A sensor position calibration system for clearance recognition

Abstract

The present invention relates to the technical field of clearance recognition, and discloses a sensor position calibration system for clearance recognition, including a data acquisition module that acquires position information data collected by each sensor; a data processing module that acquires the position information data and respectively identifies position correlation features; a data comparison module that performs consistency verification, stability verification, and redundant information verification according to the position correlation features, and sends the verification result data to a credibility evaluation module; a credibility evaluation module that calculates the credibility scores of each sensor according to a preset credibility scoring strategy, where the lower the credibility score, the more likely the corresponding sensor is abnormal; a calibration warning module that makes a judgment based on the distribution of the credibility scores. If the deviation degree between the credibility score of one sensor and the average value of the credibility scores of other sensors is greater than a preset scoring deviation threshold, then this sensor is used as a sensor to be verified and a verification instruction is generated.

Description

Technical Field

[0001] The present invention relates to the technical field of clearance recognition, and particularly to a sensor position calibration system for clearance recognition. Background Art

[0002] In fields such as railway transportation, clearance recognition is crucial, which can ensure the safe operation of vehicles such as trains. Laser clearance detection is used for vehicle clearance detection, which can determine whether the shape of the vehicle exceeds the specified safety range, thereby avoiding accidents such as collisions and scratches during vehicle operation, and ensuring that the vehicle can safely pass through structures with limited clearances such as tunnels and bridges. It is an essential process applied when new-built (or after overhaul) subway vehicles and engineering vehicles detect and accept the contour lines of the vehicles. The laser clearance detection device adopts a non-contact detection method, which can quickly collect data when the vehicle passes by and automatically complete the comparison with the standard clearance without manual intervention, greatly improving the detection efficiency.

[0003] However, due to environmental factors, equipment aging and other reasons, the position of the sensor may shift, resulting in inaccurate collected data, and further affecting the accuracy of clearance recognition. At present, existing sensor calibration systems often lack a comprehensive and accurate calibration mechanism and cannot detect and handle abnormal situations of sensors in a timely manner, which poses a potential threat to the safety of industries such as railway transportation. Therefore, an efficient and reliable sensor position calibration system for clearance recognition is needed. Summary of the Invention

[0004] Aiming at the deficiencies of the existing technology, the purpose of the present invention is to provide a sensor position calibration system for clearance recognition, which can accurately evaluate the credibility of sensors by comprehensively processing and verifying the data collected by the sensors, timely detect and handle sensors that may be abnormal, thereby improving the accuracy and reliability of clearance recognition.

[0005] To achieve the above purpose, the present invention provides the following technical solutions:

[0006] A sensor position calibration system for clearance recognition, comprising:

[0007] A data acquisition module, which acquires position information data collected by each sensor;

[0008] A data processing module, which acquires the position information data and respectively identifies position association features, and the position association features reflect the features that conform to the corresponding relationship in the position information data collected by different sensors;

[0009] A data comparison module, which performs consistency verification, stability verification and redundant information verification according to the position association features, and sends the verification result data to the credibility evaluation module,

[0010] A credibility evaluation module calculates the credibility scores of each sensor according to a preset credibility scoring strategy. The lower the credibility score, the higher the probability that the corresponding sensor is abnormal.

[0011] A calibration warning module makes a judgment based on the distribution of the credibility scores. If the deviation degree between the credibility score of one sensor and the average value of the credibility scores of other sensors is greater than a preset scoring deviation threshold, then this sensor is regarded as a sensor to be verified and a verification instruction is generated.

[0012] Further, the sensors include a number of laser profilometers and a number of area array cameras. The laser profilometers acquire three-dimensional point cloud data of the train, and the area array cameras acquire train image information of the train. The data processing module identifies key point cloud features and point cloud feature parameters in the three-dimensional point cloud data. The data processing module identifies key image features and image feature parameters in the train image information, and judges the similarity between the point cloud feature parameters and the image feature parameters. If the similarity meets a preset feature similarity threshold, then the corresponding key point cloud features and image key features / a number of key point cloud features / a number of image key features are used as the position association features. The position association features include the corner points and contour points of the vehicle.

