Safety verification method and device for environment perception device of obstacle detection system

Abstract

The application relates to a safety verification method and device for an environmental perception device of an obstacle detection system, which comprises the following steps: acquiring original data of the environmental perception device; acquiring general data of a signal system of urban rail transit; extracting data features of the original data of the environmental perception device according to the general data of the signal system; comparing the data features to determine whether the environmental perception device has a fault; and judging the integrity of the original data of the environmental perception device. The perception device can be applied to the obstacle detection safety system as a black box, without needing to pay too much attention to whether the perception device itself has safety authentication or not, and without needing to pay too much attention to the complex failure mechanism of the perception device; the original positioning and speed measuring devices of the signal system can be flexibly selected and reused, so that the cost is saved; separate positioning and speed measuring devices can also be used to further improve the availability; and the perception device itself does not need to be configured for redundancy verification, so that the cost and installation space are saved.

Description

Technical Field

[0001] This disclosure relates to the field of urban rail transit signal control technology, and in particular to a safety verification method, device and control system for an obstacle detection system environmental perception device. Background Technology

[0002] With the widespread adoption of urban rail transit, rail travel has become an indispensable mode of transportation, making the safety of rail transit operations increasingly important. Traditional rail transit safety relies not only on the protection of the signaling system but also on the correct responses of staff when necessary, especially when encountering obstacles that traditional signaling systems cannot detect, requiring the driver to take emergency evasive measures to avoid hazards. To improve operational efficiency, more and more urban rail transit systems are applying fully automated driverless technology. In environments without a driver's presence, highly reliable obstacle detection systems are crucial for ensuring the safe operation of trains.

[0003] With the rapid development of artificial intelligence and autonomous driving technology, active obstacle detection technology has become increasingly mature and sophisticated in the automotive field, and has begun to be applied in urban rail transit. As a crucial safety function, the accuracy of obstacle detection directly affects driver and passenger safety. If the environmental sensing equipment malfunctions and fails to detect obstacles, continuing to output incorrect information, it could lead to serious operational accidents. Therefore, strict safety measures must be taken to ensure the correct and effective execution of obstacle detection, and appropriate measures must be taken in the event of equipment failure or malfunction.

[0004] Currently, active obstacle detection systems primarily employ environmental sensing devices such as industrial cameras, LiDAR, and millimeter-wave radar, or a combination of several devices and technologies to collect and process obstacle data. While these environmental sensing devices are widely used in autonomous driving scenarios, very few have achieved automotive-grade certification, and they are even less likely to be directly applied to safety functions in the rail transit sector, which has higher safety requirements. Furthermore, individually certifying or analyzing the complex failure mechanisms of these environmental sensing devices requires significant time and resources. Therefore, in obstacle detection systems, a crucial research challenge is how to utilize general data from urban rail transit signaling systems, such as location, speed, and high-definition electronic maps, to verify the correctness of sensing devices. This would allow for the avoidance of separate certification of complex sensing devices or analysis of their complex failure mechanisms, effectively treating the environmental sensing devices as a black box.

[0005] Currently, there is relatively little research data on safety verification methods for active obstacle detection systems, and related patent information is also scarce. For example, patent CN108508872 B discloses a fault detection method for an autonomous vehicle information collection system. This method uses an onboard computer to analyze and compare data collected by various radars and cameras, comparing whether the identified target objects at the same time and location match, thereby determining whether the radar or camera is malfunctioning and making corresponding decisions. The shortcomings of this patent are that it requires real-time comparison of radar data and camera data. Since cameras are greatly affected by environmental factors (strong light, weak light, and strong backlight), it cannot handle scenarios with complex environmental conditions, such as ambient lighting or interference from nearby lights, easily leading to verification failures. Furthermore, a single verification requires data from both radar and camera; if verification fails, it is difficult to determine which device is malfunctioning, affecting the system's availability. For example, patent CN 112505704A discloses a method to improve the safety of a train's autonomous intelligent sensing system. This method compares point cloud data from two heterogeneous lidar systems and determines if the difference is within a certain threshold range. If the difference is less than the threshold, the final point cloud data is output based on the two data sets. The drawback of this patent is that it requires two heterogeneous lidar systems for the determination, necessitating additional equipment, increasing system costs, and burdening system operation and maintenance. Summary of the Invention

[0006] To address the aforementioned issues, this application proposes a safety verification method, apparatus, and control system for environmental sensing devices in an obstacle detection system.

