Rail transit dispatching simulation practical training and automatic examination system integrated with emergency assistance

Abstract

The invention discloses a rail transit dispatching simulation practical training and automatic examination system integrated with emergency assistance, and relates to the technical field of simulation practical training, and the system comprises a data collection module which is used for collecting basic data, operation data and historical emergency event data of a rail transit system; and the simulation environment construction module is used for constructing a high-fidelity rail transit simulation environment by applying a digital twin technology based on the data acquired by the data acquisition module, mapping a geometric field model, a physical field model, a behavior field model and a rule field model which are required by virtual equipment, constructing a four-model mapping rule base and fusing a multi-field model. According to the invention, the simulation environment construction module relies on a digital twinning technology and real acquired data, combines geometric, physical, behavioral and regular four-field models and a mapping rule base, constructs a high-fidelity simulation environment and meets diversified practical training requirements, and the emergency aid decision-making module introduces a quantum computing technology and can quickly process mass data in a complex scene.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of simulation training, in particular to a rail transit dispatching simulation training and automatic examination system integrated with emergency assistance. BACKGROUND

[0002] With the acceleration of urbanization, rail transit has become the core component of the urban public transportation system due to its large capacity, high efficiency and low pollution. The safe and stable operation of the rail transit system is directly related to the safety of millions of passengers and the smoothness of the urban transportation network. As the "central brain" of the system operation, the professional ability and emergency response level of the dispatch personnel are the key link to ensure the safety of rail transit operation. Therefore, it is a core requirement for the personnel training in the rail transit industry to carry out high-quality simulation training and scientific examination for rail transit dispatch personnel. Currently, rail transit dispatch training mainly relies on traditional training methods and primary simulation systems. However, these existing technical solutions have many problems to be solved in practical application, and it is difficult to meet the high standard requirements of modern rail transit dispatch personnel training. The specific deficiencies are as follows: the existing simulation systems mostly use simple geometric modeling or static scene simulation, and do not fully utilize advanced technologies such as digital twinning to build a high-fidelity simulation environment, resulting in the simulation environment failing to truly reproduce the operation rules of the rail transit system and the complexity of emergency scenarios, making it difficult for dispatch personnel to accumulate real operating experience in training; the training system can only simulate the occurrence of emergency events, but cannot provide real-time emergency response plan suggestions for training personnel, and is limited by the computing power bottleneck of traditional computing technology, making it difficult to truly reproduce the dynamic evolution process of emergency events when simulating complex emergency scenarios with multiple factors, resulting in a lack of effective decision support for dispatch personnel in complex emergency scenarios, and difficulty in improving emergency response capabilities. To solve this problem, we propose a rail transit dispatching simulation training and automatic examination system integrated with emergency assistance. SUMMARY

[0003] To solve the above technical problems, the rail transit dispatching simulation training and automatic examination system integrated with emergency assistance is provided, which solves the above problems.

[0004] To achieve the above purpose, the technical scheme adopted by the present application is as follows: the rail transit dispatching simulation training and automatic examination system integrated with emergency assistance, comprising: a data acquisition module for acquiring basic data, operation data and historical emergency event data of the rail transit system; The simulation environment construction module uses digital twin technology to construct a high-fidelity rail transit simulation environment based on the data obtained by the data acquisition module. The virtual device needs to use geometric field models, physical field models, behavior field models, and rule field models. The four models are mapped to a rule library, and multiple field models are fused to construct a realistic rail transit model for simulation training. The emergency auxiliary decision-making module introduces quantum computing technology to accelerate simulation operations in complex scenarios and provide real-time emergency response plan recommendations during the training process based on the simulation environment state. The training operation module allows trainees to perform dispatching operations in the simulation environment and records the operation process data. The assessment and evaluation module automatically scores based on pre-set assessment indicators and trainee operation process data. The data storage and analysis module is used to store system operation data, training records, and assessment results, and perform statistical analysis.

[0005] Preferably, the data acquisition module collects data from lines, stations, vehicles, equipment, and personnel. Line data is collected using GIS geographic information systems and laser radar scanning technology, including line direction, track type, track spacing, curve radius, slope, slope length, and turnout type. Station data is collected using a combination of building information modeling technology and high-definition camera shooting, including station building structure data, equipment layout data, functional partition data, and station staffing capacity parameters. Vehicle basic data is collected through vehicle manufacturer-provided technical manuals and OBD interfaces, including train model, body length, wheelbase, passenger capacity, traction system parameters, braking system parameters, bogie model, onboard equipment configuration, and vehicle maintenance cycle data. Device basic data includes signal systems, power supply systems, communication systems, and environmental control systems, which are collected through a combination of device management system interfaces and manual input. Personnel basic data includes job responsibilities, operating permissions, standardized work processes, and qualification requirements for personnel in various positions in the rail transit system.

