Train accurate positioning system based on multi-data interaction

Abstract

The invention discloses a train accurate positioning system based on multi-data interaction, which relates to the technical field of rail transit control and comprises a vehicle-mounted positioning terminal, a trackside cooperative positioning network and a cloud positioning decision platform. According to the invention, the multi-mode satellite receiving module, the inertial measurement unit, the millimeter wave radar and the laser radar module of the vehicle-mounted positioning terminal are used, and the intelligent transponder cluster, the optical fiber positioning sensor and the edge computing node of the trackside cooperative positioning grid are combined, so that real-time acquisition and interaction of multi-source heterogeneous data are realized; a data fusion engine of the cloud positioning decision-making platform is based on a Kalman filtering algorithm, combines a dynamic credibility evaluation model to dynamically distribute weights for data such as satellite signals, ultra wide band distance measurement, optical fiber mileage and radar feature matching according to scenes, and can output sub-meter positioning precision in different environments such as open areas, tunnels and curves. The problem of failure or error accumulation during signal shielding of a single technology is solved, and the deployment cost is remarkably reduced compared with that of a traditional scheme.

Description

Technical Field

[0001] The present invention relates to the technical field of rail traffic control, and in particular to a train precise positioning system based on multiple data interaction. Background Art

[0002] In the field of rail transit operation control, existing train positioning technologies mainly include positioning methods that rely on a single technology and some multi-sensor fusion positioning solutions. Single technologies include global satellite navigation systems (GPS, Beidou, etc.), inertial navigation systems, track circuits or query transponder positioning.

[0003] However, in terms of a single technology, satellite navigation systems will fail in signal-blocking environments such as tunnels, canyons, and urban high-rise buildings. The accumulated errors of inertial navigation systems increase significantly over long periods of operation. Track circuits or query transponders can only provide discrete position correction points. In terms of positioning accuracy, traditional positioning methods are usually at the level of hundreds of meters, which cannot meet the continuous sub-meter accuracy requirements of modern train control systems. In terms of information updates, delays in train-to-ground communication result in positioning information that cannot reflect the actual position of the train in real time. In terms of cost, the deployment and maintenance costs of some high-precision positioning solutions are too high. Summary of the Invention

[0004] Based on this, the purpose of the present invention is to provide a train precise positioning system based on multiple data interaction to solve the technical problems in the above-mentioned background technology.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a train precise positioning system based on multiple data interaction, comprising:

[0006] On-board positioning terminal: This includes a multi-mode satellite receiving module for receiving satellite positioning signals, an inertial measurement unit for measuring train acceleration and angular velocity, millimeter-wave radar and lidar modules for scanning and identifying track and surrounding environmental features, and a vehicle-to-ground communication unit that supports multiple wireless communication protocols;

[0007] Trackside collaborative positioning grid: This includes clusters of smart transponders deployed at preset intervals along the track. These transponders have built-in ultra-wideband communication chips for precise distance determination with onboard terminals; fiber-optic positioning sensors installed on the rails or trackside to continuously measure the precise mileage of trains; and edge computing nodes located at specific trackside locations to synchronize and pre-process onboard data and trackside data in a local area.

[0008] Edge computing nodes are deployed at key locations such as trackside stations and tunnel entrances, with a spacing of less than or equal to 500 meters. They are used to perform PTP time synchronization and validity filtering for onboard data and trackside data.

[0009] The digital twin track library stores track horizontal and vertical section information, fixed facility coordinates, and tunnel feature point cloud data, providing a feature matching benchmark for vehicle-mounted radar / laser scanning;

[0010] Cloud-based positioning decision-making platform: This includes a data fusion engine for fusing and processing multi-source heterogeneous data, a dynamic credibility assessment model for dynamically evaluating the credibility of each data source and assigning weights based on real-time scenarios, and a digital twin track library that stores detailed track line feature information.

[0011] The onboard positioning terminal exchanges data with the trackside collaborative positioning network via wireless communication. The edge computing node uploads the pre-processed data to the cloud-based positioning decision platform. The fused positioning results output by the cloud-based positioning decision platform are sent to the train via the train-ground communication network.