[0007] This application proposes a safety verification method for environmental sensing devices in an obstacle detection system, comprising the following steps:

[0008] Acquire raw data from environmental sensing devices;

[0009] Obtain general data for urban rail transit signaling systems;

[0010] Based on the general data of the signal system, extract the data features of the raw data from the environmental sensing device;

[0011] Compare data characteristics to determine if there is a malfunction in the environmental sensing equipment;

[0012] Determine the integrity of the raw data from environmental sensing devices.

[0013] As an optional implementation of this application, optionally, data features of the raw data from the environmental sensing device are extracted based on general data from the signal system, including:

[0014] The data type for obtaining the general data of the signal system;

[0015] Based on the data type of general data in the signal system, environmental sensing device data features corresponding to the data type are extracted from the raw data of the environmental sensing device.

[0016] The environmental sensing device data characteristics of each data type are stored.

[0017] As an optional implementation of this application, optionally, comparing data characteristics to determine whether the environmental sensing device is faulty includes:

[0018] Preset tolerance for differences;

[0019] Compare whether the data characteristics are consistent with the general data of the signal system:

[0020] If they match, proceed to the next verification step;

[0021] If there is a discrepancy, proceed to the environmental sensing device fault diagnosis step.

[0022] As an optional implementation of this application, the environmental sensing device fault judgment step may optionally include:

[0023] If the data feature is inconsistent with the general data of the signal system, then the difference value between the data feature and the general data of the signal system is calculated;

[0024] Determine whether the difference between the data feature and the general data of the signal system exceeds a preset difference tolerance:

[0025] If the difference value exceeds the preset difference tolerance, the environmental sensing device is identified as malfunctioning, and a corresponding environmental sensing device malfunction message is issued.

[0026] As an optional implementation of this application, the determination of the integrity of the original data from the environmental sensing device may include:

[0027] Acquire sensing data from environmental sensing devices deployed at preset locations in urban rail transit;

[0028] The perceived data is verified using the aforementioned data feature comparison verification method:

[0029] If the verification passes, the integrity of the original data from the environmental sensing device has been successfully verified.

[0030] Otherwise, the verification will fail.

[0031] As an optional implementation of this application, the determination of the integrity of the original data from the environmental sensing device may include:

[0032] Acquire obstacle perception data from environmental sensing devices;

[0033] The continuity of the sensed data is verified using a preset continuity verification method:

[0034] If the continuity check fails, the data does not meet the data integrity requirements.

[0035] Otherwise, the verification will succeed.

[0036] As an optional implementation of this application, it may also include:

[0037] The integrity and correctness verification cycle is calculated based on the failure rate of the sensing devices, the safety level of the obstacle detection system, and the tolerable hazard level assigned to the sensing devices.

[0038] Configure the verification period within the timer, start the timer, and complete the verification of integrity and correctness according to the verification period.

[0039] In another aspect, this application also proposes an apparatus for implementing a security verification method for the environmental sensing device of the obstacle detection system, comprising:

[0040] The raw data acquisition module is used to acquire raw data from environmental sensing devices;

[0041] A general data acquisition module is used to acquire general data from urban rail transit signaling systems.

[0042] The data feature extraction module is used to extract data features from the raw data of the environmental sensing device based on the general data of the signal system.

[0043] The fault diagnosis module is used to compare data characteristics to determine whether the environmental sensing device is faulty;

[0044] The verification module is used to determine the integrity of the raw data of the environmental sensing device and the correctness of the sensing data of the environmental sensing device at preset locations or for obstacles deployed in the urban rail.

[0045] The timing module is used to initiate integrity and correctness checks according to the verification cycle.

[0046] In another aspect, this application also proposes a control system, comprising:

[0047] processor;

[0048] Memory used to store processor-executable instructions;

[0049] The processor is configured to implement the security verification method for the environmental perception device of the obstacle detection system described above when executing the executable instructions.

[0050] Technical effects of the present invention:

[0051] This application obtains raw data from environmental sensing devices; acquires general data from urban rail transit signaling systems; extracts data features from the raw data of environmental sensing devices based on the general data of the signaling systems; compares the data features to determine whether the environmental sensing devices are faulty; and determines the integrity of the raw data of the environmental sensing devices. The sensing devices can be used as black boxes in obstacle detection safety systems, without needing to focus too much on whether the sensing devices themselves have safety certifications or complex failure mechanisms. It offers flexibility, allowing for the reuse of existing positioning and speed measurement equipment from the signaling system, saving costs; alternatively, separate positioning and speed measurement equipment can be used to further improve availability; and it does not require the sensing devices themselves to be configured for redundant verification, saving costs and installation space.