[0006] Preferably, the simulation environment construction module has the following specific steps: Data preprocessing: classify and filter the collected basic and operational data, remove redundant and abnormal data, convert unstructured data into structured format, unify numerical units and label corresponding physical entities, and form a standardized data source. The single-field model is constructed to achieve entity mapping from four dimensions. The geometric field model relies on the line GIS, station BIM and equipment drawings to restore the spatial form with three-dimensional tools; the physical field model combines vehicle dynamics and track material parameters to simulate the physical characteristics of train operation and track vibration; the behavioral field model defines the behavioral logic of train operation and signal switching through programming; and the rule field model transforms the system operation and fault rules into algorithms and clarifies the interaction constraints. Establish a mapping rule base, define the correspondence rules between data and attributes within a single-field model, formulate rules for the linkage of multiple-field models, and build bridges for the connection between models; Multi-field model fusion is performed, four types of models are imported into the digital twin platform, data interaction is configured to achieve closed-loop flow, the scene is assembled and historical data is input for debugging, parameter deviations are corrected, and an initial simulation model is formed. Model verification and iteration: static verification checks geometric deviations through 3D comparison, dynamic verification compares the consistency between simulation and real operation data, adjusts parameters to address problems, and continuously optimizes by combining new data to obtain the rail transit model.

[0007] Preferably, the steps for constructing a single-field model are as follows: Import the data obtained from the data acquisition module, including station BIM model files and 3D drawings of vehicles and equipment, and parse the data to extract spatial parameters; Based on the GIS data of the railway line and the BIM data of the station, a 3D modeling tool was used to construct the geometric framework of the virtual railway line and station, and to restore the spatial characteristics of the track laying path, station floor height and entrance / exit locations. By combining vehicle exterior drawings and laser scanning data, a geometric model of the train is constructed, and details such as the body outline, number and location of doors, and window size are refined. The 3D drawings of the equipment are then imported. For trains, physical properties are set in the simulation engine based on vehicle dynamics parameters to simulate the centrifugal force physical phenomena when the train starts to accelerate, brakes and coasts, and passes through curves. For the station environment, combined with environmental data, physical parameters such as air resistance and heat conduction are set, and the behavioral logic of virtual entities is defined using state machines and script programming; Set the switching rules for red-yellow-green lights for the signal behavior model, associate it with ground control signal data, and simulate the behavior of the signal to adjust the display status according to the train position; We analyze the operational rules of the rail transit system, transform them into executable rule algorithms, and integrate equipment fault rules.

[0008] Preferably, the steps for constructing the associated bridge are as follows: Organize the associated data and attributes of each model to form a corresponding list; design rule expressions, including direct matching, data transformation and interval constraint types; select typical entity validation rules to correct deviations; Identify and classify multi-entity interaction scenarios in rail transit, determine the combination of linkage models, define the triggering conditions and response logic for each scenario, and present the data flow path using flowcharts; The rules for single-game and multi-game scenarios are categorized and stored in the database and assigned unique identifiers; the platform interface is then developed to enable real-time interaction between the model and the rule base; the complete running scenario is simulated for testing to form a stable rule base and build a bridge between models.

[0009] Preferably, the simulation calculation steps for the emergency auxiliary decision-making module are as follows: Quantum modeling of emergency scenarios is used to extract fault parameters from the simulation environment, map the parameters to qubits, construct a Hamiltonian containing a rescue efficiency term, and transform decision optimization into a quantum ground state solution problem. Quantum algorithm selection and deployment: QAOA algorithm is selected based on scenario complexity, adaptable quantum circuits are designed, and interfaces are developed to realize classical-quantum data interaction; Quantum parallel simulation computation initializes qubits to form a superposition state to characterize multiple schemes in parallel. Based on real-time scenarios, circuit parameters are dynamically adjusted to drive the evolution of quantum states. Through multiple measurement sampling, a high-probability optimal scheme is obtained, reducing the evaluation time from exponential to polynomial level. The quantum measurement results are decoded into specific handling parameters and restored into a solution. The classical algorithm is used for secondary evaluation and ranking. The top 3 candidate solutions are selected and pushed to the training interface. At the same time, the solution is recalculated every 30 seconds in combination with the scenario update, and the handling solution is dynamically iterated.