[0013] Further, the calculation formula of the credibility scoring strategy is specifically:

[0014] ,

[0015] where S represents the credibility evaluation score, and A represents the consistency score. represents the weight parameter of the consistency score, B represents the stability score. represents the weight parameter of the stability score, C represents the redundant information verification score. represents the weight parameter of the redundant information verification score.

[0016] Further, the consistency verification steps include:

[0017] Define the current sensor as the target sensor, and obtain at least one target verification point based on the position association features.

[0018] Obtain the value of the target verification point in the world coordinate system and define it as the target verification value.

[0019] Define the sensor with key position features with respect to the target sensor as the reference sensor, and obtain the corresponding reference verification point based on the target verification point.

[0020] Obtain the value of the reference verification point in the world coordinate system and define it as the reference verification value.

[0021] Further, the specific formula for the consistency score is as follows:

[0022] ,

[0023] ,

[0024] ,

[0025] D represents the total consistency deviation of the target sensor, m represents the number of reference sensors, where the j-th position-associated feature has verification points in the i-th measurement, E is the k-th target verification value of the key feature at the j-th position of the target sensor in the i-th measurement, F is the reference verification value of the -th reference sensor compared with the target sensor under the same measurement conditions, W represents the weight of the verification times, represents the total number of verifications of all reference sensor devices in the corresponding measurement.

[0026] Further, the stability verification steps include:

[0027] Obtain position information data and extract key dimensions, where the key dimensions include width, length, and surface curvature.

[0028] Obtain historical dimensions and compare them with the key dimensions.

[0029] Further, the specific formula for the stability score is as follows:

[0030] ,

[0031] ,

[0032] where V represents the variance between the key dimension and the historical dimension, q - 1 represents the number of historical dimensions, represents the parameter value of the historical dimension, represents the parameter value of the corresponding key dimension.

[0033] Further, the redundant information verification steps include:

[0034] Obtain depth data based on the three-dimensional point cloud data and perform data preprocessing on the depth data;

[0035] Obtain a texture image based on the train image information and perform image preprocessing;

[0036] Convert the coordinates of the depth data to the same coordinate system as the texture image so that there is a corresponding relationship;

[0037] Calculate the correlation between the depth data and the texture data. The greater the correlation, the higher the degree of fit and the higher the redundancy information verification score.

[0038] Further, the credibility evaluation module is configured with a weight dynamic adjustment strategy, and the formula of the weight dynamic adjustment strategy is specifically:

[0039] ,

[0040] ,

[0041] where In, v takes values of 1, 2, and 3, representing the weight parameters of the consistency score, the weight parameters of the stability score, and the weight parameters of the redundancy information verification score respectively. represents the environmental parameter vector. represents the basic weight. represents the influence coefficient of the environmental parameter on the weight. represents the quantization function of the influence of the environmental parameter on the weight.

[0042] Further, the credibility evaluation module is configured with a stability weight parameter adjustment sub-strategy, including obtaining historical over-limit data. The historical over-limit data includes the train over-limit position. Judge the train over-limit position and the detection range of the sensor. If the number of train over-limit anomalies appearing within the detection range corresponding to the target sensor is more, the basic weight of the stability score weight parameter is smaller. If the number of times the target sensor is verified is more, the weight parameter of the consistency score is larger. If the number of sensors with position association characteristics with the target sensor is less, the basic weight of the redundancy information verification score weight parameter is larger.