[0052] Other features and aspects of this disclosure will become clear from the following detailed description of exemplary embodiments with reference to the accompanying drawings. Attached Figure Description

[0053] The accompanying drawings, which are included in and form part of this specification, illustrate exemplary embodiments, features, and aspects of this disclosure together with the specification and serve to explain the principles of this disclosure.

[0054] Figure 1 The diagram illustrates the implementation flow of the safety verification method for the environmental sensing device of the obstacle detection system of the present invention.

[0055] Figure 2 The diagram illustrates the principle of original data integrity verification for the environmental sensing device of this invention – verification through curves and slopes. Detailed Implementation

[0056] Various exemplary embodiments, features, and aspects of this disclosure will now be described in detail with reference to the accompanying drawings. The same reference numerals in the drawings denote elements that have the same or similar functions. Although various aspects of the embodiments are shown in the drawings, they are not necessarily drawn to scale unless specifically indicated otherwise.

[0057] The term “exemplary” as used herein means “serving as an example, embodiment, or illustration.” Any embodiment illustrated herein as “exemplary” is not necessarily to be construed as superior to or better than other embodiments.

[0058] Furthermore, to better illustrate this disclosure, numerous specific details are set forth in the following detailed description. Those skilled in the art will understand that this disclosure can be practiced without certain specific details. In some instances, methods, means, components, and circuits well known to those skilled in the art have not been described in detail in order to highlight the main points of this disclosure.

[0059] This invention aims to use general data from urban rail transit signaling systems, such as location, speed, and high-definition electronic maps, to perform safety verification on the correctness and effectiveness of environmental sensing devices in obstacle detection systems. This avoids the need for separate certification of complex sensing devices or analysis of their failure mechanisms. This method can effectively improve the overall safety of obstacle detection systems, ensure their safe operation, and ultimately help them achieve safety certification in the rail transit field.

[0060] Example 1

[0061] like Figure 1 As shown, this application proposes a safety verification method for an environmental sensing device in an obstacle detection system, comprising the following steps:

[0062] 1. Acquire raw data from environmental sensing devices; Different sensing devices can be selected, including but not limited to LiDAR, millimeter-wave radar, and industrial cameras, to obtain different types of raw data. Using LiDAR or millimeter-wave radar, point cloud data can be obtained, with each point including at least two-axis or three-axis coordinates. Using industrial cameras, environmental image data can be obtained.

[0063] 2. Obtain general data from the urban rail transit signaling system; different types of data can be selected depending on the environmental sensing equipment, especially data generated by safety equipment. This can be achieved using location data and speed data from the positioning system, or fixed landmarks from high-definition electronic maps. In particular, fixed landmarks should be regularly shaped and have easily identifiable colors.

[0064] 3. Based on the general data of the signal system, extract the data features of the raw data from the environmental sensing devices; based on the different types of data selected in step two, extract the corresponding sensing device data features selected in step one. The data selected in steps one and two should be obtained within the same processing cycle, or the difference in cycles should not exceed the tolerance error.

[0065] As an optional implementation of this application, optionally, data features of the raw data from the environmental sensing device are extracted based on general data from the signal system, including:

[0066] The data type for obtaining the general data of the signal system;

[0067] Based on the data type of general data in the signal system, environmental sensing device data features corresponding to the data type are extracted from the raw data of the environmental sensing device.

[0068] The environmental sensing device data characteristics of each data type are stored.

[0069] This application extracts different data features from raw data of different types of environmental sensing devices. For example:

[0070] 1) For step one, using lidar or millimeter-wave radar:

[0071] 1.1) For the signal system location data selected in step two, for the lidar or millimeter-wave radar point cloud data in step one, convert the two-axis or three-axis coordinates and the signal system location data into the same coordinate system;

[0072] Based on the location information in step two, obtain the landmarks within the detection range of the lidar or millimeter-wave radar from the high-definition electronic map database. The detection range of the lidar or millimeter-wave radar should be determined by the detection capability range of the equipment itself and the geometric characteristics of the track that affect the detection capability. The geometric characteristics of the track that affect the detection capability include curves and slopes.

[0073] Point cloud data is divided into several datasets through clustering;

[0074] The segmented point cloud dataset is compared with the acquired markers to calculate the probability of the markers being recognized, thereby inferring the confidence level of the current location.

