ontological spatiotemporal data, IoT sensing device data, and shared access data vector members, and fill them into the fourth member vector of the research and planning task object module as member vectors;

[0043] Step 6: Constraints such as the inability to modify traffic organization methods and signal control cycles are converted into constant values ​​of traffic root cause indicators or traffic measure indicators, and then filled into the fifth member vector of the research and policy task object module as a member vector.

[0044] Step 7: The traffic research and planning task object module completes instantiation, assignment, and coding system registration, and the object is output to the traffic feature diagnosis module.

[0045] The traffic feature diagnosis module is used to build a database of intuitive traffic feature indicators that represent comparative experiences such as traffic efficiency, traffic safety, traffic energy consumption, low-carbon traffic, and service level, based on research and policy task data resources. It also builds a database of traffic root cause indicators that represent the root causes of problems such as road alignment, pavement materials, road grade, traffic capacity, traffic organization, weather factors, and policy measures. The module generates and outputs a knowledge matrix that maps the full range of traffic feature indicators to traffic root cause indicators.

[0046] The specific execution steps are as follows:

[0047] Step 1: Construct a comprehensive transportation characteristic index database for megacities, which will be further subdivided into indicators for road traffic construction, operation management and service characteristics, public transportation network passenger flow and dispatch management, as well as indicators for traffic safety incidents, traffic energy consumption, traffic exhaust pollution and carbon peaking and carbon neutrality, etc. The traffic characteristic index database will be continuously adjusted and expanded as knowledge accumulates and data collection capabilities improve, and will be coded in a hierarchical manner according to field and scope of impact.

[0048] Step 2: Define the design threshold range and threshold range under working conditions for traffic characteristic indicators, and pre-compile the values ​​of each indicator, the positive and negative deviation of the threshold range, the positive and negative deviation rate of the threshold range, the positive and negative change trend, the positive and negative change intensity, and the range and variance calculation models and methods.

[0049] Step 3: Read the traffic research and policy task code and call the relevant sub-field traffic characteristic index calculation model and method. Based on the data resources, calculate the full traffic characteristic index in the traffic object space vector and time period vector and obtain the result list.

[0050] Step 4: Read the internal sub-object members of the traffic object and the external connected traffic objects, perform correlation calculations of the same traffic feature indicators, and obtain a list of feature indicator results.

[0051] Step 5: Analyze the continuous changes and abrupt changes in the indicator data in the space connecting traffic objects to determine the scope of root cause indicator analysis for traffic research and policy tasks.

[0052] Step 6: Based on knowledge or algorithms in traffic planning, traffic design, traffic engineering and traffic control, match the functional causal relationship, logical causal relationship, fuzzy feature relationship and other causal relationships between traffic feature indicators and traffic root cause indicators, establish a one-to-many causal knowledge graph between traffic feature indicators and possible traffic root cause indicators, and generate a knowledge matrix of all traffic feature indicators mapping to traffic root cause indicators.

[0053] Step 7: Sort and output a list of traffic feature index vectors and a mapped traffic root cause index matrix according to the index value weight and / or the positive and negative change range and the positive and negative change intensity, to complete the feature diagnosis and preliminary root cause analysis.

[0054] The traffic source analysis module is used to construct an operational and executable traffic measure indicator library based on the traffic objects, research objectives and constraint vectors of the research and policy task, traffic measures, traffic construction, traffic organization, traffic control, information services and traffic guidance, read traffic feature indicator vectors, complete the formation of a two-level knowledge graph data chain from feature indicators to root cause indicators and from root cause indicators to measure indicators, and output it to the strategy recommendation module.

[0055] The specific execution steps are as follows:

[0056] Step 1: Construct a comprehensive transportation planning, construction, operation, management, and service measures database for megacities. This database will be further subdivided into areas such as road traffic, rail transit, ground public transport, slow traffic, and static traffic. Traffic measures will be classified into multi-level subcategories based on major categories such as new construction, renovation, organization, control, service, and maintenance. The implementation process will be divided according to the major categories and multi-level subcategories. Traffic measure indicators will be continuously expanded as knowledge is accumulated and business expands.

[0057] Step 2: Create algorithmic models and methodologies for traffic root cause indicators, including principal component analysis, multi-factor weight analysis, factor influence factor analysis, and correlation analysis, combined with traffic domain classification coding.

[0058] Step 3: Input the list of traffic feature index vectors, perform calculations on the mapped traffic root cause index matrix, and output the list of traffic root causes.

