Rail transit star chain system based on digital twinning and communication scheme design optimization method

Abstract

The invention relates to a digital twinning-based rail transit star chain system and a corresponding communication scheme design optimization method. The system comprises a sensing network, a station center digital twin server cluster system, a station center supervision system and corresponding intelligent service businesses, wherein the sensing network covers rail transit star chain systems of trains, trackside infrastructures, station centers and the like in an omnibearing manner. According to the system and the method, all parameters of an urban rail transit system based on an orbital star chain in an actual physical domain are mapped to a virtual digital domain system, so that all-directional supervision of a full-connection system, effective utilization of multi-source data characteristics and short-time accurate modeling of a channel environment with variable position environment parameters are realized; a communication link selection strategy model is constructed in combination with a model-based reinforcement learning method, and loop optimization is realized and an optimal communication link decision scheme is finally obtained by continuously training the model in a digital domain and collecting samples and verifying the model in a physical domain.

Description

Technical Field

[0001] The present invention relates to the field of rail transit technology, and specifically to a rail transit Starlink system based on digital twins and a corresponding communication scheme design optimization method. Background Art

[0002] As the core technology supporting the Train Autonomous Operation System (TACS), vehicle-to-vehicle communication is crucial to promoting the intelligent development of rail transit, shortening train operation intervals, and ensuring driving safety. However, since the traditional rail transit train-to-ground communication system adopts a static networking method and relies on fixed infrastructure such as ground core network equipment and base station subsystems, the existing line construction cycle is long, maintenance is difficult, the fault tolerance rate is low, the self-healing anti-destruction disaster tolerance is poor, the compatibility is poor, and the upgrade and maintenance cost is high. Many defects make it difficult to meet the development needs of the Train Autonomous Operation System (TACS).

[0003] The Starlink self-organizing communication system supports flexible self-organizing networks and dynamic addition and deletion of network equipment. The network has disaster recovery and self-healing functions and supports intelligent selection of the optimal communication path. It has many advantages such as rapid deployment, flexible networking, self-healing, strong anti-destruction and disaster recovery, short construction period, low maintenance cost and low difficulty of trackside equipment. It can ensure high reliability and large-capacity communication between vehicles, vehicles and ground, and ground and air, while effectively solving the difficulties and pain points faced by rail transit systems based on vehicle-to-vehicle communication, such as high construction difficulty, high maintenance cost, and poor disaster recovery.

[0004] In the Starlink-based rail transit system, there are multiple communication mechanisms and available communication solutions. In the complex rail transit environment, the optimal communication link selection solution is needed to ensure the safety and stability of the system. However, due to the complex state parameters and unknown environmental models in the Starlink-based rail transit system, there is currently no mature and feasible communication link selection technology solution for the Starlink-based rail transit system. Summary of the invention

[0005] The present invention provides a rail transit Starlink system and a communication scheme design optimization method based on digital twins to overcome the problems of the prior art.

[0006] In order to achieve the above-mentioned purpose, the present invention adopts the following technical solutions and technical methods.

[0007] A rail transit Starlink system based on digital twins, characterized in that the rail transit Starlink system comprises: an on-board sensor system 1 and an on-board multi-mode communication module installed on each train, a rail Starlink sensor system 2 and a trackside multi-mode communication module installed in each trackside Starlink device, ground communication infrastructure, a station center sensor system 3, a station center digital twin server cluster system 5 and a station center supervision system 6;

[0008] The on-board sensor system 1 is used to collect train system status information and upload the above train system status information to the station center digital twin server cluster system 5 through the on-board multi-mode communication module and the ground communication infrastructure; the train system status information includes train speed, train equipment status, passenger load factor, train safety status and train position;

[0009] The track starlink sensing system 2 is used to collect trackside line status information, and upload the trackside line status information to the station center digital twin server cluster system 5 through the trackside multi-mode communication module and the ground communication infrastructure; the trackside line status information includes trackside equipment temperature, trackside equipment current, track vibration, trackside switch information and track obstacle information;

[0010] The station center sensor system 3 is used to collect the status data of the station system and the planning data of the dispatching center, and upload it to the station center digital twin server cluster system 5 through the station center private network communication transceiver device 43;

[0011] The above-mentioned vehicle-mounted multi-mode communication module, trackside multi-mode communication module and station center private network communication transceiver equipment constitute a dedicated sensor communication network 4;

[0012] The station center digital twin server cluster system 5 is used to collect the status information and operation data of the entire rail transit Starlink system, and based on the digital domain-real domain structure, map the actual rail transit Starlink system into a digital twin rail transit Starlink system in the virtual digital domain, which is further applied to system intrusion detection, system equipment fault diagnosis and system communication optimization.

