System and method for preventing mistaken opening of high-speed train door based on multi-sensor fusion

Abstract

The invention discloses a system and method for preventing mistaken opening of a high-speed rail vehicle door based on multi-sensor fusion, and belongs to the technical field of rail transit, the system comprises a control center module, a data acquisition and recognition module connected with the control center module, a vehicle door opening module, an early warning module and a multi-sensor module connected with the data acquisition and recognition module; a multi-sensor module is used for obtaining information of a train and a platform, whether the platform is located on the left side or the right side of the train is judged through a weighted voting decision-making mechanism, and when a train driver presses a button for opening a door mistakenly, a train door is controlled to prevent the train door from being opened mistakenly. Through various types of sensors and a weighted voting decision-making mechanism, the identification accuracy of the train and the platform is improved, the misoperation of a driver is reduced, then the train door is prevented from being opened by mistake, the safety and the intelligent level of high-speed train door control are improved, the risk of manual and equipment misoperation is effectively reduced, and the safety of a high-speed train is improved. And reliable guarantee is provided for safe operation of a high-speed rail system.

Description

Technical Field

[0001] The present invention belongs to the technical field of rail transit, especially the field related to the door control of high-speed rail trains, and specifically relates to a high-speed rail door anti-mis-opening system and method based on multi-sensor fusion. Background Art

[0002] As a modern rail transit tool, high-speed rail trains are widely used in the high-speed transportation networks of various countries around the world. Their efficient and safe operation is directly related to the safety of passengers and the railway system. To ensure the safety of passengers, when a high-speed rail train arrives at a station, the docking between the train doors and the station platform must be precise. If the wrong side doors are opened after the train stops, it will cause passengers to fall into the track area, posing serious safety hazards to the main line operation and passengers.

[0003] Currently, for the opening of high-speed rail train doors, the driver needs to first press the "Release Door X" (X is left or right) button, and then press the "Open Door X" button to open all the doors on the corresponding side through bus centralized control. However, in practical applications, human misoperations still occur frequently. In addition, the existing door control lacks a multi-level verification and warning mechanism, and the driver is easily affected by external conditions or himself / herself, resulting in the problem of incorrect door opening not being completely and effectively avoided.

[0004] Therefore, there is an urgent need for a system that can detect the left and right side positions of the platform relative to the vehicle after a high-speed rail arrives at the station based on multi-sensor data fusion, and verify and control when the driver opens the door. Through such a system, the driver's misoperation of opening the wrong side doors can be effectively avoided, thus ensuring the safety of passengers and train operation and improving the operation reliability of the train. Summary of the Invention

[0005] Aiming at the above deficiencies in the prior art, the high-speed rail door anti-mis-opening system and method based on multi-sensor fusion provided by the present invention solve the problem of incorrect opening of high-speed rail doors causing safety hazards to passengers and train operation.

[0006] To achieve the above invention purpose, the technical solution adopted by the present invention is: a high-speed rail door anti-mis-opening system based on multi-sensor fusion, including a control center module, a data acquisition and recognition module, a door opening module, and a warning module connected to the control center module, and a multi-sensor module connected to the data acquisition and recognition module; The multi-sensor module is used for train position information, train running speed, information on whether the train has stopped running, and relative position information between the train body and the platform; The data acquisition and recognition module is used for obtaining different data collected by the multi-sensor module and status data of the train driver pressing the door button, and weighted-fusing different data collected by the multi-sensor module to obtain the actual position information of the platform; The control center module is used to obtain the status data of the train driver pressing the door button and the actual platform position information output by the data acquisition and recognition module, compare and judge them, and control the door opening module and the warning module according to the judgment results; The door opening module is used to control the opening and closing of the doors; The warning module is used to send audible and visual warning information to the driver.