[0059] Step 4: Analyze the direct primary root cause, direct secondary root cause, indirect primary root cause, and indirect secondary root cause of the single characteristic indicator in the traffic characteristic indicator list. Optimize the weights of each root cause in the causal knowledge graph according to the three levels of functional relationship, numerical connection, or logical association, and sort them to obtain the sorted causal graph.

[0060] Step 5: The list of all traffic feature indicators is further optimized by filtering high-frequency root causes, merging similar root causes, and organizing the time series of root causes, based on the single feature sorting causal graph. This results in a source root cause graph of traffic feature indicators and weighted traffic root cause indicators.

[0061] Step 6: Sort the root causes according to their direct and indirect primary and secondary causes, and obtain a one-to-many root cause and control map by sorting the root cause indicators and matching them with the indicator library.

[0062] Step 7: Introduce root cause indicators for each policy research task, and establish a dynamic programming function to optimize the measures by considering comprehensive constraint factors such as optimal effect, lowest economic cost, shortest time, lowest complexity, best operability, and lowest impact.

[0063] Step 8: Adjust the weights of the constraint factors of the measures according to the research and policy objectives and constraints, and output the vector of the preferred measures index to further optimize and rank the knowledge graph of traffic root cause indicators and preferred traffic measures indicators.

[0064] Step 9: Connect feature indicators to root cause indicators, and root cause indicators to measure indicators to form a two-level knowledge graph data chain, which is then output to the traffic strategy generation node module.

[0065] The traffic strategy generation module is used to generate and output a set of optimized preliminary traffic strategies based on the input of the sorted traffic root cause indicators and the optimized traffic measure indicator map.

[0066] The specific execution steps are as follows:

[0067] Step 1: Read the two-level knowledge graph, optimize the research and development objectives, and arrange, extract, merge, and organize all traffic measure indicators according to the feasible upstream and downstream connections and process design of the measures indicators. This generates the traffic measure indicator process, i.e., the strategy workflow.

[0068] Step 2: In the single-measure indicator process, Cartesian product is used to adjust the comprehensive constraint factor and the optimal dynamic programming function is repeatedly called to obtain the maximum solution from the root cause.

[0069] Step 3: Using traffic characteristic indicators to respond to traffic research and policy objectives as constraints, further optimize the traffic measure indicator process to obtain preliminary traffic strategies.

[0070] Step 4: Generate and output the optimized set of preliminary traffic strategies;

[0071] The traffic simulation and deduction module is used to conduct a full-process digital experiment based on the input set of preliminary traffic strategies, establish multimodal simulation task items for each preliminary traffic strategy, and output them to the traffic strategy evaluation module.

[0072] The specific execution steps are as follows:

[0073] Step 1: Integrate multiple simulation platform tools, plan the data input and output protocol of the simulation system based on the digital foundation, and establish a multimodal simulation and deduction working mode of single task single simulation system, single task multiple simulation system, multi-strategy multi-index single simulation system, and multi-strategy multi-index multi-simulation system.

[0074] Step 2: Analyze traffic characteristic index data after strategy execution in the digital twin environment, and create simulation tasks with traffic research and policy tasks as the primary primary label and traffic strategies as the secondary sub-labels.

[0075] Step 3: Edit and input the traffic strategy, start the simulation system, calculate and analyze traffic characteristic indicators, and output the optimal solution or optimal feasible solution for the strategy objective.

[0076] Step 4 Figure 5 As shown, a simple strategy for a single traffic research task can output results in a single simulation using a single simulation software.

[0077] Step 5 Figure 5 As shown, complex strategies for a single traffic research task, such as the continuous process analysis of traffic flow, parking, and passenger flow at traffic hubs, require the use of a combination of passenger flow simulation software and traffic software to perform segmented simulations according to the sequence of passenger flow to vehicle flow and vehicle flow to passenger flow. The output of the first stage of simulation serves as the input for the second stage of simulation, and the second stage of simulation provides the analysis results of traffic characteristic indicators.

[0078] Step 6: Obtain accurate traffic characteristic index data after quantitative analysis by simulation tools, and output it to the traffic strategy evaluation module.

[0079] The traffic strategy evaluation module is used to analyze and recommend strategies based on accurate traffic characteristic index data;

[0080] The specific execution steps are as follows:

[0081] Step 1: After importing real data to train the strategy, obtain the comparison values ​​of each indicator between the simulation tool analysis output data and the real data, and then optimize to obtain the quasi-release strategy.

[0082] Step 2: Adjust the weights of traffic characteristic indicators multiple times and repeatedly execute the traffic simulation module to obtain the optimal solution or the optimal feasible solution recommendation strategy for the comprehensive impact factor;

[0083] Step 3: Deploy the strategy. The execution strategy is obtained through practical use and adjustment.