[0013] On the basis of the above scheme,

[0014] The vehicle-mounted sensing system 1 is installed on the train and includes a millimeter wave radar 11, a secondary protection radar 12, an image sensor 13, and a door status sensor 14;

[0015] The vehicle-mounted multi-mode communication module is used to: establish a direct communication link based on the PC5 interface with adjacent trains to achieve vehicle-to-vehicle communication; establish a direct communication link based on the PC5 interface or the Uu interface with nearby trackside Starlink equipment; establish a communication link based on the Uu interface with the communication infrastructure; the above-mentioned Uu interface is divided into a public network Uu interface and a private network Uu interface;

[0016] The track star chain sensing system 2 includes a trackside equipment acquisition module 21, an image sensor 22 and a vibration sensor 23;

[0017] The trackside multi-mode communication module is used to: establish a direct communication link based on the PC5 interface or Uu interface with each nearby train; establish a direct communication link based on the PC5 interface with the adjacent trackside Starlink device;

[0018] The station center sensor system 3 includes a dispatch center server 31 and a subsystem server 32;

[0019] The ground communication infrastructure is a ground base station, and the ground communication infrastructure adopts 5G, LTE or WLAN networking mode, and the networking mode includes public network 5G, private network 5G, private network LTE, private network WLAN and public network LTE;

[0020] The station center digital twin server cluster system 5 includes multiple secure cloud servers, a server cluster system and multiple digital domain-real domain interfaces;

[0021] The station center supervision system 6 includes multiple local hosts and display devices; the station center supervision system 6 is used to visualize the digital twin rail transit Starlink system and provide supervision services.

[0022] An application of a rail transit Starlink system based on digital twins, characterized in that the application is fault diagnosis of trackside equipment or on-board equipment, specifically comprising the following steps:

[0023] Step 1: Transmit the trackside line status information collected by the track starlink sensor system 2 or the train system status information collected by the vehicle-mounted sensor system 1 to the station digital twin server cluster system 5;

[0024] Step 2: Compare the trackside line status information and train system status information collected in step 1 with the actual environment model data in the station digital twin server cluster system 5 to quickly identify abnormal behavior of the trackside Starlink equipment or abnormal behavior of the on-board equipment.

[0025] A design optimization method for rail transit Starlink system and communication solution based on digital twin, the design ideas are detailed as follows:

[0026] The digital twin-based rail transit Starlink system is used to collect system parameter data and perceive channel environment of trains, trackside and ground centers. Based on the real rail transit Starlink system, a virtual simulation system highly consistent with the real scene is established in the station center, which can be used for communication optimization, fault diagnosis, danger forecasting and other services.

[0027] The communication scheme design and optimization method described is based on the digital twin rail transit Starlink system. First, based on the sample data of the sensor system during the train operation, and secondly, the actual environment model is trained in the digital twin server cluster system of the station center. Then, the optimal communication link selection and switching scheme is trained based on the environmental model, business characteristics, etc. After a period of time, the communication link selection scheme of the actual system is updated. Finally, the algorithm performance is evaluated based on the performance of the actual system.

[0028] The digital twin-based rail transit Starlink system includes a dedicated sensor communication network, an on-board sensor system, a rail Starlink sensor system, a station center sensor system, a station center digital twin server cluster system, a station center supervision system, and service algorithms for different businesses;

[0029] The vehicle-mounted terminal comprises:

[0030] Dedicated sensor communication network, including on-board private network communication transceiver equipment, trackside private network communication transceiver equipment, station center private network communication transceiver equipment, etc., based on the traditional 5G, WLAN or LTE communication of urban rail transit, is used to upload the status information of trains, trackside, station centers and collected channel information to the station center digital twin server cluster system;

[0031] The on-board sensor system, including millimeter-wave radar, secondary protection radar, image sensor, door status sensor and other sensors, collects train system status information from multiple aspects, including train speed, train equipment status, passenger load factor, train safety status, train location, etc.