[0012] The vehicle-ground communication unit supports 5G-R, LTE-R, and WLAN communication protocols. After communication is restored, the vehicle terminal automatically uploads local fusion data and receives positioning results calibrated in the cloud.

[0013] The dynamic credibility assessment model is configured as follows:

[0014] When the satellite signal quality meets the preset open area condition, the satellite positioning data is assigned a first weight value;

[0015] When the train is within the effective coverage of the ultra-wideband signal of the intelligent transponder, a second weight value is assigned to the ultra-wideband ranging data;

[0016] When the train running speed exceeds a preset speed threshold, a third weight value is assigned to the continuous mileage data provided by the optical fiber positioning sensor;

[0017] When the system determines that the scene is a tunnel, a curve, or a densely built-up area, assigning a fourth weight value to the environmental feature matching data identified by the millimeter-wave radar and the lidar module;

[0018] When other primary positioning signal sources fail or their quality falls below a preset threshold, the weight of the inertial measurement unit data is increased.

[0019] Furthermore, the data fusion engine adopts a fusion algorithm based on the Kalman filter principle, and uses the weight factors assigned by the dynamic credibility assessment model to perform weighted processing on the state estimation information of each data source participating in the fusion.

[0020] Furthermore, the vehicle-mounted positioning terminal further includes a local decision module configured to switch to a local fusion positioning mode that relies on the inertial measurement unit data and the optical fiber positioning sensor data during a period of vehicle-ground communication interruption.

[0021] Furthermore, the deployment spacing of the smart transponder cluster along the track line is less than or equal to fifty meters, and it has a built-in ultra-wideband communication chip and absolute position coordinates, and uses UWB technology to measure distance or broadcast position with the vehicle-mounted positioning terminal.

[0022] Furthermore, the fiber optic positioning sensor is a fiber optic Bragg grating sensor array laid along the rail, which generates continuous mileage counting information by sensing the wheel passing over, and the measurement resolution is less than or equal to ten centimeters.

[0023] In summary, the present invention has the following main beneficial effects: the present invention utilizes the multi-mode satellite receiving module, inertial measurement unit, millimeter-wave radar, and lidar module of the vehicle-mounted positioning terminal, combined with the intelligent transponder cluster, fiber-optic positioning sensor, and edge computing node of the trackside collaborative positioning grid, to achieve real-time collection and interaction of multi-source heterogeneous data. The data fusion engine of the cloud-based positioning decision platform is based on the Kalman filter algorithm and combines it with a dynamic credibility assessment model to dynamically assign weights to satellite signals, ultra-wideband ranging, fiber-optic mileage, radar feature matching, and other data according to the scenario. It can output sub-meter positioning accuracy in different environments such as open areas, tunnels, and curves, solving the problem of failure or error accumulation of a single technology when the signal is blocked. In addition, the on-board local decision module can rely on inertial data and fiber-optic mileage to maintain positioning when vehicle-to-ground communication is interrupted, and automatically calibrate after communication is restored to ensure positioning continuity. The digital twin track library provides a feature matching benchmark for radar scanning, further improving positioning reliability in complex scenarios. The overall system meets the requirements of modern train control systems for high-precision positioning, and the deployment cost is significantly reduced compared to traditional solutions. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 Schematic diagram of the overall system architecture of the present invention;

[0025] Figure 2 This is a main flow chart of the train precise positioning method of the present invention;

[0026] Figure 3 This is a schematic diagram of the matching of vehicle-mounted radar / laser scanning tunnel wall feature points with a digital map. DETAILED DESCRIPTION

[0027] The following will be combined with the accompanying drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and should not be understood as limiting the present invention.

[0028] The following describes an embodiment of the present invention based on its overall structure.