[0075] 1.2) For the signal system velocity data selected in step two, the point cloud data of lidar or millimeter-wave radar in step one is divided into several datasets by clustering.

[0076] The point cloud data from the previous period of the lidar or millimeter-wave radar point cloud data in step one is also divided into several datasets through clustering.

[0077] Select a suitable dataset and calculate the two-dimensional or three-dimensional velocity obtained from the data of the sensing device through differential calculation.

[0078] 2) For step one, an industrial camera is used:

[0079] Based on the location information in step two, obtain the location and size information of the markers within the detection range of the industrial camera from the electronic map. The detection range of the industrial camera should be determined by the detection capability range of the equipment itself and the geometric characteristics of the track that affect the detection capability. The track geometry features that affect the detection capability include curves and slopes.

[0080] By using coordinate transformation and distance calculation, the location of the marker is determined in the image data obtained in step one. The image is divided into several regions to find the region where the marker is located.

[0081] The shape and color of the marker are determined by feature recognition.

[0082] 4. Compare data characteristics to determine if the environmental sensing equipment is faulty; for different combinations of selections in the above steps, different judgment criteria are used respectively.

[0083] Verify the correctness of the perceived data.

[0084] As an optional implementation of this application, optionally, comparing data characteristics to determine whether the environmental sensing device is faulty includes:

[0085] Preset tolerance for differences;

[0086] Compare whether the data characteristics are consistent with the general data of the signal system:

[0087] If they match, proceed to the next verification step;

[0088] If there is a discrepancy, the process proceeds to the environmental sensing device fault diagnosis step. As an optional embodiment of this application, the environmental sensing device fault diagnosis step may optionally include:

[0089] If the data feature is inconsistent with the general data of the signal system, then the difference value between the data feature and the general data of the signal system is calculated;

[0090] Determine whether the difference between the data feature and the general data of the signal system exceeds a preset difference tolerance:

[0091] If the difference value exceeds the preset difference tolerance, the environmental sensing device is identified as malfunctioning, and a corresponding environmental sensing device malfunction message is issued.

[0092] For 1.1), compare the inferred current position and its confidence level with the difference between the position of the signal system in step three;

[0093] For 1.2), compare the inferred current velocity and its confidence level with the difference between the velocity of the signal system in step 3;

[0094] For 2), compare the differences in the size and color of the markers.

[0095] 5. Assess the integrity of the original data from the environmental sensing devices.

[0096] Since the location of markers can only be limited to the outside or below of the track, and the straight track is in the area outside the detection range of the sensing equipment, it is necessary to determine the integrity of the data before judging the correctness of the sensing data.

[0097] There are two options to choose from:

[0098] As an optional implementation of this application, the determination of the integrity of the original data from the environmental sensing device may include:

[0099] Acquire sensing data from environmental sensing devices deployed at preset locations in urban rail transit;

[0100] The perceived data is verified using the aforementioned data feature comparison verification method:

[0101] If the verification passes, the integrity of the original data from the environmental sensing device has been successfully verified.

[0102] Otherwise, the verification will fail.

[0103] As an optional implementation of this application, the determination of the integrity of the original data from the environmental sensing device may include:

[0104] Acquire obstacle perception data from environmental sensing devices;

[0105] The continuity of the sensed data is verified using a preset continuity verification method:

[0106] If the continuity check fails, the data does not meet the data integrity requirements.

[0107] Otherwise, the verification will succeed.

[0108] 5.1) Regularly check the system using ramps and curves;

[0109] Ramps and curves allow markers to be positioned within the detection range of the sensing device. By periodically collecting sensing data from ramps and curves, and applying the steps one through four for verification, if the comparison in step four is successful, the sensing data covers the entire detection area of ​​the sensing device, and the data integrity verification is successful. Otherwise, the data integrity verification fails.

[0110] 5.2) Verify data continuity.

[0111] Train obstacles can be categorized into two types: those located in front of the train and those that cross over onto the track from the side. Regardless of the type, the obstacle should exhibit continuous change. If the corresponding sensing data suddenly disappears within the detected area or the amount of sensing data decreases by more than a certain percentage, the continuity check fails, and the data does not meet the integrity requirement. Otherwise, the data integrity check succeeds.

[0112] This embodiment does not limit the verification of the correctness of the perceived data.

[0113] The verification of completeness and correctness can be carried out using a periodic calculation model.