[0007] The beneficial effects of the present invention are as follows: By using various types of sensors and selecting preprocessing algorithms and weighted voting decision mechanisms suitable for different sensors, the present invention improves the recognition accuracy of trains and platforms, reduces driver's misoperations, and thus prevents the misopening of train doors, ensuring the safety of train operation and passengers. At the same time, it improves the safety and intelligence level of high-speed rail door control, facilitates passing vehicles to pass quickly, avoids safety accidents caused by misoperations and subsequent processing costs, effectively reduces the risks of human and equipment misoperations, and provides a reliable guarantee for the safe operation of the high-speed rail system.

[0008] Furthermore: The multi-sensor module includes a Beidou positioning unit, an MVB bus unit, and a distance detection unit; The Beidou positioning unit is used to obtain train position information, including the current position information of the train and the arrival platform information of the train; The MVB bus unit is used to obtain the train running speed, and based on the train running speed, obtain information on whether the train has stopped running; The distance detection unit is used to obtain the relative position information between the train body and the platform.

[0009] The beneficial effects of the above further solution are as follows: By collecting data through the multi-sensor module, the present invention can achieve precise train positioning, operation status monitoring, and environmental perception, improve the safety and accuracy of door control, and reduce the risk of misoperations.

[0010] Furthermore: The data acquisition and recognition module includes a collection unit, a calculation unit, and a decision-making unit connected in sequence; The collection unit is used to obtain different data collected by the multi-sensor module and the status data of the train driver pressing the door button, and set the collection frequency to generate a current collection time mark for the data collected by the multi-sensor module and the status data of the train driver pressing the door button; The calculation unit is used to obtain the platform position information corresponding to different data through corresponding data preprocessing methods according to the different data collected by the multi-sensor module; The decision-making unit is used to obtain the actual platform position information through weighted voting decision according to the platform position information corresponding to different data.

[0011] The beneficial effects of the above further solution are as follows: By dividing the data acquisition and recognition module into an acquisition unit, a calculation unit, and a decision-making unit, the present invention realizes the efficient processing of data and accurate decision-making, improving the response speed and decision-making accuracy of the system.

[0012] Furthermore: The control center module includes a data receiving unit, a judgment unit, a control unit, a mode switching unit, and a data storage unit that are connected in sequence; The data receiving unit is used to receive the actual platform position information and the status data of the train driver pressing the door button; The judgment unit is used to judge the decision conclusion of whether the door is opened according to the actual platform position information and the status data of the train driver pressing the door button; The control unit is used to send control information to the door opening module and the warning module according to the decision conclusion of whether the door is opened; The mode switching unit is used to control the data acquisition and recognition module and the control center module to enter the working mode and the shielding mode; The data storage unit is used to store different data collected by the multi-sensor module, the status data of the train driver pressing the door button, the control information sent by the control unit, and the mode switching data of the mode switching unit.

[0013] The beneficial effects of the above further solution are as follows: Through the control center module, the present invention can realize the control and management of the door opening decision-making and record data, which is convenient for fault diagnosis and performance optimization.

[0014] The present invention also provides a method for preventing incorrect opening of high-speed rail doors based on multi-sensor fusion, including the following steps: S1: Use multi-sensors to respectively obtain the train position information, train running speed, information on whether the train has stopped running, and the relative position information between the train body and the platform, and perform weight self-adjustment; S2: According to the different types of multi-sensors, adopt corresponding data preprocessing methods to process the information obtained by the multi-sensors in S1 to obtain the platform position information of the multi-sensors; S3: According to the platform position information of the multi-sensors and the prior information, use the weighted voting decision-making mechanism to obtain the actual position information of the train and the platform. In response to a tie in the weighted voting decision-making mechanism, introduce a conflict handling mechanism to handle the tie situation; S4: Judge whether the actual position information of the train and the platform is consistent with the door opening button pressed by the train driver. If so, the door opening button is valid and the train door opens. Otherwise, the door opening button is invalid and the train door cannot open, and an audible and visual warning is given.