[0010] Preferably, the fault parameters include static and dynamic characteristics. Static characteristics include fault type, fault location, and fault range, while dynamic characteristics include fault evolution rate and duration. The mapping to qubits is achieved through quantum phase encoding, which maps continuous parameter values ​​to the phase of the quantum state. The core variables of emergency decision-making are clearly defined, and a Hamiltonian containing the decision objective is constructed. The emergency objective and constraints are transformed into energy terms, quantifying rescue efficiency. Higher efficiency corresponds to lower energy. The constraint penalty term adds additional energy penalties for decisions that violate constraints such as resource limits and time windows. The ground state is solved using quantum algorithms. The QAOA algorithm is used to construct a quantum superposition state with parameters on quantum hardware / simulator. The energy is calculated by substituting the Hamiltonian into the parameters. The parameters are iteratively optimized to gradually reduce the energy to convergence. The quantum state at this point is measured, and the resulting bit combination is the optimal emergency decision scheme.

[0011] Preferably, the training operation module adopts a three-column layout of main monitoring area + function operation area + information prompt area. The main monitoring area presents the real-time status of the simulation environment in a three-dimensional visualization form. The function operation area is divided into function modules according to role responsibilities. The information prompt area pushes information. The data of the operation process includes operation subject data, behavior data, object data and result data. Multi-dimensional tags are added to each piece of data before storage.

[0012] Preferably, the assessment steps for the evaluation module are as follows: The performance evaluation system is broken down into dimensions based on operational standardization, decision-making effectiveness, response timeliness, and resource utilization. Each dimension's indicators are quantitatively defined, weights are assigned, and additional deduction rules for serious operational errors are set to construct a quantitative performance evaluation framework. Relevant data is extracted from the training operation database, key fields are filtered, unstructured data is transformed into a quantifiable format, and the correspondence between data and indicators is established through anomaly verification and correction. The single-dimensional score is calculated according to the quantitative standard. The accuracy rate is calculated as (number of correct steps / total number of steps) × 100%. The accuracy rate is combined with the weight to calculate the comprehensive score. If there is a serious error, additional points will be deducted. The score is checked to see if it deviates from the normal range. Generate a structured report containing basic information, detailed dimensional scores, and key operational analysis; identify weaknesses and propose improvement suggestions; and archive the scoring data.

[0013] Preferably, the data storage and analysis module is categorized according to data characteristics. System operation data is stored in a time-series database, which supports querying by time and periodic cleanup. Training records are stored in a relational database and object storage, storing operation data and screenshots respectively. Assessment results are summarized in a relational database and a data warehouse, and the stored data is encrypted using dual encryption methods.

[0014] Compared with the prior art, the beneficial effects of the present invention are as follows: This invention proposes a simulation environment construction module that relies on digital twin technology and real-world data collection, combining geometric, physical, behavioral, and rule-based models and a mapping rule base to construct a high-fidelity simulation environment. This allows training to closely resemble real-world scenarios, enhancing operational realism, while avoiding the wear and tear, operational interference, and safety risks associated with relying on real equipment, significantly reducing training costs. It can also flexibly reproduce various operational and emergency scenarios to meet diverse training needs. The emergency decision support module incorporates quantum computing technology, enabling rapid processing of massive amounts of data in complex scenarios, solving the problem of low efficiency in traditional computing, and ensuring smooth simulation. Simultaneously, it provides real-time emergency response plans based on the simulation environment's state, offering professional guidance to trainees, especially beginners, reducing operational errors. Furthermore, the plans are generated based on objective conditions, avoiding biases from human experience and ensuring scientific decision-making. Attached Figure Description

[0015] Figure 1 This is a framework diagram of the automatic assessment system of the present invention. Detailed Implementation

[0016] The following description is intended to disclose the invention and enable those skilled in the art to implement it. The preferred embodiments described below are merely examples, and other obvious variations will occur to those skilled in the art.

[0017] Reference Figure 1 As shown, the integrated emergency-assisted rail transit dispatching simulation training and automatic assessment system includes: The data acquisition module is used to collect basic data, operational data, and historical emergency event data of the rail transit system. The simulation environment construction module, based on the data acquired by the data acquisition module, uses digital twin technology to build a high-fidelity rail transit simulation environment. Mapping virtual equipment requires geometric field models, physical field models, behavioral field models, and rule field models. It constructs a four-model mapping rule library, integrates multiple field models, and builds a realistic rail transit model for simulation training. The emergency decision support module incorporates quantum computing technology to accelerate simulation operations in complex scenarios and provides real-time emergency response plan suggestions based on the state of the simulation environment during the training process. The practical training module allows trainees to perform scheduling operations in a simulation environment and record the data of the operation process. The assessment module automatically scores trainees based on preset assessment indicators and their operational process data. The data storage and analysis module is used to store system operation data, training records, and assessment results, and to perform statistical analysis.