[0029] like Figure 1-3 As shown in the figure, this system mainly consists of the following parts:

[0030] On-board positioning terminal: installed on the train, it is the direct data collection terminal;

[0031] Multi-mode satellite receiving module: Receives and processes signals from multiple global navigation satellite systems, such as GPS, BeiDou, GLONASS, and Galileo;

[0032] Inertial measurement unit: usually contains a three-axis accelerometer and a three-axis gyroscope, used to measure the train's acceleration and angular velocity and calculate position changes;

[0033] Trackside feature recognition module: Integrates millimeter-wave radar and lidar to actively scan fixed facilities on both sides of and above the track, such as tunnel walls, catenary pillars, platforms, and signal machines, to obtain environmental feature point cloud data for matching and positioning with preset digital maps;

[0034] Train-to-ground communication unit: supports multiple communication protocols such as 5G-R (5G for railways), LTE-R (LTE for railways), and WLAN, enabling high-speed and reliable data interaction with trackside equipment and cloud platforms;

[0035] Trackside collaborative positioning network: deployed along the railway line to provide local reference and auxiliary positioning information;

[0036] Smart transponder clusters: Deployed at high density along the track, for example, every 50 meters or less, each transponder has built-in precise absolute position coordinates and an ultra-wideband communication chip. When a train passes over it, it uses UWB technology to accurately measure the distance or broadcast the location to the onboard terminal.

[0037] Fiber Bragg Grating (FBG): Special fiber optic sensors, such as fiber Bragg Grating (FBG) sensor arrays, are installed along the rails. When a train wheel passes over the rails, the sensors can sense and accurately count the number of times a train wheel passes over them, providing continuous, high-resolution mileage information.

[0038] Edge computing nodes: Computing devices deployed at intervals of 500 meters at key locations along the line, such as stations, section signal points, and tunnel entrances, are responsible for receiving and processing data from nearby onboard terminals and trackside equipment, performing localized calculations such as time synchronization, data filtering, and preliminary fusion, reducing the burden on the cloud and improving response speed.

[0039] Cloud-based positioning decision platform: located in a data center or cloud environment;

[0040] Multi-source data fusion engine: The core processing unit receives pre-processed data from edge computing nodes. It uses advanced fusion algorithms, such as the federated Kalman filter, to organically integrate satellite positioning data, inertial navigation data, UWB ranging / positioning data, fiber optic mileage data, and radar / laser point cloud matching data to calculate the optimal train position estimate;

[0041] Dynamic Credibility Assessment Model: This model analyzes the train's operating environment in real time. Based on information such as location, speed, received signal strength, feature matching, and the quality of each data source, such as satellite signal strength, IMU error estimation, and UWB signal quality, it dynamically assigns a "weight" or "credibility coefficient" to each participating data source according to preset rules and logic. This weight reflects the data's reliability in the current scenario. For example, satellite data is given a higher weight in open areas, while radar matching and fiber optic data are given a higher weight in tunnels.

[0042] Digital Twin Track Library: This stores high-precision electronic track map data, including horizontal and vertical section information of the line, such as curve radius, slope, and coordinates of fixed facilities, such as signals, transponders, switches, platforms, and tunnel feature points, providing a benchmark for on-board feature matching and position verification.

[0043] Positioning method process

[0044] The following steps are involved:

[0045] Data collection and upload:

[0046] The vehicle-mounted positioning terminal continuously collects satellite positioning information, raw data from inertial sensors, and environmental perception point cloud data obtained through radar and laser scanners;

[0047] When a train approaches, the trackside intelligent transponder sends its own precise location information or performs distance measurement interaction to the onboard terminal via ultra-wideband signals;

[0048] Fiber optic positioning sensors sense the passage of train wheels in real time and generate continuous and accurate mileage counting information;

[0049] The on-board terminal uploads the collected satellite data, IMU data, and radar / laser point cloud data to the nearest trackside edge computing node through the vehicle-ground communication unit; the smart transponder information and fiber grating information are also transmitted to the edge computing node.