[0032] The rail Starlink sensing system includes multiple modules such as trackside equipment acquisition modules, image sensors, vibration sensors, etc., which collect trackside line status information such as trackside switches and track obstacles and send it to the station center digital twin server cluster system;

[0033] The station center sensing system includes a server dedicated to collecting the status of dispatch center equipment and systems and other subsystems, and designs a secure interface to upload the status data of the station system and the planning data of the dispatch center to the station center digital twin server cluster system;

[0034] The digital twin server cluster system of the station center includes multiple secure cloud servers, which build a server cluster system and multiple digital domain-real domain interfaces to collect the status information and operation data of the entire system in real time. Based on various digital domain-real domain interfaces, the real orbit Starlink system is mapped to the simulation system for intrusion detection, fault diagnosis, communication optimization and other services.

[0035] The station center supervision system, including multiple local hosts and display or projection devices, visualizes the digital twin rail transit Starlink system and provides supervision services by remotely accessing the station center digital twin server cluster system.

[0036] The services that the train equipment or trackside Starlink equipment needs to transmit include:

[0037] The data services transmitted by train equipment include: train operation control services, passenger information services, operating environment status perception services, train facility and equipment status perception services and other operation and maintenance services. Among them, train operation control services have the highest priority, train facility and equipment status perception services and operating environment status perception services are high priority, and passenger information services are low priority.

[0038] The data services transmitted by the trackside Starlink devices include: train operation control services, passenger information services, operating environment status perception services, train facility and equipment status perception services, and Starlink equipment group status perception services. Among them, train operation control services have the highest priority, train facility and equipment status perception services and operating environment status perception services are high priority, Starlink equipment group status perception services are medium priority, and passenger information services are low priority.

[0039] In the rail transit Starlink system based on digital twin:

[0040] The station center digital twin server cluster system collects the status data of the entire track system in real time through a dedicated sensor communication network, and maps the rail transit Starlink system in the actual physical domain to a digital twin rail transit Starlink system in the virtual digital domain through data preprocessing, data screening, and multi-source data feature fusion. In addition to the scene display function, the system in the virtual digital domain can also provide optimization solutions for communication solutions, intelligent services (intrusion detection, fault diagnosis, etc.), auxiliary decision-making and other services based on characteristic parameters, and present the results in a visual form to the supervisory administrator through the station center supervision system.

[0041] In the communication scheme design optimization method:

[0042] The rail transit Starlink system based on digital twins includes an actual physical domain system and a virtual digital domain system. The virtual digital domain system can collect parameters based on the status of the actual physical domain, and provide early warning, fault diagnosis, and system optimization functions for different services.

[0043] The rail transit Starlink system based on digital twin is characterized by:

[0044] The train is equipped with a multi-mode communication module, which can establish a direct communication link based on the PC5 interface with adjacent trains to achieve vehicle-to-vehicle communication, and can also establish a communication link based on the PC5 interface or Uu interface with trackside equipment. It can also establish a communication link based on the Uu interface with the communication infrastructure. Uu is divided into public network Uu interface and private network Uu interface;

[0045] Trackside Starlink equipment is equipped with a multi-mode communication module. In addition to being able to establish a direct communication link with the train, trackside Starlink equipment can also establish a direct communication link based on the PC5 interface. At the same time, the trackside Starlink equipment is connected to the ground communication infrastructure through cables (such as optical fiber, etc.);

[0046] Ground communication infrastructure uses existing 5G, LTE or WLAN networking methods, including public 5G, private 5G, private LTE, private WLAN and public LTE.