[0084] Step 4: The system continuously accumulates and evaluates three types of strategy sets: pre-selected strategy data, release strategy data, and execution strategy data.

[0085] The traffic strategy release module is used to generate a strategy release report based on recommended strategy data, by invoking traffic research and policy task vector sets, feature indicators from traffic feature diagnosis, root cause indicators from traffic source analysis, measure indicators from traffic strategy generation, simulation tools and processes from traffic simulation simulation, and quantitative analysis and adjustment processes from traffic strategy evaluation. Figure 4 As shown, the traffic research and policy system outputs and publishes policies to the business system via a digital platform.

Claims

1. A comprehensive intelligent transportation system for megacities, characterized in that: include: The Traffic Research and Planning Task Module is used to establish preset and specified traffic research and planning tasks for existing urban traffic entities in the traffic geographic information system or 2D / 3D traffic spatiotemporal models, and / or future digital research and planning tasks using the traffic planning blueprints, project design drawings, or traffic transformation drawings of roads, bridges, and hub stations to build GIS / BIM digital models. In the preset, customized, and future digital research and planning tasks, the following five vector members are used: the overall outer contour of the traffic object, the internal traffic sub-objects and their boundaries in 2D / 3D coordinate vectors, the start and end time of the research and planning task and the minimum analysis step size period vectors, the research and planning objectives such as the traffic flow increase ratio or the accident rate reduction vectors, the available spatiotemporal data of the data resources, the data of IoT sensing devices, and the shared access data vectors, and the constraint conditions such as the requirement that the traffic organization method cannot be modified and the signal control cycle vectors. These five vector members together constitute the traffic research and planning task object matrix. The traffic feature diagnosis module is used to construct a comparative and intuitive traffic feature index library based on the data resources of the research and policy task. This library represents traffic efficiency, traffic safety, traffic energy consumption, low-carbon transportation, and service level. It also constructs a traffic root cause index library representing the fundamental causes of problems, such as road alignment, pavement materials, road grade, traffic capacity, traffic organization, weather factors, and policy measures. The module calculates all feature indicators for traffic research and policy tasks based on the data resource vector for a specified traffic object and time period. It outputs the calculated values ​​of each indicator, the positive and negative deviations within the threshold interval, the positive and negative deviation rates within the threshold interval, the positive and negative trends, the intensity of positive and negative changes, and the range and variance row vectors. It also correlates and calculates the sub-object members within the traffic object. The same traffic characteristic indicators of peripheral connected traffic objects are used to obtain a list of characteristic indicator results. The continuous changes and abrupt changes of indicator data in the traffic object space are analyzed to determine the traffic research and policy task and expand the scope of root cause indicator analysis. Based on knowledge or algorithms in the fields of traffic planning, traffic design, traffic engineering and traffic control, the functional causal relationship, logical causal relationship, fuzzy feature relationship and other causal relationship between traffic characteristic indicators and traffic root cause indicators are matched. A one-to-many causal knowledge graph between traffic characteristic indicators and possible traffic root cause indicators is established item by item. A knowledge matrix of traffic characteristic indicators mapping to traffic root cause indicators is generated. The traffic characteristic indicator vector list and the mapping traffic root cause indicator matrix are output according to the weight and / or positive and negative change range. The traffic source tracing analysis module is used to construct a library of operable and executable traffic measures indicators representing traffic construction, traffic organization, traffic control, information services, and traffic guidance based on the traffic objects, research objectives, and constraint vectors of the research and policy tasks. It takes a list of traffic characteristic indicator vectors as input and performs single-feature principal component analysis, multi-factor weight analysis, factor influence analysis, and correlation analysis on the mapped traffic root cause indicator matrix. It analyzes the direct primary root causes, direct secondary root causes, indirect primary root causes, and indirect secondary root causes of a single characteristic indicator. It optimizes the weights of each root cause in the causal knowledge graph according to three levels: functional relationship, numerical connection, or logical association, and sorts them to obtain a sorted causal graph. The entire list of traffic characteristic indicators is then used to filter high-frequency root causes, merge similar root causes, and organize root causes. The data is arranged in a sequence and further optimized to obtain a root cause graph of traffic feature indicators and weighted traffic root cause indicators. The root causes are sorted according to direct, indirect, primary and secondary root causes. The root cause indicators are matched with the measure indicator library to obtain a one-to-many root cause measure graph. For each root cause indicator, a dynamic programming function for measure selection is established by introducing comprehensive constraint factors such as research and policy task constraints, optimal effect, lowest economic cost, shortest time, lowest complexity, best operability, and lowest correlation. The weight of the measure constraint factors is adjusted according to the research and policy objectives and constraints, and the optimal measure indicator vector is output. The knowledge graph of traffic root cause