[0015] The beneficial effects of the present invention are as follows: According to the multi-sensor data fusion and weighted voting decision-making mechanism, the present invention can improve the safety and accuracy of high-speed rail door control, accurately judge the position information of the train and the platform, and then correctly judge the door opening operation, prevent the wrong opening of the door, and effectively prevent the occurrence of safety accidents; And through the self-adjustment of weights, adjusting the different weights of multi-sensors, it can adapt to different high-speed rail operation scenarios and achieve high efficiency and intelligence in the door management of high-speed rail trains.

[0016] Furthermore: The multi-sensors include ultrasonic radars, millimeter-wave radars, and cameras installed on both sides of the body of each carriage, and lidar sensors installed at the head or tail of the train; The ultrasonic radar uses the Kalman filtering algorithm for data preprocessing; The millimeter-wave radar uses the least squares algorithm for data preprocessing; The camera uses the image enhancement algorithm for data preprocessing; The lidar sensor uses the point cloud filtering algorithm for data preprocessing.

[0017] The beneficial effects of the above further solution are as follows: By introducing the data preprocessing processes and methods of the Kalman filtering algorithm, the least squares algorithm, the image enhancement algorithm, and the point cloud filtering algorithm, the present invention improves the accuracy of the data and the robustness of the system, and enhances the accuracy and effectiveness of the door opening decision.

[0018] Furthermore: The expression of the weighted voting decision-making mechanism is as follows: ; ; ; Wherein, is the actual position information of the train and the platform, is the score that the platform is on the left side of the train, is the score that the platform is on the right side of the train, is the weighting coefficient of the i-th sensor , is the output of the i-th sensor for judging the platform position, taking a value of 0 or 1. When the current platform is on the left side, it is 1, and when it is on the right side, it is 0, is the total weight of the prior information, is the prior information. When the current platform is on the left side, it is 1, and when it is on the right side, it is 0, is the number of the sensor, , is the number of sensors of the multi-sensors.

[0019] The beneficial effects of the above further solution are as follows: The present invention calculates the scores of the platform on the left and right sides of the train through a weighted voting decision-making mechanism, introduces a preprocessing method, weight self-adjustment, prior information, and a conflict handling mechanism to determine the actual position information of the train and the platform, and performs fusion judgment on multi-source data to improve the reliability of the decision-making result under different environmental conditions. Brief Description of the Drawings

[0020] Figure 1 It is a structural diagram of a high-speed rail door anti-misopening system based on multi-sensor fusion. Detailed Embodiment

[0021] The following describes the detailed embodiment of the present invention to facilitate those skilled in the art of this technical field to understand the present invention. However, it should be clear that the present invention is not limited to the scope of the detailed embodiment. For those of ordinary skill in the art of this technical field, as long as various changes are within the spirit and scope of the present invention defined and determined by the appended claims, these changes are obvious, and all inventions and creations using the concept of the present invention are within the scope of protection.

[0022] As Figure 1 shown, a high-speed rail door anti-misopening system based on multi-sensor fusion includes a control center module, a data acquisition and recognition module, a door opening module, and a warning module connected to the control center module, and a multi-sensor module connected to the data acquisition and recognition module; Among them, the multi-sensor module is used to obtain the train position information, train running speed, information on whether the train stops running, and the relative position information between the train body and the platform; The data acquisition and recognition module is used to obtain different data collected by the multi-sensor module and the status data of the train driver pressing the door button, and perform weighted fusion on the different data collected by the multi-sensor module to obtain the actual position information of the platform; The control center module is used to obtain the status data of the train driver pressing the door button and the platform position information output by the data acquisition and recognition module, and perform comparison and judgment, and control the door opening module and the warning module according to the judgment result; The door opening module is used to control the opening and closing of the door; The warning module is used to send audible and visual warning information to the driver.