[0047] The specific steps of the rail transit Starlink system and communication solution design optimization method based on digital twins include:

[0048] (1) For different rail transit services, the services that need to be transmitted by train equipment or trackside Starlink equipment are prioritized. The priority of the task will serve as an important reference indicator for the communication optimization plan;

[0049] (2) The station center digital twin server cluster system collects the status information parameters of the train and the trackside Starlink equipment in the actual physical domain. The status information parameters include the train position, the trackside Starlink equipment position, the transmission power of different equipment, path loss, small-scale fading parameters, the average workload of each service, the half-duplex error rate for the PC5 interface direct communication, the signal strength, and the train service priority obtained in step 1;

[0050] (3) In the virtual data domain, define a communication link selection strategy network. The input of the strategy network is the short-term environmental parameters and the actual static parameters. The action output of the strategy network is the selection scheme of the communication link. The goal is to minimize the service delay and maximize the service success rate. Define an environment model. The input of the model is the state information of the previous moment and the action output of the strategy network. The output is the state information of the next moment.

[0051] The short-term environmental parameters include path loss, small-scale fading parameters, successful transmission probability and successful perception probability of the PC5 link, and the data volume of the service that needs to be transmitted by the train equipment or the trackside Starlink equipment in step (1); the actual static parameters include the position and transmission power of the trackside Starlink equipment, or the transmission power of the train transmitting terminal;

[0052] (4) In the virtual digital domain, the policy network interacts with the environment model to obtain the data output by the environment model. In the actual physical domain, the train and trackside Starlink equipment collect real environment data by implementing the policy network solution. Based on the real environment data and the data output by the environment model, the environment model is trained and optimized by minimizing the error.

[0053] (5) Define the reward function of the policy network and train the policy network using the deep Q learning method based on the trained environment model;

[0054] (6) deploying the strategy network to each train and trackside Starlink device through a dedicated sensor communication network 4;

[0055] (7) Repeat the above optimization arrangement (steps 4 to 6) by iteratively training the environment model and the policy network until the actual policy converges to the optimal one, and then evaluate the optimal communication link selection scheme.

[0056] The present invention provides a rail transit Starlink system and communication scheme design optimization method based on digital twin, and its beneficial effects are:

[0057] Through digital twin technology, comprehensive simulation and dynamic monitoring of rail transit systems can be achieved, the actual operating environment can be accurately simulated, and solutions can be provided for the optimization of various services of the train autonomous operation system. Regarding the communication optimization solution, the present invention solves the problem of poor optimization accuracy of communication strategies caused by complex and changeable state parameters and unknown environmental models in rail transit systems, and realizes safe, efficient and flexible selection and adjustment of data transmission links between trains and trackside Starlink equipment. This optimization method makes the communication of the Starlink system more efficient, data transmission smoother, and the real-time and accuracy of information significantly improved. Overall, the optimization design based on digital twins enhances the dispatching and command capabilities of the rail transit system, reduces delays and failure rates, and improves operational efficiency and passenger satisfaction. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] The present invention has the following accompanying drawings:

[0059] Figure 1 A schematic diagram of a rail transit Starlink system based on digital twins provided in an embodiment of the present invention;

[0060] Figure 2 A schematic diagram of the structure of a rail transit Starlink system based on digital twins provided in an embodiment of the present invention;

[0061] Figure 3 A schematic diagram of delay optimization results of a communication solution provided in an embodiment of the present invention;

[0062] Figure 4A schematic diagram of a train operation control service optimization result provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0063] The embodiments of the present application are described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present application, and cannot be interpreted as limiting the present application.

[0064] Those skilled in the art will appreciate that, unless otherwise stated, the singular forms "a", "an", "said" and "the" used herein may also include plural forms. It should be further understood that the term "comprising" used in the specification of the present application refers to the presence of the features, integers, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or groups thereof. The term "and / or" used herein includes any unit and all combinations of one or more associated listed items.

[0065] It will be understood by those skilled in the art that, unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as those generally understood by those skilled in the art to which this application belongs. It should also be understood that terms such as those defined in common dictionaries should be understood to have meanings consistent with the meanings in the context of the prior art, and will not be interpreted with idealized or overly formal meanings unless defined as herein.