[0114] 6. As an optional embodiment of this application, it may also include:

[0115] The integrity and correctness verification cycle is calculated based on the failure rate of the sensing devices, the safety level of the obstacle detection system, and the tolerable hazard level assigned to the sensing devices.

[0116] Configure the verification period within the timer, start the timer, and complete the verification of integrity and correctness according to the verification period.

[0117] The frequency of accuracy and integrity verification depends on the failure rate of the sensing devices, the safety level of the obstacle detection system, and the tolerable hazard level assigned to the sensing devices (all set by the user or directly identifiable). The verification cycle must at least ensure that the sensing devices experience fewer failures than the specified hazard level during the period.

[0118] The verification cycle is configured using a timer. This timer is started whenever an integrity verification is completed. When the timer exceeds the required verification cycle, a data integrity verification failure is output to the fault detection system.

[0119] The following is a specific implementation plan to demonstrate and illustrate the above technical principles and verification methods:

[0120] The safety verification method for environmental sensing equipment in an obstacle detection system includes the following steps:

[0121] Step 1: Acquire raw data from environmental sensing devices

[0122] This example uses a combination of a LiDAR (Light Detection and Ranging) system without security certification and an industrial camera for environmental sensing. The LiDAR is used to identify obstacles and distances, while the industrial camera is used to identify traffic signals. The overall safety integrity level requirement for the obstacle detection system is SIL2.

[0123] The performance parameters of the lidar are: the detection range along the straight track is (Sl, Sh).

[0124] The train is running along track GD0.

[0125] Point cloud data from lidar is collected. For a specific collection period ti, each point D contains three-axis coordinate information (Xdi, Ydi, Zdi). The point cloud data collected in period ti is the dataset {(Xdi, Ydi, Zdi)}.

[0126] Image data from industrial cameras is collected. For a specific acquisition period ti, each image contains two-dimensional image information of the environment in front of the train. Image information in the middle of the period is selected to form an image pixel dataset A.

[0127] Step 2: Obtain general data for the signal system

[0128] This example uses a combination of fixed landmarks on electronic maps and location data from a positioning system to provide a basis for the verification of sensing devices.

[0129] Data is collected from the positioning system. For a specific collection period ti, the three-dimensional position information of the positioning system is (Xwi, Ywi, Zwi), and the mileage along the current orbit GD0 direction is S0.

[0130] Read the electronic map information, calculate the visual limit affected by curves and slopes based on the three-dimensional location information, and take the limit mileage along the current track GD0 direction as S1. Take the smaller of S1 and S0+Sh as S2.

[0131] Step 3: Extract the data features of the sensing devices based on the signal system data.

[0132] According to (3.1.1):

[0133] (1) For the acquisition period ti, the dataset {(Xdi, Ydi, Zdi)} in step one and the three-dimensional position (Xwi, Ywi, Zwi) in step two are transformed into data {(Xdi', Ydi', Zdi')} and (Xwi', Ywi', Zwi') in the same coordinate system, so as to calculate the installation position (Xlwi', Ylwi', Zlwi') of the lidar.

[0134] (2) Obtain all landmark information from S0+Sl to S2 from the electronic map, including the three-dimensional positions of the landmark center points (Xb1, Yb1, Zb1), (Xb2, Yb2, Zb2), ..., (Xbn, Ybn, Zbn). Transform the three-dimensional positions of the landmark center points to the same coordinate system as in (1) (Xb1', Yb1', Zb1'), (Xb2', Yb2', Zb2'), ..., (Xbn', Ybn', Zbn').

[0135] (3) The point cloud data {(Xdi', Ydi', Zdi')} is divided into several datasets through clustering. The centroid of each dataset is calculated and denoted as (Xf1', Yf1', Zf1'), (Xf2', Yf2', Zf2'), ..., (Xfm', Yfm', Zfm'). In particular, the process of clustering and dividing the datasets can be optimized by using the features of the markers to improve the probability that the point cloud data of the markers is correctly classified.

[0136] (4) Compare the classified point cloud data centers (Xf1', Yf1', Zf1'), (Xf2', Yf2', Zf2'), ..., (Xfm', Yfm', Zfm') with the marker centers (Xb1', Yb1', Zb1'), (Xb2', Yb2', Zb2'), ..., (Xbn', Ybn', Zbn'), calculate the probability that the markers are correctly identified, and select... Select *s* markers with a recognition probability greater than a certain level, denoted as (Xfs1', Yfs1', Zfs1'), (Xfs2', Yfs2', Zfs2'), ..., (Xfss', Yfss', Zfss') and (Xbs1', Ybs1', Zbs1'), (Xbs2', Ybs2', Zbs2'), ..., (Xbss', Ybss', Zbss'). Then calculate the confidence level for the current lidar position as (Xlwi', Ylwi', Zlwi').