[0023] In an embodiment of the present invention, the multi-sensor module includes a Beidou positioning unit, an MVB bus unit, and a distance detection unit; The Beidou positioning unit is used to obtain the train position information, including the current position information of the train and the arrival platform information of the train. Among them, the arrival platform information of the train can be input in advance, recording the position of the platform when the train arrives, including that the platform is on the left side of the train and the platform is on the right side of the train, and the arrival platform information of the train can be obtained by querying the relevant platform information; The MVB bus unit is used to obtain the train running speed, and based on the train running speed, obtain the information on whether the train has stopped running; when it is detected that the train has stopped running, a high-speed rail door anti-mis-opening system based on multi-sensor fusion of the present invention enters a fully working state to monitor the position information of the train and the platform; The distance detection unit is used to obtain the relative position information between the train body and the platform, including: the distance data between the train body and the platform, the visual image data of the platform, and the three-dimensional shape data of the platform. Among them, a radar is used to identify through the density, shape and distribution of point cloud data to obtain the distance data between the train body and the platform and the three-dimensional shape data of the platform, and a camera is used to identify and obtain the visual image data of the platform through pixel and color information.

[0024] In an embodiment of the present invention, the data acquisition and recognition module includes a collection unit, a calculation unit and a decision unit connected in sequence; The collection unit is used to obtain different data collected by the multi-sensor module and the status data of the train driver pressing the door button, and set the collection frequency to generate a current collection time mark for the data collected by the multi-sensor module and the status data of the train driver pressing the door button; among them, the status data of the train driver pressing the door button includes "release the left door", "release the right door", "open the left door" and "open the right door", and the default collection frequency is 10Hz, which can be changed according to actual use; The calculation unit is used to obtain the platform position information corresponding to different data through corresponding data preprocessing methods according to different data collected by the multi-sensor module; among them, the platform position information corresponding to different data is the judgment result of the data collected by each sensor in the multi-sensor module.

[0025] The decision unit is used to obtain the actual platform position information through weighted voting decision according to the platform position information corresponding to different data. Among them, the weighted voting decision can be used to judge whether the current platform is on the left side or the right side of the train to obtain the actual platform position information.

[0026] In an embodiment of the present invention, the control center module includes a data receiving unit, a judgment unit, a control unit, a mode switching unit and a data storage unit connected in sequence; The data receiving unit is used to receive the actual platform position information and the status data of the train driver pressing the door button; A judgment unit, which is used to judge the decision conclusion of whether the door is opened according to the actual platform position information and the status data of the train driver pressing the door button, specifically as follows: When the actual platform position information is "the platform is on the left side of the train" and the status data of the train driver pressing the door button is "release the left door", it is judged as consistent, the decision conclusion is to open, the "open left door" button is valid, and after the train driver presses the "open left door", the train door opens; When the actual platform position information is "the platform is on the left side of the train" and the status data of the train driver pressing the door button is "release the right door", it is judged as inconsistent, the decision conclusion is not to open, the "open right door" button is invalid, and after the train driver presses the "open right door", the train door remains closed and an audible and visual warning is given; When the actual platform position information is "the platform is on the right side of the train" and the status data of the train driver pressing the door button is "release the left door", it is judged as inconsistent, the decision conclusion is not to open, the "open left door" button is invalid, and after the train driver presses the "open left door", the train door remains closed and an audible and visual warning is given; When the actual platform position information is "the platform is on the right side of the train" and the status data of the train driver pressing the door button is "release the right door", it is judged as consistent, the decision conclusion is to open, the "open right door" button is valid, and after the train driver presses the "open right door", the train door opens; A control unit, which is used to send control information to the door opening module and the warning module according to the decision conclusion of whether the door is opened; A mode switching unit, which is used to control the data acquisition and recognition module and the control center module to enter the working mode and the shielding mode. Among them, in the working mode, the anti-misoperation opening system of the high-speed rail door based on multi-sensor fusion works normally. In the shielding mode, the anti-misoperation opening system of the high-speed rail door based on multi-sensor fusion pauses working and will not generate any control and audible and visual warning information. It is applicable to scenarios such as door maintenance, repair, and car washing of the train. At the same time, the original door opening and closing process is not affected, improving the applicability and safety of the system; A data storage unit, which is used to store different data collected by the multi-sensor module, the status data of the train driver pressing the door button, the control information sent by the control unit, and the mode switching data of the mode switching unit; when the management personnel check the historical record data, they can quickly locate and verify various data and conclusions through the time mark.