[0066] To facilitate the understanding of the embodiments of the present application, the exemplary embodiments of the present application are further described in detail below in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than an exhaustive list of all the embodiments. It should be noted that the embodiments in the present application and the features in the embodiments can be combined with each other without conflict.

[0067] The present invention proposes a rail transit Starlink system based on digital twins. The station center digital twin server cluster system collects status data of the entire system through a dedicated sensor network.

[0068] Figure 1 The network architecture diagram of the rail transit Starlink system based on digital twin provided for this application. Figure 1 As shown in the figure, the system architecture mainly includes but is not limited to: an intelligent trackside perception unit with integrated sensing and sensing integrated on the trackside, intelligent on-board terminals integrated on the front and rear of the vehicle, ground communication infrastructure, a ground center and a digital twin cluster.

[0069] First, based on the sample data of the sensor system during the train's travel, the on-board sensor system collects train system status information from multiple aspects, such as train speed, train equipment status, passenger load factor, train safety status, and train location. The rail Starlink sensor system collects trackside line status information such as trackside switches and track obstacles. The station center sensor system collects station system status data and dispatch center plan data. The dedicated sensor communication network is based on the traditional WLAN or LTE communication of urban rail transit, and uploads the status information and channel information collected by the train, trackside, and station center to the station center digital twin server cluster system.

[0070] Secondly, the station center digital twin server cluster system collects the status information and operation data of the entire system in real time, and based on various digital domain-real domain interfaces, through data preprocessing, data screening, multi-source data feature fusion and other means, maps the rail transit Starlink system in the actual physical domain to the digital twin rail transit Starlink system in the virtual digital domain, trains the actual environment model in the digital twin server cluster system, and uses it for intrusion detection, fault diagnosis, communication optimization and other services. The station center supervision system visualizes the digital twin rail transit Starlink system and provides supervision services by remotely accessing the station center digital twin server cluster system.

[0071] Then, the optimal communication link selection and switching scheme is trained based on the environmental model, business characteristics, etc., and the communication link selection scheme of the actual system is updated after a period of time. Finally, the algorithm performance is evaluated based on the performance of the actual system.

[0072] In addition, Figure 2 As shown, a rail transit Starlink system based on digital twins provided in this embodiment includes an on-board sensor system 1, a rail Starlink sensor system 2, a station center sensor system 3, a dedicated sensor communication network 4, a station center digital twin server cluster system 5, and a station center supervision system 6.

[0073] The on-board sensing system 1 includes a millimeter-wave radar 11, a secondary protection radar 12, an image sensor 13, a door status sensor 14 and other modules. Through multiple modules, it collects all-round train system status information such as train speed, train equipment status, passenger load factor, train safety status, and train position. After the collection is completed, the collected data is periodically uploaded to the station center digital twin server cluster system 5 through the on-board private network communication transceiver equipment 41.

[0074] The track starlink sensing system 2 includes multiple modules such as the trackside equipment acquisition module 21, the image sensor 22, the vibration sensor 23, etc., wherein the trackside refers to the side of the track, or around the track. The trackside line status information such as the trackside switch and track obstacles are collected through multiple modules. After the collection is completed, the collected data is periodically sent to the station center digital twin server cluster system 5 through the trackside dedicated network communication transceiver equipment 42.

[0075] The station center sensing system 3 includes a server 31 dedicated to collecting information about the equipment and system status of the dispatch center, and a server 32 for collecting information about the status of other subsystems. A security interface is designed to periodically upload the status data of the station system and the planning data of the dispatch center to the station center digital twin server cluster system 5 through the station center private network communication transceiver equipment 43.

[0076] The dedicated sensor communication network 4, including the on-board private network communication transceiver equipment 41, the trackside private network communication transceiver equipment 42, the station center private network communication transceiver equipment 43, etc., is based on the traditional WLAN or LTE communication of urban rail transit, and is used to upload the train, trackside, station center status information and collected channel information collected by the above-mentioned transceiver equipment to the station center digital twin server cluster system 5.

[0077] The station center digital twin server cluster system 5, including multiple secure cloud servers, server cluster systems and multiple digital domain-real domain interfaces, collects the status information and operation data of the entire system in real time, and maps the actual orbital Starlink system to the simulation system based on various digital domain-real domain interfaces for intrusion detection, fault diagnosis, communication optimization and other services.