[0050] Edge preprocessing:

[0051] The edge computing node receives data streams from onboard terminals and trackside equipment;

[0052] Perform strict time synchronization on all received data, such as using the PTP precision time protocol, to ensure that the data has a unified time reference;

[0053] Perform preliminary validity checks and filtering on the data, such as removing satellite data with signal strength below the threshold and IMU data that is clearly beyond the reasonable range.

[0054] Data upload and scenario evaluation:

[0055] The edge computing node uploads the pre-processed and time-synchronized data to the cloud-based positioning decision platform;

[0056] The cloud platform's dynamic credibility assessment model analyzes and determines the train's current operating scenario, such as "high-speed operation on open plains", "medium-speed operation in tunnels", and "urban curved areas", based on received data such as train speed, satellite signal strength, UWB signal status, and preset geographic fence information such as tunnel signs.

[0057] Dynamic weight allocation:

[0058] The dynamic credibility evaluation model assigns a dynamic weight factor to each currently available data source. The allocation rules are as follows:

[0059] Scenario "Open Plain Highway": Satellite positioning data has a high weight, such as 0.7, and fiber optic mileage data has a high weight, such as 0.8. Due to the high speed, radar matching data has a low weight, such as 0.2.

[0060] Scenario "Tunnel Entrance": UWB data has a high weight, such as 0.9. As the transponder is triggered, the radar matching data weight begins to increase, such as 0.6, and the satellite data weight begins to decrease, such as 0.4.

[0061] Scenario "Mid-Tunnel": UWB data has a weight of 0, radar matching data has a high weight, such as 0.75, fiber optic mileage data has a high weight, such as 0.85, IMU data has a medium weight, such as 0.6, and satellite data has a weight of 0;

[0062] Scenario "City Curve": The weight of radar matching data is increased, for example, to 0.75. The weight of satellite data may decrease or fluctuate due to occlusion.

[0063] Scenario "Major signal failure": When the satellite and UWB signals fail simultaneously for more than a set time, such as 3 seconds, the IMU data weight is significantly increased, such as a linear increase from 0.5 to 0.9).

[0064] Multi-source data fusion:

[0065] The data fusion engine receives the status information of each data source with assigned weights;

[0066] Adopting a fusion algorithm based on the Kalman filter principle, such as the federated Kalman filter or its variants;

[0067] The algorithm uses the state prediction and measurement values ​​of each data source, combined with its assigned weight factor, and performs iterative calculations. Data sources with higher weights have a greater impact on the final result.

[0068] Output the optimal real-time train position, speed and attitude estimation values ​​at the current moment.

[0069] Results distribution and calibration:

[0070] The high-precision positioning results obtained by fusion calculation are sent to the train's onboard equipment through the train-ground communication network for use by the train operation control system;

[0071] The fused positioning results are used to perform online calibration of the accumulated error of the vehicle's inertial measurement unit, especially when passing by a smart transponder with precise position.

[0072] Communication interruption handling (degraded mode):

[0073] When the system detects a disruption in the vehicle-to-ground communication link, the local decision-making module of the vehicle-mounted positioning terminal automatically starts. The local decision-making module mainly relies on the vehicle's inertial measurement unit data and the continuous mileage data received from the fiber optic positioning sensor;

[0074] Simplified fusion algorithms, such as simplified Kalman filtering, are executed on the vehicle side to perform position estimation and maintain basic positioning functions. Once vehicle-ground communication is restored, the vehicle terminal reconnects to the cloud fusion system, uploads local data and receives the fusion positioning results sent from the cloud.

[0075] Although an embodiment of the present invention has been shown and described, this specific embodiment is merely an explanation of the present invention and is not a limitation of the invention. The specific features, structures, materials or characteristics described may be combined in an appropriate manner in any one or more embodiments or examples. After reading this specification, those skilled in the art may make modifications, substitutions and variations to the embodiment without creative contribution as needed without departing from the principles and purpose of the present invention. However, as long as they are within the scope of the claims of the present invention, they are protected by patent law.