[0018] This application's data acquisition module provides a real and comprehensive data source, while the data storage and analysis module achieves data retention, analysis, and optimization, forming a data closed loop. The simulation environment construction module utilizes digital twins and four-field models to create highly realistic training scenarios that can reproduce diverse scenarios, solving the problem of the disconnect between theory and practice. The emergency auxiliary decision-making module improves computational efficiency through quantum computing and provides real-time solution suggestions, lowering the training threshold. The training operation module records operational data, and the assessment module automatically scores, avoiding subjective bias and helping to identify shortcomings. The system requires no physical equipment, avoiding safety risks, significantly reducing costs, and efficiently supporting the training of rail transit talents.

[0019] The data acquisition module collects data on lines, stations, vehicles, equipment, and personnel. Line data is collected using a GIS geographic information system and lidar scanning technology, including line alignment, track type, track spacing, curve radius, gradient, slope length, and turnout type. The station data is collected by combining building information modeling technology with high-definition camera shooting, including station building structure data, equipment layout data, functional zoning data and station passenger capacity parameters. Vehicle basic data is collected through the technical manuals provided by the vehicle manufacturer and the on-board diagnostic system (OBD) interface, including train model, body length, wheelbase, passenger capacity, traction system parameters, braking system parameters, bogie model, on-board equipment configuration, and vehicle maintenance cycle data. Basic equipment data includes signal system, power supply system, communication system and environmental control system, which is collected through a combination of equipment management system interface and manual data entry; Personnel basic data collection includes data on the division of responsibilities, operating authority, standardized operating procedures, and qualification requirements of personnel in various positions within the rail transit system.

[0020] This application employs adaptive technologies and methods for different targets to ensure comprehensive data coverage and high accuracy: Line data utilizes GIS and LiDAR to accurately acquire key parameters such as alignment, gradient, and turnout model; station data is collected comprehensively through BIM and high-definition cameras, capturing detailed architectural structures and equipment layouts; vehicle data combines manufacturer manuals and OBD interfaces, covering core information such as vehicle type and traction / braking parameters; equipment and personnel data are supplemented by system interfaces and manual input, ensuring the completeness of signal and power supply system parameters, personnel responsibilities, and operational procedures. This comprehensive and accurate data provides a vital source for building a high-fidelity training environment, preventing simulation scene distortion due to missing or biased data. Line data supports track operation scheduling training, station data assists in station operation process simulation, vehicle data meets vehicle operation and maintenance training, equipment data ensures signal and power supply system fault handling drills, and personnel data provides a basis for job responsibility division and standardized operation training. This precise correspondence between "data-scenario-job" makes the training content highly matched with actual work, helping trainees quickly master the core skills of their corresponding jobs.

[0021] The specific steps for building the simulation environment construction module are as follows: The collected basic data (line, station and vehicle parameters) and operational data (train speed and signal status) are classified and filtered to remove duplicate and redundant data and outliers that are outside the reasonable range, so as to avoid invalid information from interfering with the modeling. Unstructured data is format-converted, station images are converted into modelable image coordinates, and personnel operation text is decomposed into programmable nodes to ensure data compatibility with modeling tools; finally, numerical units are standardized, and corresponding physical entities are labeled for each set of data to form a structured data source. The single-field model construction achieves comprehensive entity mapping across four dimensions: geometry, physics, behavior, and rules. It is the core component of simulation scene restoration. The geometric field model relies on the line GIS data, station BIM models, and equipment drawings, using Unity 3D tools to restore the spatial form. From small details like gate locations to large changes in line gradient, everything is accurately matched to the real scene. The physics field model integrates vehicle dynamics parameters, track material properties, and environmental parameters, simulating physical characteristics through the PhysX physics engine. It simulates train acceleration based on traction power and calculates braking distance based on friction coefficient, ensuring the realism of physical phenomena. The behavior field model uses Python tools to programmatically define entity behavior logic, achieving automated behavior simulation. The rule-based field model transforms system operation and fault rules into algorithms, clarifying the entity interaction logic; The establishment of a mapping rule base is a key bridge connecting distributed models. It is divided into two layers of rules: rules within a single field and rules between multiple fields. Within a single field, the correspondence between data and attributes is defined, such as a gradient of 15‰ corresponding to the track tilt angle in the geometric field, and a braking pressure of 0.8MPa corresponding to the braking force output in the physical field. Linkage rules are formulated between multiple fields. For example, when a train reaches a switch, the rule field switching logic, geometric field state update, and physical field resistance parameter adjustment are triggered simultaneously to avoid data disconnect or logical conflicts between models and to achieve multi-dimensional collaboration. Multi-field model fusion relies on a digital twin platform to complete the initial simulation model construction. Four types of models are imported into the platform, and data interaction interfaces are configured to ensure that the physical field velocity data is synchronized to the behavioral field, and that changes in the behavioral field signal trigger the rule field judgment, thus realizing the closed-loop flow of data. The model is assembled according to actual scenarios (such as morning rush hour operation and equipment failure emergency) to form a complete virtual framework. Finally, historical data is input for debugging to observe whether the simulation logic conforms to the real situation, correct parameter deviations (such as train arrival time errors), and obtain an initial model that can be preliminarily run. Model validation and iteration ensure high-fidelity simulation through static and dynamic dual validation and continuous data updates. Static validation uses 3D comparison tools to compare the geometric field model with real-world photos to ensure that the spatial parameter deviation is less than 5%. Dynamic validation compares the simulation with real-world operating data (such as fault handling time), and if the deviation exceeds 10%, the model parameters or logic are adjusted retrospectively. At the same time, newly collected data (such as new line parameters and new equipment characteristics) are imported regularly to iterate the model, ensuring that the simulation environment is always synchronized with the real system and meets the needs of simulation training. The process is interconnected, realizing the transformation from data to a high-fidelity training environment.