indicators and optimal traffic measure indicators is further sorted. Thus, a two-level knowledge graph data chain is formed from feature indicators to root cause indicators and from root cause indicators to measure indicators. The data is then output to the traffic strategy generation node module. The traffic strategy generation module is used to generate a traffic strategy workflow based on the input of the sorted traffic root cause indicators and the optimized traffic measure indicators. With the goal of optimizing the research and development, and according to the upstream and downstream connection relationship and process design of the measure indicators, it fully arranges, extracts, merges and organizes the traffic measure indicators to generate the traffic measure indicator process, i.e., the strategy workflow. In a measure indicator process, the Cartesian product is used to adjust the comprehensive constraint factor and repeatedly call the optimized dynamic programming function to meet the goal of optimizing the traffic characteristic indicators in the research and development task, and maximizing the solution of direct primary root causes, direct secondary root causes and indirect primary root causes. The module generates and outputs the optimized preliminary traffic strategy set. The traffic simulation and extrapolation module is used to conduct full-process digital experiments based on the input set of preliminary traffic strategies. It establishes multimodal simulation tasks for each preliminary traffic strategy, analyzes traffic characteristic index data after strategy execution in a digital twin environment, and creates simulation tasks with traffic research and development tasks as the primary primary label and traffic strategies as secondary sub-labels. The module allows users to edit input traffic strategies, start the simulation system, calculate and analyze traffic characteristic indicators, and output the optimal or optimal feasible solution for the research and development objectives. Simple strategies for a single traffic research and development task can be output in a single simulation using a single simulation software. Complex strategies for a single traffic research and development task, such as the continuous process analysis of traffic flow, parking, and passenger flow at traffic hubs, require a combination of passenger flow simulation software and traffic software, with simulations divided into segments according to passenger flow to vehicle flow and vehicle flow to passenger flow. The output of the first-segment simulation serves as the input for the second-segment simulation, which in turn provides the analysis results of traffic characteristic indicators. This leads to multimodal simulation and extrapolation tasks, including single-task single-simulation systems, single-task multi-simulation systems, multi-strategy multi-indicator single-simulation systems, and multi-strategy multi-indicator multi-simulation systems. The module obtains accurate traffic characteristic index data after quantitative analysis by simulation tools and outputs it to the traffic strategy evaluation module. The traffic strategy evaluation module is used to analyze and recommend strategies based on accurate traffic feature index data. After importing real data to train the strategy, the simulation tool analysis output data is compared with the real data item by item. After optimization, a quasi-release strategy is obtained. The weights of traffic feature index items are adjusted multiple times and the traffic simulation simulation module is repeatedly executed to obtain the optimal solution or optimal feasible solution of the comprehensive impact factor and recommend the strategy. The release strategy is used and adjusted in actual combat to obtain the execution strategy. The system continuously accumulates and evaluates three types of strategy sets: pre-selected strategy data, release strategy data, and execution strategy data. The traffic strategy release module is used to generate a strategy release report based on the recommended strategy data. It calls upon the traffic research and policy task vector set, feature indicators in traffic feature diagnosis, root cause indicators in traffic source tracing analysis, measure indicators in traffic strategy generation, simulation tools and simulation process in traffic simulation deduction, and quantitative analysis and adjustment process in traffic strategy evaluation. The traffic research and policy system outputs the released strategy to the business system through the digital platform.

2. The intelligent digital research and development system for integrated transportation in megacities as described in claim 1, characterized in that, The spatial objects include two-dimensional and three-dimensional GIS boundaries, containing traffic objects and external connecting objects. The time period includes historical time selection, current real-time, and future predefined time. The research and development objectives include levels such as improvement, problem solving, and predictive optimization, used to determine the operational boundaries and exit conditions. The data conditions include collectable, monitorable, and processable data resources and business processes in the physical traffic environment, as well as data provided by information technology equipment that can be expected to be built, and simulation data obtained from the simulation and deduction module.

3. The intelligent digital research and development system for integrated transportation in megacities as described in claim 1, characterized in that, The traffic business system and the traffic simulation module achieve data collaboration and strategy release feedback through an information and digital foundation, forming the local integrated model.

4. The intelligent digital research and development system for integrated transportation in megacities as described in claim 1, characterized in that, The traffic business system and the traffic simulation module are integrated in different locations through information and digital infrastructure to achieve data interconnection and complementary capabilities, forming a type 1 remote integration model.

5. The intelligent digital research and development system for integrated transportation in megacities as described in claim 3, characterized in that, The two local integration models are further integrated to form the off-site integration model type 2.

Images