[0137] According to 3.2):

[0138] (1) Obtain all landmark information from S0+S1 to S2 from the electronic map. This example uses rectangular landmarks as an example. The three-dimensional positions of the landmark points are as follows:

[0139] {(Xb 11 Yb 11 Zb 11 ), (Xb 12 Yb 12 Zb 12 ), (Xb 13 Yb 13 Zb 13 ), (Xb 14 Yb 14 Zb 14 )},{(Xb 21 Yb 21 Zb 21 ), (Xb 22 Yb 22 Zb 22 ), (Xb 23 Yb 23 Zb 23 ), (Xb 24 Yb 24 Zb 24 )},…,{(Xb n1 Yb n1 Zb n1 ), (Xb n2 Yb n2 Zb n2 ), (Xb n3 Ybn3 Zb n3 ), (Xb n4 Yb n4 Zb n4 The three-dimensional positions of the marker points and the three-dimensional positions (Xwi, Ywi, Zwi) of the positioning system in step two are converted into data in the same coordinate system.

[0140] {(Xb 11 ',Yb 11 ',Zb 11 '), (Xb 12 ',Yb 12 ',Zb 12 '), (Xb 13 ',Yb 13 ',Zb 13 '), (Xb 14 ',Yb 14 ',Zb 14 ')},{(Xb 21 ',Yb 21 ',Zb 21 '), (Xb 22 ',Yb 22 ',Zb 22 '), (Xb 23 ',Yb 23 ',Zb 23 '), (Xb 24 ',Yb 24 ',Zb 24 ')},…,{(Xb n1 ',Yb n1 ',Zb n1 '), (Xb n2 ',Yb n2 ',Zb n2 '), (Xb n3 ',Yb n3 ',Zb n3 '), (Xb n4 ',Yb n4 ',Zb n4 The camera mounting positions (Xxwi', Yxwi', Zxwi') are calculated by using (Xwi', Ywi', Zwi') and (Xwi', Ywi', Zwi').

[0141] (2) Based on the camera installation location and imaging parameters, select any marker point and calculate the position of the marker point in the image:

[0142] {(Xbt n1 Ybt n1 ), (Xbt n2 Ybtn2 ), (Xbt n3 Ybt n3 ), (Xbt n4 Ybt n4 )};

[0143] The image pixel dataset is filtered to segment out the smallest region covering the above points, forming the target pixel dataset A1;

[0144] (3) Perform feature recognition on A1, calculate the shape of the target marker, compare it with the shape determined by the position of the marker in the image in (2), and determine the probability that the two shapes are the same. If a certain level is met, extract and calculate the main color of the target marker.

[0145] (4) Repeat (2) and (3) until the main color data of all target markers that meet the shape requirements are calculated;

[0146] Step four: Compare data characteristics to determine if the sensing device is faulty.

[0147] For step three above, the position confidence calculated in 3.1.1 is evaluated to see if it meets the specified threshold. If it does not, the lidar is considered to be malfunctioning and needs to be output to the obstacle detection system for appropriate protection. If it meets the threshold, the lidar is considered to be functioning normally, but the detection distance can be output to the obstacle detection system based on the farthest distance of the identified landmarks in 3.1.1 to improve the security of the obstacle recognition function.

[0148] In step three above, the main color of the target marker that satisfies the shape calculated in 3.2 is compared with the color marked on the electronic map. If the error exceeds the specified upper limit, the camera is considered to have malfunctioned and needs to be output to the obstacle detection system for appropriate protection. Since the camera is greatly affected by ambient light, the upper limit of the error can be appropriately increased to reduce false malfunctions. In particular, if the camera's color information is not needed in practical applications, the color verification process in step three (3.2) above can be omitted.

[0149] Step 5: Determine the integrity of the original data from the sensing device.

[0150] This example uses periodic checks on ramps and curves to determine the integrity of the original data.

[0151] Taking a curve as an example, the design places a marker W at a designated location on the curve. When the train reaches this specific location, the marker is positioned in a relatively central area within the detection range of the sensing equipment, as shown in the attached diagram. Figure 2 As shown in the middle.