[0022] The simulation calculation steps of the emergency auxiliary decision-making module are as follows: Quantum modeling of emergency scenarios is used to extract fault parameters from the simulation environment, map the parameters to qubits, construct a Hamiltonian containing a rescue efficiency term, and transform decision optimization into a quantum ground state solution problem. Quantum algorithm selection and deployment: QAOA algorithm is selected based on scenario complexity, adaptable quantum circuits are designed, and interfaces are developed to realize classical-quantum data interaction; Quantum parallel simulation computation initializes qubits to form a superposition state to characterize multiple schemes in parallel. Based on real-time scenarios, circuit parameters are dynamically adjusted to drive the evolution of quantum states. Through multiple measurement sampling, a high-probability optimal scheme is obtained, reducing the evaluation time from exponential to polynomial level. The quantum measurement results are decoded into specific handling parameters and restored into a solution. The classical algorithm is used for secondary evaluation and ranking. The top 3 candidate solutions are selected and pushed to the training interface. At the same time, the solution is recalculated every 30 seconds in combination with the scenario update, and the handling solution is dynamically iterated.

[0023] Fault parameters include static and dynamic characteristics. Static characteristics include fault type, fault location, and fault range, while dynamic characteristics include fault evolution rate and duration. The mapping to qubit steps is achieved through quantum phase encoding, mapping continuous parameter values ​​to the phase of the quantum state. The core variables for emergency decision-making are clearly defined, and a Hamiltonian containing the decision objective is constructed. Emergency objectives and constraints are transformed into energy terms, quantifying rescue efficiency; higher efficiency corresponds to lower energy. A constraint penalty term adds extra energy penalties for decisions that violate constraints such as resource limits and time windows. The ground state is solved using quantum algorithms. The QAOA algorithm is used to construct a quantum superposition state with parameters on quantum hardware / simulator. The energy is calculated by substituting the Hamiltonian into the parameters. The parameters are iteratively optimized to gradually reduce the energy to convergence. The quantum state at this point is measured, and the resulting bit combination is the optimal emergency decision scheme.

[0024] The core of the QAOA algorithm revolves around the cost Hamiltonian (Hamiltonian quantity). ) and mixed Hamiltonian ( The alternating evolution of the formulas is used to gradually approximate the optimal solution of the combinatorial optimization problem. The key formula system is as follows: The total evolutionary Hamiltonian is composed of a weighted average of the cost term and the mixed term, as shown in the formula: ; in The cost Hamiltonian weight parameter controls the strength of the approximation to the target optimization direction; To control the quantum state exploration capability by mixing Hamiltonian weight parameters and avoid getting trapped in local optima; To achieve the cost Hamiltonian, the emergency decision-making objective is encoded as a quantum mechanical operator; To determine the mixed Hamiltonian, take... , The Pauli-X operator is used to realize the superposition and transition of quantum states, thereby expanding the search range; The total evolutionary Hamiltonian; QAOA uses alternating applications and The evolution operator generates parameterized quantum states. The formula is: ; in The number of layers in QAOA is the highest level of accuracy in approximating the optimal solution. For quantum evolution operators, realize quantum states in Hamiltonian The evolution over time; For the initial quantum state ( ), The number of qubits corresponds to the variable dimension of the emergency decision-making problem; The objective function formula is: The expected value of the objective function is calculated by measuring the quantum state, and this value serves as the basis for parameter optimization. The formula is as follows:

[0025] in Given the expected cost value under the current parameters, the optimization objective is to adjust... ( )and ( To minimize this value, in actual calculations, multiple quantum measurement samples are needed to approximate the expected value using a statistical average. quantum state The left-hand vector form, used in quantum mechanics to describe the dual space representation of a quantum state, The parameterized quantum state is the quantum state obtained by alternately applying the evolution operator of the cost Hamiltonian and the mixed Hamiltonian; In emergency decision support for rail transit By combining fault types, such as signal faults, train derailment coding, rescue resource allocation, and path planning, the above formula can be used to achieve quantum-accelerated solutions for optimal handling schemes, significantly reducing computation time in complex scenarios.