[0043] Advantages of the present invention:

[0044] The present invention verifies each other through the sensors set in the vehicle clearance recognition system, judges the association between the data characteristics collected by each sensor device by identifying the position association characteristics, uses the sensors to verify each other, calculates the credibility scores respectively, and sets several evaluation items for the credibility scores, including the consistency score, the stability score, and the redundancy information verification score. If the deviation degree of the average value of the credibility score of one sensor device is greater than the preset score deviation threshold, it indicates that the deviation degree of this sensor device from other devices is high and it is more likely to generate errors. Therefore, it is judged whether this device needs to be verified and calibrated to ensure the detection accuracy of each sensor; the system setting of the present invention can perform self-detection during the regular detection process, judge the credibility score and output a verification instruction, which is convenient for timely discovery of faults, has a high degree of automation and improves the reliability of sensor detection. Description of the Drawings

[0045] Figure 1 It is a schematic diagram of the system architecture of the present invention;

[0046] Figure 2 It is a schematic flowchart of the consistency verification step in the present invention;

[0047] Figure 3 It is a schematic flowchart of the stability verification step in the present invention;

[0048] Figure 4 It is a schematic flowchart of the redundant information verification step in the present invention. Detailed implementation manners

[0049] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0050] It should be noted that when a component is referred to as being "fixed to" another component, it can be directly on the other component or there may also be an intermediate component. When a component is considered to be "connected to" another component, it can be directly connected to the other component or there may be an intermediate component at the same time. When a component is considered to be "disposed on" another component, it can be directly disposed on the other component or there may be an intermediate component at the same time. The terms "vertical", "horizontal", "left", "right" and similar expressions used herein are only for the purpose of illustration.

[0051] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the technical field to which the present invention belongs. The terms used in the description of the present invention herein are only for the purpose of describing specific embodiments, and are not intended to limit the present invention. The term "and / or" used herein includes any and all combinations of one or more of the related listed items.

[0052] As Figures 1 to 4 shown, a sensor position calibration system for limit recognition in this embodiment includes:

[0053] A data acquisition module that acquires position information data collected by each sensor;

[0054] A data processing module that acquires the position information data and respectively identifies position association features, where the position association features reflect the features that conform to the corresponding relationship in the position information data collected by different sensors;

[0055] The data comparison module performs consistency verification, stability verification, and redundant information verification based on the position correlation features, and sends the verification result data to the credibility evaluation module.

[0056] The credibility evaluation module calculates the credibility scores of each sensor according to the preset credibility scoring strategy. The lower the credibility score, the more likely the corresponding sensor is abnormal.

[0057] The calibration warning module makes a judgment based on the distribution of the credibility scores. If the deviation degree of the credibility score of one sensor from the average value of the credibility scores of other sensors is greater than the preset scoring deviation threshold, then this sensor is used as the sensor to be verified and a verification instruction is generated.

[0058] Multiple evaluation items can be set for the credibility score. For example, in this application, consistency, stability, and redundancy verification items are set. Consistency: Compare the same geometric feature parameters extracted by two devices and calculate their deviation degree. The smaller the deviation, the higher the consistency of the measurement results of the two devices and the higher the credibility. Stability index: Analyze the fluctuation of the measurement data of each device during multiple measurements. The smaller the data fluctuation, the better the stability of the device measurement and the higher the credibility. Redundant information verification index: Utilize the complementarity of the measurement information of two devices, such as the texture information of the area array camera and the depth information of the three-dimensional laser profiler, for mutual verification. The more the verification result conforms to the expectation, the higher the credibility of the device measurement.

[0059] The following is an introduction according to the modules of the sensor position calibration system for boundary recognition:

[0060] 1. Data acquisition module

[0061] The function of this data acquisition module is to obtain the position information data collected by each sensor. For example, in the railway boundary recognition scenario, the sensors may include several laser profilers and several area array cameras. The laser profiler can obtain the three-dimensional point cloud data of the train, and the area array camera can obtain the train image information of the train. Suppose multiple laser profilers and cameras are installed at a railway tunnel entrance. The laser profilers scan the train in real time to generate the three-dimensional point cloud data of the train surface, and the area array cameras take pictures of the train. The data acquisition module collects these data.