[0152] Repeat steps one through four. If the marker in step three includes the special marker W for the curve, and the selected marker position also includes the position of W, and the sensing device does not malfunction as determined in step four, then an integrity verification is considered to have passed.

[0153] Set a timer to start each time an integrity check is completed. When the timer exceeds the required check period, output "Data integrity check failed" to the fault detection system.

[0154] Step 6: Calculate the verification period.

[0155] The LiDAR selected in this example has a failure rate of 26.3 x 10^6. -6 The entire fault detection system needs to achieve a Safety Integrity Level of 2 (SIL2) per hour, with a corresponding THR of less than 10. -6 Considering the system configuration, the TFFR of the lidar data acquisition function should be less than 10 per hour. -7 / Hour.

[0156] Verification period T<10 -7 / 26.3x10 -6 = 13.7 seconds.

[0157] It should be noted that although radar equipment and industrial cameras have been used as examples to illustrate the acquisition of raw sensing data, those skilled in the art will understand that this disclosure is not limited to this. In fact, users can flexibly set the type of sensing device to acquire corresponding data according to the actual application scenario, as long as the technical functions of this application can be achieved by following the above technical methods.

[0158] Example 2

[0159] Based on the implementation principle of Embodiment 1, an apparatus for implementing the safety verification method of the environmental sensing device of the obstacle detection system is also proposed, comprising:

[0160] The raw data acquisition module is used to acquire raw data from environmental sensing devices;

[0161] A general data acquisition module is used to acquire general data from urban rail transit signaling systems.

[0162] The data feature extraction module is used to extract data features from the raw data of the environmental sensing device based on the general data of the signal system.

[0163] The fault diagnosis module is used to compare data characteristics to determine whether the environmental sensing device is faulty;

[0164] The verification module is used to determine the integrity of the raw data of the environmental sensing device and the correctness of the sensing data of the environmental sensing device at preset locations or for obstacles deployed in the urban rail.

[0165] The timing module is used to initiate integrity and correctness checks according to the verification cycle.

[0166] For details on the functions and interaction principles of the above modules, please refer to the description in Example 1.

[0167] Obviously, those skilled in the art should understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, it can include the processes of the embodiments of the control methods described above. The modules or steps of the present invention described above can be implemented using general-purpose computing devices. They can be centralized on a single computing device or distributed across a network of multiple computing devices. Optionally, they can be implemented using computer-executable program code, thereby storing them in a storage device for execution by a computing device, or fabricating them separately as individual integrated circuit modules, or fabricating multiple modules or steps into a single integrated circuit module. Thus, the present invention is not limited to any specific hardware and software combination.

[0168] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, it can include the processes of the embodiments of the control methods described above. The storage medium can be a magnetic disk, optical disk, read-only memory (ROM), random access memory (RAM), flash memory, hard disk drive (HDD), or solid-state drive (SSD), etc.; the storage medium can also include combinations of the above types of memory.

[0169] Example 3

[0170] Furthermore, this application also proposes a control system, comprising:

[0171] processor;

[0172] Memory used to store processor-executable instructions;

[0173] The processor is configured to implement the security verification method for the environmental perception device of the obstacle detection system described above when executing the executable instructions.

[0174] This disclosure discloses an embodiment of a control system including a processor and a memory for storing processor-executable instructions. The processor is configured to implement, when executing the executable instructions, a security verification method for an environmental perception device in an obstacle detection system as described above.

[0175] It should be noted here that the number of processors can be one or more. Furthermore, the control system in this embodiment may also include input devices and output devices. The processors, memory, input devices, and output devices can be connected via a bus or other means, without specific limitations herein.

[0176] As a computer-readable storage medium, the memory can be used to store software programs, computer-executable programs, and various modules, such as the program or module corresponding to the safety verification method of an environmental sensing device in an obstacle detection system according to an embodiment of this disclosure. The processor executes various functional applications and data processing of the control system by running the software programs or modules stored in the memory.

[0177] Input devices can be used to receive input digital numbers or signals. These signals can be key signals related to user settings and function control of the device / terminal / server. Output devices can include display devices such as screens.

[0178] The various embodiments of this disclosure have been described above. These descriptions are exemplary and not exhaustive, and are not limited to the disclosed embodiments. Many modifications and variations will be apparent to those skilled in the art without departing from the scope and spirit of the described embodiments. The terminology used herein is chosen to best explain the principles, practical applications, or technical improvements to the technology in the market, or to enable others skilled in the art to understand the embodiments disclosed herein.