[0078] The station center supervision system 6 includes multiple local hosts and display or projection devices. It visualizes the digital twin rail transit Starlink system and provides supervision services by remotely accessing the station center digital twin server cluster system.

[0079] The present invention proposes a rail transit Starlink system based on digital twins. The station center digital twin server cluster system collects the status data of the whole system through a dedicated sensor network to support a variety of smart services based on big data. Taking smart fault diagnosis as an example:

[0080] When implementing fault diagnosis of trackside equipment, various types of sensors are used to collect various parameter data of railway equipment, such as temperature, vibration, current, etc. These sensors transmit data to the trackside Starlink device. By comparing the actual collected data with the expected data in the virtual simulation system, we can quickly identify any abnormal or abnormal equipment behavior. For example, when the temperature of a trackside device rises above a predetermined threshold, or the vibration frequency fluctuates abnormally, the system will automatically trigger an alarm and transmit the abnormal information to the ground center. In the digital twin server cluster system at the station center, we have trained a real-world environment model that can update the model based on real-time data changes. This enables us to predict possible equipment failures and identify potential causes of failures through further analysis of the model.

[0081] When implementing on-board equipment fault diagnosis, the train needs to send real-time information on the operating status to the trackside Starlink equipment, such as the train's location information, speed information, communication link status, control system log, etc. The trackside Starlink equipment analyzes the collected data and detects possible abnormalities or faults by monitoring data changes. For more complex data analysis, such as system-level faults, fault trend prediction, etc., the trackside Starlink equipment sends the data to the station center system. In the digital twin server cluster system at the station center, the actual environment model is used to model and predict the operation of the on-board equipment.

[0082] The present invention proposes a rail transit Starlink system based on digital twins, and proposes a communication scheme design optimization method based on digital twins for the instant communication needs in the rail transit Starlink system. The method is specifically implemented by the following steps:

[0083] by Figure 1 Take the data processing as an example to illustrate: for any mobile train n, it has the ability to communicate with adjacent trains n+1 and train n-1, M nearby trackside Starlink devices, and 1 communication infrastructure to meet the needs of different services. At the same time, for any fixed trackside Starlink device m, it has the ability to communicate with N nearby trains, two adjacent track Starlink devices m-1 and m+1, and 1 communication infrastructure to meet the needs of different services. Any trackside Starlink device, communication infrastructure, and train can be used as a relay to support different services. The train communicates with the communication infrastructure equipment using the Uu interface, and the trackside Starlink equipment and the communication infrastructure equipment use cable (such as optical fiber) equipment for communication. The remaining interfaces are all direct communication links based on the PC5 interface. The communication data relayed by the communication infrastructure needs to pass through the core network.

[0084] Take train equipment as the data generation end as an example: during communication, the data service priority of the train transmission is first defined. The services include: train operation control service, passenger information service, operating environment status perception service, train facility equipment status perception service and other operation and maintenance services. Among them, the train operation control service has the highest priority, the train facility equipment status perception service and the operating environment status perception service have high priority, and the passenger information service has low priority. The priority is determined by PO n，L (t) indicates.

[0085] The train needs to consider its own position L n (t), the target position L n，dd (t), the signal strength queue RSRP of the train communicating with different devices n (t) = {RSRP n，dd (T), RSRP m，1 (t), ..., RSRP m，M (t), RSRP i (t)}, and the average traffic volume of the four services B n (t) = {B 1 (t), B 2 (t), B 3 (t), B 4 (t)}.

[0086] For the communication based on Uu interface, the communication success rate is calculated by taking OFDM as an example. For the communication transmission power PR on a given single subcarrier, n , communication path loss PL n (t), small-scale fading parameter E in communication n，ssf And the sum of the power of interference and noise Φ n (t), then the signal to interference noise ratio is expressed as:

[0087]

[0088] The packet error rate of communication data can be expressed as:

[0089] PER n (t) = η 1 exp(η 2 γ n，eff )

[0090] Among them, η 1 and η 2 Parameters determined by the adaptive modulation and coding mechanism in the system, γ n，eff is the indexed signal-to-interference-noise ratio level.