[0062] 2. Data processing module

[0063] This module obtains location information data and respectively identifies location-related features. Location-related features reflect the features that conform to the corresponding relationships in the location information data collected by different sensors. Specifically, the data processing module identifies key point cloud features and point cloud feature parameters in the three-dimensional point cloud data, and identifies key image features and image feature parameters in the train image information. For example, in the three-dimensional point cloud data of the train, key point cloud features such as the corner points and contour points of the train are identified, and their point cloud feature parameters such as coordinates and angles are calculated; in the train image information, key image features such as the corner points and contour points of the train are also identified, and image feature parameters are calculated. Then, the similarity between the point cloud feature parameters and the image feature parameters is judged. If the similarity meets the preset feature similarity threshold, the corresponding key point cloud features and key image features / several key point cloud features / several key image features are used as location-related features. For example, when the similarity between the coordinates of the corner points of the train head in the point cloud data and the coordinates of the corresponding corner points in the image is more than 90%, these two corner points are used as location-related features.

[0064] 3. Data comparison module

[0065] This module performs consistency verification, stability verification, and redundant information verification based on the location-related features, and sends the verification result data to the credibility evaluation module.

[0066] Consistency verification: Consistency verification compares the same geometric feature parameters extracted by two devices and calculates their deviation degree. The smaller the deviation, the higher the consistency and credibility of the measurement results of the two devices.

[0067] The steps of consistency verification include:

[0068] Define the current sensor as the target sensor, and obtain at least one target verification point based on the location-related features.

[0069] Obtain the value of the target verification point in the world coordinate system and define it as the target verification value.

[0070] Define the sensor with key location features relative to the target sensor as the reference sensor, and obtain the corresponding reference verification point based on the target verification point.

[0071] Obtain the value of the reference verification point in the world coordinate system and define it as the reference verification value.

[0072] The specific formula for the consistency score is:

[0073] ,

[0074] ,

[0075] ,

[0076] Let D denote the total consistency deviation of the target sensor, and m denote the number of reference sensors. Among them, the j-th position-associated feature has verification points in the i-th measurement. Let E be the k-th target verification value of the key feature at the j-th position of the target sensor in the i-th measurement, F be the reference verification value of the l-th reference sensor compared with the target sensor under the same measurement conditions, and W denote the weight of the number of verifications. denotes the total number of verifications of all reference sensor devices in the corresponding measurement. The smaller the deviation and the more times it is verified, the higher the consistency score, which also reflects the reliability of the device as a reference for other devices.

[0077] Stability verification: During multiple measurements, analyze the fluctuation of the measurement data of each device. The smaller the data fluctuation, the better the stability of the device measurement and the higher the credibility.

[0078] The stability verification steps include:

[0079] Obtain position information data and extract key dimensions, where the key dimensions include width, length, and surface curvature.

[0080] Obtain historical dimensions and compare them with the key dimensions.

[0081] The specific calculation formula for the stability score is:

[0082] ,

[0083] ,

[0084] where V represents the variance between the key dimension and the historical dimension, q - 1 represents the number of historical dimensions, represents the parameter value of the historical dimension, represents the parameter value of the corresponding key dimension.

[0085] Redundant information verification: Redundant information verification utilizes the complementarity of the measurement information of two devices, such as the texture information of an area array camera and the depth information of a three-dimensional laser profiler, for mutual verification. The more the verification result conforms to the expectation, the higher the credibility of the device measurement.

[0086] The redundant information verification steps include:

[0087] Obtain depth data based on three-dimensional point cloud data and perform data preprocessing on the depth data;

[0088] Obtain texture images based on train image information and perform image preprocessing;

[0089] Convert the coordinates of the depth data to the same coordinate system as the texture image so that there is a corresponding relationship.

[0090] Calculate the correlation between the depth data and the texture data. The greater the correlation, the higher the degree of fit and the higher the redundancy information verification score.