Claims

1. A safety verification method for environmental sensing equipment in an obstacle detection system, applied to an urban rail transit obstacle detection system, characterized in that, The method of performing security verification on environmental sensing devices as black boxes eliminates the need for security certification or failure mechanism analysis of the devices themselves. The method includes the following steps: S1 Multi-source raw data synchronous acquisition: Acquire raw data from at least one environmental sensing device in the obstacle detection system, including lidar, millimeter-wave radar, and industrial cameras. The raw data includes point cloud data from lidar / millimeter-wave radar and environmental image data from industrial cameras, wherein each sampling point of the point cloud data contains at least two-axis or three-axis coordinate information; Synchronously acquire general data from the urban rail transit signaling system, including real-time train positioning data, real-time speed data, and high-definition electronic map data. The high-definition electronic map data includes information on fixed landmarks at preset locations along the track and geometric feature data of curves and slopes of the track line, and the difference in acquisition cycle between the general data and the raw data from the environmental sensing devices does not exceed a preset tolerance error. S2 Coordinate System Adaptation and Differentiated Data Feature Extraction: The data type of the general data of the signal system is obtained, and the original data of the environmental sensing device and the general data of the signal system are converted to the same coordinate system. Environmental sensing device data features corresponding to the data type are extracted from the original data of the environmental sensing device, and the environmental sensing device data features corresponding to each data type are saved. Specifically, for point cloud data from lidar / millimeter-wave radar, combined with real-time train positioning data and high-definition electronic map data, point cloud clustering features of fixed landmarks within the detection range, train speed features calculated based on point cloud differences, and current train position confidence features are extracted. For environmental image data from industrial cameras, combined with real-time train positioning data and high-definition electronic map data, shape and color features of fixed landmarks within the detection range are extracted. S3 Device Fault Judgment with Differential Tolerance: A differential tolerance matching the type of environmental sensing device is preset. The data characteristics are compared with the general data of the signal system. If they are consistent, the data integrity verification step is initiated. If they are inconsistent, the difference between the data characteristics and the general data of the signal system is calculated, and it is determined whether the difference exceeds the preset differential tolerance. If the difference exceeds the preset differential tolerance, the environmental sensing device is identified as having malfunctioned, and the corresponding environmental sensing device fault information is issued. If the difference does not exceed the preset differential tolerance, the data integrity verification step is initiated. S4 Data Integrity Verification: Determining the integrity of raw data from environmental sensing devices includes: Acquire sensing data from environmental sensing devices deployed at preset locations in urban rail transit; The perceived data is verified using the aforementioned data feature comparison verification method: If the verification passes, the integrity of the original data from the environmental sensing device has been successfully verified. Otherwise, the verification will fail; or Determining the integrity of raw data from environmental sensing devices includes: Acquire obstacle perception data from environmental sensing devices; The continuity of the sensed data is verified using a preset continuity verification method: If the continuity check fails, the data does not meet the data integrity requirements. Otherwise, the verification was successful; S5. Verification cycle control for matching security level: Based on the failure rate of environmental sensing equipment, the security integrity level of obstacle detection system, and the tolerable hazard level allocated to environmental sensing equipment, calculate the verification cycle for data correctness and integrity; configure the verification cycle in a timer, start the timer, and trigger the verification process of steps S1-S4 to be executed cyclically according to the verification cycle through the timer. When the timer exceeds the verification cycle and fails to complete the valid integrity verification, output data integrity verification failure information to the obstacle detection system.

2. An apparatus for implementing the safety verification method of the environmental sensing device of the obstacle detection system as described in claim 1, characterized in that, include: The raw data acquisition module is used to acquire raw data from environmental sensing devices; A general data acquisition module is used to acquire general data from urban rail transit signaling systems. The data feature extraction module is used to extract data features from the raw data of the environmental sensing device based on the general data of the signal system. The fault diagnosis module is used to compare data characteristics to determine whether the environmental sensing device is faulty; The verification module is used to determine the integrity of the raw data of the environmental sensing device and the correctness of the sensing data of the environmental sensing device at preset locations or for obstacles deployed in the urban rail. The timing module is used to initiate integrity and correctness checks according to the verification cycle.

3. A control system, characterized in that, include: processor; Memory used to store processor-executable instructions; The processor is configured to implement the security verification method for the environmental perception device of the obstacle detection system as described in claim 1 when executing the executable instructions.

Images