[0091] For the communication based on PC5 interface, taking the existing V2X communication as an example, the success rate of data packet transmission is:

[0092]

[0093] in, is the probability of successful transmission, δ HD is the half-duplex error, is the probability of successful perception, PSR(d n ) is the packet perception rate. HD Under the premise of a given data frame length and service arrival rate, the speed is constant. When the train and other equipment are in a stable state, the service arrival rate is a fixed value, so δ HD In this system it can be considered as a fixed parameter. and PSR(L n ) It is a dynamic parameter because it takes into account the dynamic position and multipath between devices.

[0094] Define the policy network as η θ , the network output is action a n (t) indicates that any device is selected as the communication relay for this transmission. The station center digital twin server cluster system first performs environmental training modeling on the parameters without models, and the state parameters of the environmental model , the state at the next moment is s n，d (t+1). The environment model is trained in the virtual digital domain using the collected real sample data; the learning goal of the environment model is to find the environment model parameter φ that minimizes the prediction loss:

[0095]

[0096] Getting the environment model Based on this, the policy network ηθ is retrained. In addition to considering the existing state parameters, the priority PO of the transmission task needs to be considered in the training of the policy network ηθ. n，L (t). In this scheme, the train operation control business is defined as having the highest priority. The reward function can be defined as

[0097]

[0098] Among them, PE n (t) is the actual transmission success rate, It is about the actual communication delay T n is a decreasing function with different parameters depending on the service level. Given the state of the past T moments, the policy network η θ The output of is the actual selected communication link to maximize the cumulative expected reward

[0099] Apply the optimal communication link selection strategy obtained in the above steps to the real rail transit Starlink system, collect more data about the actual environment model, add them to the real sample data set, and continuously optimize the strategy network in the repetitive process. Compare and evaluate the switching performance indicators of the system before and after optimization. Figure 3 The simulation diagram of the average service delay and Figure 4 Schematic diagram of control business optimization results shown.

[0100] When the trackside Starlink device is the data generation end, the optimal communication link selection plan can also be obtained by referring to the above-mentioned business optimization steps with the train equipment as the data generation end.

[0101] The contents not described in detail in this specification belong to the prior art known to professional and technical personnel in this field.

Claims

1. A rail transit Starlink system based on digital twin, characterized in that: The rail transit Starlink system comprises: an on-board sensor system (1) and an on-board multi-mode communication module installed on each train, a rail Starlink sensor system (2) and a trackside multi-mode communication module installed in each trackside Starlink device, ground communication infrastructure, a station center sensor system (3), a station center digital twin server cluster system (5) and a station center supervision system (6); The on-board sensor system (1) is used to collect train system status information, and upload the train system status information to the station center digital twin server cluster system (5) through the on-board multi-mode communication module and the ground communication infrastructure; the train system status information includes train speed, train equipment status, passenger load factor, train safety status and train position; The track starlink sensing system (2) is used to collect trackside line status information, and upload the trackside line status information to the station center digital twin server cluster system (5) through the trackside multi-mode communication module and the ground communication infrastructure; the trackside line status information includes trackside equipment temperature, trackside equipment current, track vibration, trackside switch machine information and track obstacle information; The station center sensor system (3) is used to collect status data of the station system and planning data of the dispatching center, and upload them to the station center digital twin server cluster system (5) through the station center private network communication transceiver device (43); The above-mentioned vehicle-mounted multi-mode communication module, trackside multi-mode communication module and station center dedicated network communication transceiver equipment constitute a dedicated sensor communication network (4); The station center digital twin server cluster system (5) is used to collect the status information and operation data of the entire rail transit Starlink system, and based on the digital domain-real domain structure, map the real rail transit Starlink system into a digital twin rail transit Starlink system in the virtual digital domain, which is further applied to system intrusion detection, system equipment fault diagnosis and system communication optimization.