[0026] The training operation module adopts a three-column layout of main monitoring area + function operation area + information prompt area. The main monitoring area presents the real-time status of the simulation environment in a three-dimensional visualization form. The function operation area is divided into function modules according to role responsibilities. The information prompt area pushes information. The data of the operation process includes operation subject data, behavior data, object data and result data. Multi-dimensional tags are added to each piece of data before storage.

[0027] The main monitoring area of ​​this application presents the real-time status of the simulation environment in a three-dimensional visualization format, allowing trainees to intuitively see the train operation, station equipment status, and other conditions, as if they were in a real rail transit scenario. This enhances the immersion and sense of involvement in the training, and helps trainees better understand and master the actual operation process. For example, when simulating train dispatching, trainees can clearly see the real-time position of the train on the line and changes in its running speed, thereby enabling them to perform dispatching operations more accurately.

[0028] The assessment steps for the evaluation module are as follows: The performance evaluation system is broken down into dimensions based on operational standardization, decision-making effectiveness, response timeliness, and resource utilization. Each dimension's indicators are quantitatively defined, weights are assigned, and additional deduction rules for serious operational errors are set to construct a quantitative performance evaluation framework. Relevant data is extracted from the training operation database, key fields are filtered, unstructured data is transformed into a quantifiable format, and the correspondence between data and indicators is established through anomaly verification and correction. The single-dimensional score is calculated according to the quantitative standard. The accuracy rate is calculated as (number of correct steps / total number of steps) × 100%. The accuracy rate is combined with the weight to calculate the comprehensive score. If there is a serious error, additional points will be deducted. The score is checked to see if it deviates from the normal range. Generate a structured report containing basic information, detailed dimensional scores, and key operational analysis; identify weaknesses and provide improvement suggestions; and archive the scoring data.

[0029] This application's quantitative assessment framework breaks down indicators into four dimensions, quantifies and defines them, and assigns weights. It also includes a serious error penalty rule to eliminate subjective bias and ensure the fairness and credibility of the assessment. Rigorous data processing filters key fields, transforms unstructured data, and corrects anomalies, establishing a correspondence between data and indicators. This ensures that scoring is based on real operational data, improving accuracy. Scientific scoring calculation first calculates the accuracy rate of each dimension, then combines it with weights to obtain a comprehensive score. Strict error penalty and scoring verification are implemented, taking into account both details and the overall picture to ensure reasonable results. Structured reports identify weaknesses and provide suggestions, and data archiving facilitates subsequent analysis, discovering common problems to optimize training courses, forming an "assessment-optimization" closed loop to maximize the value of the assessment.

[0030] The data storage and analysis module is categorized according to data characteristics. System operation data is stored in a time-series database, which supports querying by time and periodic cleanup. Training records are stored in a relational database and object storage, storing operation data and screenshots respectively. Assessment results are summarized in a relational database and a data warehouse, and the stored data is encrypted using dual encryption methods.

[0031] This application selects an appropriate storage solution based on the characteristics of different data types, avoiding the inefficiency caused by a "one-size-fits-all" approach to storage. System operation data (such as real-time train speed and equipment status) is characterized by high-frequency generation and time-series correlation. Using a time-series database for storage can optimize read and write performance in the time dimension, supporting fast queries by time period (such as "querying line operation data for the morning peak from 7:30 to 8:30"). It also periodically cleans up expired data (such as automatically deleting non-critical operation data older than 6 months) to reduce storage redundancy. The training records include structured operation data (such as operation steps and timestamps) and unstructured screenshot files, employing a "closed" approach. The combination of relational database and object storage enables efficient retrieval and correlation analysis of operational data (such as "correlation of trainee IDs and operation records") through relational databases, while object storage allows for flexible storage of large-capacity screenshots, balancing the needs of both structured and unstructured data management. In assessment results, detailed data (such as scores for each dimension and reasons for deductions) requires frequent querying and modification, so relational databases ensure data consistency. Summary data (such as average scores for each training session and rankings of scores for each position) needs to support multi-dimensional statistical analysis, so data warehouses are used to integrate the data, allowing data for different purposes to be properly organized, significantly improving management efficiency.