[0027] The beneficial effects of the present invention are as follows: By using various types of sensors and selecting preprocessing algorithms and weighted voting decision mechanisms applicable to different sensors, the present invention improves the recognition accuracy of trains and platforms, reduces the misoperations of drivers, and thus prevents the accidental opening of train doors, avoiding the risks of passenger chaos and panic as well as personnel falling onto the track. At the same time, it facilitates passing trains to pass quickly, avoids safety accidents caused by misoperations and subsequent processing costs, significantly improves the safety and intelligent level of high-speed rail door control, effectively reduces the risks of human and equipment misoperations, and provides a reliable guarantee for the safe operation of the high-speed rail system.

[0028] Embodiment 2 The present invention provides a method for preventing accidental opening of high-speed rail doors based on multi-sensor fusion, including the following steps: S1: Using multi-sensors, respectively obtain train position information, train running speed, information on whether the train has stopped running, and the relative position information between the train body and the platform; S2: According to the different types of multi-sensors, adopt corresponding data preprocessing methods to process the information obtained by the multi-sensors in S1, obtain the platform position information of the multi-sensors, and perform weight self-adjustment; S3: According to the platform position information of the multi-sensors and prior information, use a weighted voting decision mechanism to obtain the actual position information of the train and the platform. In response to a tie in the weighted voting decision mechanism, introduce a conflict handling mechanism to handle the tie situation; S4: Judge whether the actual position information of the train and the platform is consistent with the door opening button pressed by the train driver. If so, the door opening button is valid and the train doors open. Otherwise, the door opening button is invalid, the train doors cannot open, and an audible and visual warning is given.

[0029] In S1, the multi-sensors include ultrasonic radars, millimeter-wave radars, and cameras installed on both sides of the body of each carriage, lidar sensors installed at the front or rear of the train, and Beidou positioning sensors to obtain the position information and speed information of the train; In S2, corresponding data preprocessing methods are adopted to process the information obtained by the multi-sensors in S1. The specific preprocessing methods are as follows: The ultrasonic radar uses the Kalman filtering algorithm for data preprocessing; The millimeter-wave radar uses the least squares algorithm for data preprocessing; The camera uses an image enhancement algorithm for data preprocessing; The lidar sensor uses a point cloud filtering algorithm for data preprocessing.

[0030] The weight self - adjustment is adaptively optimized through machine learning algorithms. The weights of each sensor can be optimized and adjusted according to data such as environmental factors, sensor performance, and historical judgment accuracy to improve the accuracy, adaptability, stability, and reliability of the system; the weight self - adjustment means , is the environmental factor (light, weather, etc.), is the self - inspection status (whether the line is broken, whether frames are lost, etc.), is the historical accuracy rate (based on historical judgment statistics), which can be adaptively optimized and adjusted through machine learning.

[0031] The value range of the total weight of various sensors under typical working conditions is shown in Table 1.

[0032] Table 1

[0033] All weights satisfy , The weight is a dynamically configurable parameter, and the system adjusts its value range according to factors such as real - time working conditions, sensor status, and self - inspection results as shown in Table 2.

[0034] Table 2

[0035] The typical value is 0.15 and can float between 0.1 and 0.3. The system ensures that in the fusion expression, is satisfied to maintain the numerical balance of the fusion decision. This mechanism improves the stability and fault - tolerance ability of the system when the perception ability is limited or conflict data occurs frequently.