2. A rail transit Starlink system based on digital twin according to claim 1, characterized in that: The vehicle-mounted sensing system (1) is installed on a train and comprises a millimeter wave radar (11), a secondary protection radar (12), an image sensor (13), and a door status sensor (14); The vehicle-mounted multi-mode communication module is used to: establish a direct communication link based on the PC5 interface with adjacent trains to achieve vehicle-to-vehicle communication; establish a direct communication link based on the PC5 interface or Uu interface with nearby trackside Starlink equipment; Establish a communication link based on the Uu interface with the communication infrastructure; the above-mentioned Uu interface is divided into a public network Uu interface and a private network Uu interface; The track star chain sensing system (2) comprises a trackside equipment acquisition module (21), an image sensor (22) and a vibration sensor (23); The trackside multi-mode communication module is used to: establish a direct communication link based on the PC5 interface or Uu interface with each nearby train; establish a direct communication link based on the PC5 interface with the adjacent trackside Starlink device; The station center sensing system (3) comprises a dispatch center server (31) and a subsystem server (32); The ground communication infrastructure is a ground base station, and the ground communication infrastructure adopts 5G, LTE or WLAN networking mode, and the networking mode includes public network 5G, private network 5G, private network LTE, private network WLAN and public network LTE; The station center digital twin server cluster system (5) includes multiple secure cloud servers, a server cluster system and multiple digital domain-real domain interfaces; The station center supervision system (6) includes multiple local hosts and display devices; the station center supervision system (6) is used to visualize the digital twin rail transit Starlink system and provide supervision services.

3. An application of the rail transit Starlink system based on digital twin as claimed in claim 1 or 2, characterized in that: The application is fault diagnosis of trackside equipment or onboard equipment, which specifically includes the following steps: Step 1, transmitting the trackside line status information collected by the track starlink sensor system (2) or the train system status information collected by the vehicle-mounted sensor system (1) to the station digital twin server cluster system (5); Step 2: Compare the trackside line status information and train system status information collected in step 1 with the actual environment model data in the station digital twin server cluster system (5) to quickly identify abnormal behavior of the trackside Starlink equipment or abnormal behavior of the on-board equipment.

4. A communication scheme design optimization method for a rail transit Starlink system based on digital twins as claimed in claim 1 or 2, characterized in that: The steps include: Step 1: Prioritize the services that need to be transmitted by train equipment or trackside Starlink equipment; Step 2: The station center digital twin server cluster system collects the status information parameters of the train or trackside Starlink equipment in the actual physical domain. The status information parameters include: The train position, the trackside Starlink device position, the transmission power, path loss, and small-scale fading parameters of the train equipment or the trackside Starlink device, the average workload of the services that the train equipment or the trackside Starlink device needs to transmit in step 1, the half-duplex error rate and signal strength of the direct communication of the PC5 interface, and the priority level of the services that the train equipment or the trackside Starlink device needs to transmit obtained in step 1; Step 3, in the virtual data domain, define a communication link selection strategy network; the input of the strategy network is the short-term environmental parameters and the actual static parameters, and the action output of the strategy network is the selection scheme of the communication link; with the goal of minimizing the service delay and maximizing the service success rate, define the environment model, the input of the model is the state information of the previous moment and the action output of the strategy network, and the output is the state information of the next moment; The short-term environmental parameters include path loss, small-scale fading parameters, successful transmission probability and successful perception probability of the PC5 link, and the data volume of the service that needs to be transmitted by the train equipment or the trackside Starlink equipment in step 1; the actual static parameters include the position and transmission power of the trackside Starlink equipment, or the transmission power of the train transmitting terminal; Step 4: In the virtual digital domain, the policy network interacts with the environment model to obtain the data output by the environment model. In the actual physical domain, the train equipment or trackside Starlink equipment collects real environment data by implementing the policy network solution, and trains and optimizes the environment model based on the real environment data and the data output by the environment model by minimizing the error. Step 5: Define the reward function of the policy network. Based on the environment model trained in step 4, train the policy network through the deep Q learning method to obtain the communication link selection scheme. Step 6: deploy the communication link selection scheme obtained in step 5 to each train device or trackside Starlink device through a dedicated sensor communication network; Step 7: Repeat steps 4 to 6 above, and iteratively train the environment model and the policy network until the optimal communication link selection scheme is obtained.

Images