[0032] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the claimed invention. The scope of protection claimed by the appended claims and their equivalents is defined.

Claims

1. A rail transit dispatching simulation training and automatic examination system integrated with emergency assistance, characterized in that, Comprise: Data acquisition module for collecting basic data, operation data and historical emergency event data of rail transit system; Simulation environment construction module, based on the data obtained by the data acquisition module, uses digital twin technology to construct a high-fidelity rail transit simulation environment, maps the virtual device to the geometric field model, the physical field model, the behavior field model and the rule field model, constructs a four-model mapping rule library, fuses multiple field models, and constructs a realistic rail transit model for simulation training; Emergency decision support module, introduces quantum computing technology to accelerate simulation operations in complex scenarios, and provides real-time emergency disposal scheme suggestions according to the simulation environment state during the training process; Training operation module for trainees to perform dispatching operations in the simulation environment and record the operation process data; Assessment and evaluation module automatically scores according to the pre-set assessment indicators and operation process data of the trainees; Data storage and analysis module for storing system operation data, training records and assessment results, and conducting statistical analysis.

2. The integrated emergency assistance rail transit dispatch simulation training and automatic examination system according to claim 1, characterized in that: The data acquisition module collects data including lines, stations, vehicles, equipment and personnel. Line data is collected through GIS geographic information system and laser radar scanning technology, including line direction, track type, track spacing, curve radius, slope, slope length and turnout type. Station data is collected using building information modeling technology combined with high-definition camera shooting, including station building structure data, equipment layout data, functional partition data and station fixed staff capacity parameters. Vehicle basic data is collected through vehicle manufacturer's technical manual and OBD interface, including train model, body length, wheelbase, passenger capacity, traction system parameters, braking system parameters, bogie model, vehicle-mounted equipment configuration and vehicle maintenance cycle data. Equipment basic data includes signal system, power supply system, communication system and environmental control system, which is collected through device management system interface and manual input. Personnel basic data includes job responsibilities, operation permissions, standardized operation processes and qualification requirements of personnel in each post of the rail transit system.

3. The integrated emergency assistance rail transit dispatch simulation training and automatic examination system according to claim 1, characterized in that, The specific steps of the simulation environment construction module are as follows: Data preprocessing, classifying and filtering the collected basic and operation data, removing redundant and abnormal data, converting unstructured data into structured format, unifying numerical units and labeling corresponding physical entities to form a standardized data source; Single field model construction, entity mapping from four dimensions, geometric field model relies on line GIS, station BIM and equipment drawings to restore spatial form with three-dimensional tools; physical field model simulates train running and track vibration physical properties based on vehicle dynamics and track material parameters; behavior field model defines train operation and signal switching behavior logic through programming; rule field model converts system operation and fault rules into algorithms to clarify interaction constraints; Establish mapping rule library, define corresponding rules of data and attributes in single field model, develop multi-field model linkage rules, and build model interconnection bridge; Multi-scene model fusion is conducted, four types of models are imported into the digital twin platform, data interaction is configured to realize closed-loop circulation, scenes are assembled and historical data is input for debugging, parameter deviation is corrected, and an initial simulation model is formed; Model verification and iteration: static verification checks geometric deviation through three-dimensional comparison, dynamic verification compares simulation and real operation data for consistency, parameters are adjusted according to problems, and new data is continuously optimized to obtain the rail transit model.

4. The integrated emergency assistance rail transit dispatch simulation training and automatic examination system according to claim 3, characterized in that, The single-scene model construction steps are as follows: Import the data collected by the data acquisition module, including station BIM model files, vehicle and equipment three-dimensional drawings, and extract spatial parameters by analyzing the data; Based on line GIS data and station BIM data, use three-dimensional modeling tools to build the geometric framework of virtual lines and stations, restore the spatial characteristics of track laying paths, station floor heights, and entrance and exit positions; Combine vehicle appearance drawings and laser scanning data to construct train geometric models, refine vehicle body contours, door quantity and position, and window size details, and import equipment three-dimensional drawings; For trains, set physical properties in the simulation engine based on vehicle dynamics parameters to simulate train start-up acceleration, braking sliding, and centrifugal force physical phenomena when passing curves; For station environment, set air resistance and heat conduction physical parameters based on environmental data, and define the behavior logic of virtual entities using state machines and script programming; Set the switching rules of red light-yellow light-green light for the signal behavior model, associate with ground control signal data, and simulate the behavior of the signal adjusting the display state according to the train position; Sort out the operation rules of the rail transit system and convert them into executable rule algorithms, and integrate equipment fault rules.