[0036] Regarding the weight ratio of lidar and millimeter - wave radar, lidar has a high - precision advantage in the recognition result boundary at low speed and short distance (such as platform contour and three - dimensional coordinates). Combining with the point - cloud filtering algorithm, it is still robust even in rainy days. The system can self - adjust the weight ratio of lidar and other sensors according to the actual application working conditions.

[0037] In S3, the expression of the weighted voting decision mechanism is as follows: ; ; ; Among them, is the actual position information of the train and the platform, is the score that the platform is on the left side of the train, is the score that the platform is on the right side of the train, is the weighted coefficient of the i - th sensor , is the output for the i-th sensor to judge the platform position, with a value of 0 or 1. When the current platform is on the left side, it is 1, and when it is on the right side, it is 0. is the total weight of the prior information. is the prior information. When the current platform is on the left side, it is 1, and when it is on the right side, it is 0. is the sensor number. , is the number of sensors of the multi-sensor.

[0038] The conflict handling mechanism includes two typical situations: one is that the output results of some sensors are inconsistent, that is, some judge the platform to be on the left side, and some judge the platform to be on the right side. The system filters and suppresses the conflict information sources through means such as confidence analysis, anomaly recognition, weight reallocation, and the main sensor priority mechanism; the other is that the fusion voting calculation results are equal on both the left and right sides and cannot be clearly judged. The system corrects the historical preference , increases the weight of the prior information , adopts the majority voting mechanism or sets the default safe behavior and other means to handle such conflicts, ensuring that the system can output a unique and reliable judgment result in any case.

[0039] In an embodiment of the present invention, after performing corresponding data preprocessing methods according to the information obtained by the multi-sensor, the platform position information of each sensor in the multi-sensor is obtained. The following is an example to illustrate how the whole method ensures the reliability of the result when a certain sensor is abnormal: 4 ultrasonic radars, 2 millimeter-wave radars, 1 lidar, and 2 cameras are installed on the train, plus the prior information , assuming that the platform is actually on the left side of the train, and due to the influence of the environment or the platform shape, etc., the millimeter-wave radar detects that there are platforms on both the left and right sides 100 meters away and outputs abnormal situations, while the other sensors output normally; The system first performs preprocessing operations on the original data obtained by the above sensors to obtain the detection results of each sensor; Ultrasonic radar: It detects that the platform is on the left side. Therefore, the platform position information of the first ultrasonic radar , and the platform position information of the second ultrasonic radar , and the platform position information of the third ultrasonic radar , and the platform position information of the fourth ultrasonic radar ; Millimeter-wave radar: It detects that there are platforms on both sides. Therefore, the platform position information of the first millimeter-wave radar installed on the left side , and the platform position information of the second millimeter-wave radar installed on the right side ; Lidar sensor: It detects that the platform is on the left side. Therefore, the platform position information of the lidar sensor , the judgment result is on the left; Camera: It detects that the platform is on the left, so the platform position information of the first camera , the platform position information of the first camera ; Prior information: The current platform is on the left, ; In the above situation, a certain sensor data significantly deviates from or is inconsistent with other data. Because in actual operation, due to differences in the layout angles of sensors, different detection principles, and external environmental interferences (such as rain, fog, occlusion, reflection interference, platform shape, etc.) and its own failures, etc., there may be judgment differences or conflicts in the local detection outputs of each sensor, which is a common phenomenon in multi-source fusion systems; During the self-adjustment process of the system weights, when a certain sensor data significantly deviates from or is inconsistent with other data, its influence is reduced or ignored in the weighted system; Assume that the current experimental time is at night (the working condition environment is night), then the total weights of each sensor are shown in Table 3 below: Table 3