Claims

1. A train precise positioning system based on multiple data interaction, characterized in that: include: On-board positioning terminal: This includes a multi-mode satellite receiving module for receiving satellite positioning signals, an inertial measurement unit for measuring train acceleration and angular velocity, millimeter-wave radar and lidar modules for scanning and identifying track and surrounding environmental features, and a vehicle-to-ground communication unit that supports multiple wireless communication protocols; Trackside collaborative positioning grid: This includes clusters of smart transponders deployed at preset intervals along the track. These transponders have built-in ultra-wideband communication chips for precise distance determination with onboard terminals; fiber-optic positioning sensors installed on the rails or trackside to continuously measure the precise mileage of trains; and edge computing nodes located at specific trackside locations to synchronize and pre-process onboard data and trackside data in a local area. Cloud-based positioning decision-making platform: This includes a data fusion engine for fusing and processing multi-source heterogeneous data, a dynamic credibility assessment model for dynamically evaluating the credibility of each data source and assigning weights based on real-time scenarios, and a digital twin track library that stores detailed track line feature information. Among them, the on-board positioning terminal interacts with the trackside collaborative positioning network through wireless communication, the edge computing node uploads the pre-processed data to the cloud-based positioning decision platform, and the fused positioning results output by the cloud-based positioning decision platform are sent to the train through the vehicle-ground communication network.

2. The train precise positioning system based on multiple data interaction according to claim 1, characterized in that: The dynamic credibility assessment model is configured as follows: When the satellite signal quality meets the preset open area condition, the satellite positioning data is assigned a first weight value; When the train is within the effective coverage of the ultra-wideband signal of the intelligent transponder, a second weight value is assigned to the ultra-wideband ranging data; When the train running speed exceeds a preset speed threshold, a third weight value is assigned to the continuous mileage data provided by the optical fiber positioning sensor; When the system determines that the scene is a tunnel, a curve, or a densely built-up area, assigning a fourth weight value to the environmental feature matching data identified by the millimeter-wave radar and the lidar module; When other primary positioning signal sources fail or their quality falls below a preset threshold, the weight of the inertial measurement unit data is increased.

3. The train precise positioning system based on multiple data interaction according to claim 2, characterized in that: The data fusion engine adopts a fusion algorithm based on the Kalman filter principle and uses the weight factors assigned by the dynamic credibility assessment model to perform weighted processing on the state estimation information of each data source involved in the fusion.

4. The train precise positioning system based on multiple data interaction according to claim 1, characterized in that: The vehicle-mounted positioning terminal further includes a local decision module configured to switch to a local fusion positioning mode that relies on the inertial measurement unit data and the optical fiber positioning sensor data during a vehicle-ground communication interruption.

5. The train precise positioning system based on multiple data interaction according to claim 1, characterized in that: The deployment spacing of the smart transponder cluster along the track line is less than or equal to 50 meters, and it has a built-in ultra-wideband communication chip and absolute position coordinates, and uses UWB technology to measure distance or broadcast position with the on-board positioning terminal.

6. The train precise positioning system based on multiple data interaction according to claim 1, characterized in that: The fiber optic positioning sensor is a fiber optic Bragg grating sensor array laid along the rails. It generates continuous mileage counting information by sensing the wheel passing over the rails, and the measurement resolution is less than or equal to ten centimeters.

7. The train precise positioning system based on multiple data interaction according to claim 1, characterized in that: The edge computing nodes are deployed at key locations such as trackside stations and tunnel entrances, with a spacing of less than or equal to 500 meters, and are used to perform PTP time synchronization and validity filtering on-board data and trackside data.

8. The train precise positioning system based on multiple data interaction according to claim 1, characterized in that: The digital twin track library stores track horizontal and vertical section information, fixed facility coordinates, and tunnel feature point cloud data, providing a feature matching benchmark for vehicle-mounted radar / laser scanning.

9. The train precise positioning system based on multiple data interaction according to claim 1, characterized in that: The vehicle-ground communication unit supports 5G-R, LTE-R and WLAN communication protocols. After the communication interruption is restored, the on-board terminal automatically uploads local fusion data and receives the positioning results calibrated in the cloud.

Images