5. The integrated emergency assistance rail transit dispatch simulation training and automatic examination system according to claim 3, characterized in that, The association bridge construction steps are as follows: Sort out the associated data and attributes of each model to form a corresponding list; design rule expressions, including direct matching, data conversion, and interval constraint types; Select typical entities to verify and correct rules; Identify and classify rail transit multi-entity interaction scenarios, determine the combination of linked models, define the trigger conditions and response logic of each scenario, and present the data flow path in a flowchart; Store single-scene and multi-scene rules in the database and assign unique identifiers; Realize real-time interaction between models and rule libraries through platform interfaces; Test the complete operation scenario simulation to form a stable rule library and build the association bridge between models.

6. The integrated emergency assistance rail transit dispatch simulation training and automatic examination system according to claim 1, characterized in that, The simulation operation steps of the emergency auxiliary decision-making module are as follows: Quantum modeling of emergency scenarios: extract fault parameters from the simulation environment, map the parameters to quantum bits, construct a Hamiltonian containing rescue efficiency terms, and convert decision optimization to quantum ground state solving problems; Quantum algorithm selection and deployment: select QAOA algorithm based on scene complexity, design adaptive quantum circuits, and develop interfaces to realize classical-quantum data interaction; Quantum parallel simulation operation: initialize quantum bits to form a superposition state to represent multiple schemes in parallel, dynamically adjust circuit parameters to drive quantum state evolution based on real-time scene, and obtain high-probability optimal solutions through multiple measurements and sampling, reducing evaluation time from exponential to polynomial level; The quantum measurement results are decoded into specific treatment parameters and restored to the scheme, and the top 3 candidate schemes are selected and pushed to the practical training interface through secondary evaluation and ranking with classical algorithms. At the same time, the treatment scheme is dynamically iterated and updated every 30 seconds.

7. The integrated emergency assistance rail transit dispatch simulation training and automatic examination system according to claim 6, characterized in that: The fault parameters include static and dynamic characteristics. The static characteristics include fault type, fault location and fault range. The dynamic characteristics include fault evolution rate and duration. The quantum bit steps are mapped to the phase of the quantum state through quantum phase encoding. The core variables of emergency decision-making are determined, and the Hamiltonian containing the decision-making objective is constructed. The emergency objective and constraints are converted into energy terms, and the rescue efficiency is quantified. The higher the efficiency, the lower the energy. The constraint penalty term increases the additional energy penalty for decisions that violate resource limits and time windows. The ground state is solved by quantum algorithm. Through QAOA algorithm, a quantum superposition state containing parameters is constructed on quantum hardware / simulator. The Hamiltonian is substituted to calculate the energy, and the parameters are iteratively optimized to gradually reduce the energy to convergence. The quantum state is measured at this time, and the obtained bit combination is the optimal emergency decision-making scheme.

8. The integrated emergency assistance rail transit dispatch simulation training and automatic examination system according to claim 1, characterized in that: The practical training operation module adopts a three-column layout of main monitoring area + functional operation area + information prompt area. The main monitoring area presents the real-time state of the simulation environment in a three-dimensional visual form. The functional operation area divides the functional modules according to the role and responsibility. The information prompt area pushes information. The data of the operation process includes operation subject data, behavior data, object data and result data. Each data is stored after adding multi-dimensional labels.

9. The integrated emergency assistance rail transit dispatch simulation training and automatic examination system according to claim 1, characterized in that, The examination and evaluation module has the following steps: The examination index system is disassembled, and the dimensions are divided according to operation standardization, decision effectiveness, response timeliness and resource utilization rate. The index of each dimension is quantitatively defined, weighted, and the additional deduction rule for serious operation errors is set to build a quantitative examination framework. Related data is extracted from the practical training operation database, key fields are selected, unstructured data is converted into quantifiable format, and the correspondence between data and indicators is established through abnormality checking and correction. Calculate the single-dimension score according to the quantification standard. Calculate the correct rate = correct step number / total step number × 100%. Calculate the comprehensive score by combining the weight of the correct rate. If there is a serious error, deduct additional points. Check if the score deviates from the normal range. Generate a structured report containing basic information, dimension score details and key operation analysis. Identify weak links and provide improvement suggestions. Archive the score data.

10. The integrated emergency assistance rail transit dispatch simulation training and automatic examination system according to claim 1, characterized in that: The data storage and analysis module is classified according to data characteristics. System running data is stored in time series database, supporting time query and regular cleaning. Practical training records are stored in relational database + object storage, storing operation data and screenshots respectively. The examination results are stored in relational database and data warehouse, and the stored data is encrypted through double encryption.

Images