[0040] The total weight of all ultrasonic radars is set to , then , , , ; The total weight of all millimeter-wave radars is set to , then , ; The total weight of the lidar sensor is set to , ; The total weight of all cameras is set to , then , ; The weight of the prior information is set to , then ; The scores calculated for the platform being on the left side of the train and the platform being on the right side of the train are as follows respectively: ; ; Among them, represents the weight of the first ultrasonic radar, represents the weight of the second ultrasonic radar, represents the weight of the third ultrasonic radar, represents the weight of the fourth ultrasonic radar, represents the weight of the first millimeter-wave radar, Represents the weight of the second millimeter-wave radar, Represents the weight of the lidar, Represents the weight of the first camera, Represents the weight of the second camera, Represents the weight of the prior information; during the weight calculation process, when a certain sensor data is significantly deviated or inconsistent with other data, its influence is naturally suppressed or even ignored in the weighting system; Perform left and right score comparison: ; Therefore, the actual position information of the train and the platform is obtained as: the platform is on the left side of the train; In this embodiment, if occurs, the following method can be used to solve the situation of tied votes; 1) Increase the historical preference correction amount to ensure the uniqueness of the result; for example: initial calculation: , ; Since historical data shows that the platform at the same coordinate position last time was on the left, after adjustment: ; 2) The weight of the prior information (prior data) can be increased to ensure the uniqueness of the result; 3) Adopt a majority voting mechanism: count the results of the 4 types of sensor data and the prior information : If the majority judges that the platform is on the left, select the left; if the majority judges that the platform is on the right, select the right; 4) The default safe behavior is to further increase the proportion of the weight of the sensor with a larger total weight and recalculate the left and right scores; The present invention adopts technical means such as multi-sensor data fusion, abnormal elimination, weight self-adjustment, conflict handling mechanism, weighted voting decision-making, etc., to ensure that the system can output a unique and credible judgment result in any case.

[0041] In S4, if the door opening button pressed by the train driver is "release the left door", which is consistent with the actual position information of the train and the platform, both are on the left side, the "open left door" button is valid. After the train driver presses the "open left door" button, the train door opens. If the door opening button pressed by the train driver is "release the right door", which is inconsistent with the actual position information of the train and the platform, the "open right door" button is invalid. After the train driver presses the "open right door" button, the train door cannot be opened, and an audible and visual warning is given.

[0042] The beneficial effects of the present invention are as follows: By using various types of sensors and selecting preprocessing methods and weighted voting decision-making mechanisms applicable to different sensors, and by introducing weight self-adjustment and conflict handling mechanisms, the present invention improves the accuracy of platform position recognition, reduces driver's misoperations, thereby preventing the accidental opening of train doors, avoiding passenger chaos, panic, and the risk of personnel falling onto the track. At the same time, it facilitates passing vehicles to pass quickly, avoids safety accidents caused by misoperations and subsequent processing costs, significantly improves the safety and intelligent level of high-speed rail door control, effectively reduces the risks of human and equipment misoperations, and provides a reliable guarantee for the safe operation of the high-speed rail system.

Claims

1. A high-speed rail door anti-misopening system based on multi-sensor fusion, comprising a control center module, a data acquisition and identification module, a door opening module and an early warning module connected to the control center module, and a multi-sensor module connected to the data acquisition and identification module; The multi-sensor module is used for train position information, train running speed, whether the train has stopped running, and relative position information between the train body and the platform; The data acquisition and identification module is used to obtain different data collected by the multi-sensor module and the state data of the train driver pressing the door button, and weightedly fuse the different data collected by the multi-sensor module to obtain the actual position information of the platform; The control center module is used to obtain the status data of the train driver pressing the door button and the actual platform position information output by the data acquisition and identification module, and make a comparison and judgment, and control the door opening module and the early warning module according to the judgment result; The door opening module is used to control the opening and closing of the door; The warning module is used to send sound and light warning information to the driver.

2. The high-speed rail door anti-misopening system based on multi-sensor fusion according to claim 1 is characterized in that: The multi-sensor module includes a Beidou positioning unit, an MVB bus unit and a distance detection unit; The Beidou positioning unit is used to obtain train location information, including the current location information of the train and the arrival platform information of the train; The MVB bus unit is used to obtain the running speed of the train and obtain information on whether the train has stopped running based on the running speed of the train; The distance detection unit is used to obtain the relative position information between the train body and the platform.

3. The high-speed rail door anti-misopening system based on multi-sensor fusion according to claim 1 is characterized in that: The data acquisition and identification module includes an acquisition unit, a calculation unit and a decision unit connected in sequence; The acquisition unit is used to obtain different data collected by the multi-sensor module and the state data of the train driver pressing the door button, and set the acquisition frequency to generate a current acquisition time mark for the data collected by the multi-sensor module and the state data of the train driver pressing the door button; The calculation unit is used to obtain the station location information corresponding to the different data collected by the multi-sensor module through a corresponding data preprocessing method; The decision-making unit is used to obtain actual platform location information through weighted voting decision based on the platform location information corresponding to different data.

4. The high-speed rail door anti-misopening system based on multi-sensor fusion according to claim 1 is characterized in that: The control center module includes a data receiving unit, a judgment unit, a control unit, a mode switching unit and a data storage unit connected in sequence; The data receiving unit is used to receive the actual platform position information and the status data of the train driver pressing the door button; The judgment unit is used to determine whether the door is open based on the actual platform position information and the state data of the train driver pressing the door button; The control unit is used to send control information to the door opening module and the early warning module according to the decision conclusion of whether the door is opened; The mode switching unit is used to control the data acquisition and identification module and the control center module to enter the working mode and the shielding mode; The data storage unit is used to store different data collected by the multi-sensor module, state data of the train driver pressing the door button, control information sent by the control unit, and mode switching data of the mode switching unit.

5. A method for preventing mis-opening of high-speed rail doors based on multi-sensor fusion, characterized in that: The following steps are involved: S1: Use multiple sensors to obtain train location information, train running speed, whether the train has stopped running, and the relative position information between the train body and the platform, and adjust the weights automatically; S2: According to different types of multi-sensors, the corresponding data preprocessing method is used to process the information obtained by the multi-sensors in S1 to obtain the station location information of the multi-sensors; S3: Based on the platform location information and prior information of multiple sensors, a weighted voting decision mechanism is used to obtain the actual location information of the train and the platform. In response to a tie in the weighted voting decision mechanism, a conflict handling mechanism is introduced to handle the tie situation; S4: Determine whether the actual position information of the train and the platform is consistent with the door opening button pressed by the train driver. If so, the door opening button is valid and the train door is opened. Otherwise, the door opening button is invalid, the train door cannot be opened, and an audible and visual warning is given.

6. The method for preventing mis-opening of high-speed rail doors based on multi-sensor fusion according to claim 5 is characterized in that: The multi-sensor includes ultrasonic radars, millimeter-wave radars and cameras installed on both sides of the body of each vehicle, and a laser radar sensor installed at the front or rear of the vehicle; The ultrasonic radar uses a Kalman filter algorithm to perform data preprocessing; The millimeter wave radar uses a least squares algorithm to perform data preprocessing; The camera uses an image enhancement algorithm to perform data preprocessing; The laser radar sensor uses a point cloud filtering algorithm to perform data preprocessing.

7. The method for preventing mis-opening of high-speed rail doors based on multi-sensor fusion according to claim 5 is characterized in that: The expression of the weighted voting decision mechanism is as follows: in, The actual location information of the train and platform, is the score for a platform located on the left side of the train, is the score for a platform located on the right side of the train, is the weighting coefficient of the i-th sensor, , is the output of the i-th sensor to determine the platform position, and its value is 0 or 1. It is 1 when the current platform is on the left and 0 on the right. is the total weight of the prior information, is the prior information, which is 1 when the current platform is on the left and 0 when the current platform is on the right. is the sensor number, , is the number of sensors for multi-sensor.

Images