[0091] Calculate the local surface normal vector of the depth data, which can reflect the surface orientation information. The depth change rate can also be extracted, that is, the difference in depth values between adjacent points, to characterize the surface curvature. For example, when dealing with the surface shape of a train, the surface normal vector can indicate the inclination direction of the car body surface, and the depth change rate can reflect the undulation degree of the surface.

[0092] For the texture image, extract Scale-Invariant Feature Transform (SIFT) features, Speeded-Up Robust Features (SURF) features, etc. These features are invariant to image rotation, scaling, and illumination changes and can effectively describe the uniqueness of the texture.

[0093] Use the extracted features for matching. For example, use data structures such as KD-trees to quickly match the feature points of the depth data and the texture data. For example, match the surface normal vector feature points in the depth data with the SIFT feature points in the texture data to find the corresponding relationship at the feature level.

[0094] Calculate the correlation between the depth values of the matching point pairs and the texture feature intensity. For example, count whether the texture feature intensity corresponding to the points with larger depth values shows a certain regular change among a certain number of matching point pairs. The consistency of the change trends of the depth data and the texture data in the same region can also be calculated, such as calculating the consistency of the gradient directions in a certain local region of the two, and evaluating the degree of fit between the two through a quantitative index.

[0095] 4. Credibility Evaluation Module

[0096] Calculate the credibility scores of each sensor according to the preset credibility scoring strategy. The lower the credibility score, the more likely the corresponding sensor is abnormal. The specific calculation formula of the credibility scoring strategy is:

[0097] ,

[0098] where S represents the credibility evaluation score, A represents the consistency score, represents the weight parameter of the consistency score, B represents the stability score, represents the weight parameter of the stability score, C represents the redundancy information verification score, represents the weight parameter of the redundancy information verification score.

[0099] The credibility evaluation module is configured with a weight dynamic adjustment strategy, and the formula of the weight dynamic adjustment strategy is specifically as follows:

[0100] ,

[0101] ,

[0102] where in which v takes values of 1, 2, and 3, representing the weight parameters of the consistency score, the weight parameters of the stability score, and the weight parameters of the redundant information verification score respectively, represents the environmental parameter vector, represents the basic weight, represents the influence coefficient of the preset environmental parameters on the weight, represents the quantization function of the influence of environmental parameters on the weight. For example, if the environmental vibration is large, the stability score may be more important, and the weight of stability should be increased; or in the case of insufficient light, redundant information: such as the correlation between depth data and texture may be more critical, and the weight of redundant information needs to be increased.

[0103] For example, a total of three environmental parameters are introduced this time, including light intensity, vibration, and humidity. Corresponding photosensitive sensors, accelerometers, and humidity sensors need to be set. Among them, light intensity will affect the quality of texture data, vibration will affect the stability of the point cloud, and humidity will affect the laser reflectivity. represents the quantization function of the influence of environmental parameters on the weight, which is a preset function. Specific examples are given below:

[0104] is set as the quantization function of light intensity, ;

[0105] is set as the quantization function of vibration intensity, ;

[0106] is set as the quantization function of humidity, ;

[0107] where represents the weight parameter of light intensity, represents the light intensity, () is specifically a trigonometric function formula, represents the weight parameter of vibration intensity, represents the vibration intensity, represents the weight parameter of humidity, represents the humidity.

[0108] The credibility evaluation module is configured with a stability weight parameter adjustment sub-strategy, including obtaining historical over-limit data, where the historical over-limit data includes the train over-limit position, determining the train over-limit position and the detection range of the sensor. If the number of train over-limit anomalies occurring within the detection range corresponding to the target sensor is larger, the base weight of the stability score weight parameter is smaller. If the number of times the target sensor is verified is larger, the weight parameter of the consistency score is larger. If the number of sensors with a position association feature with the target sensor is smaller, the base weight of the redundant information verification score weight parameter is larger.

[0109] 5. Calibration warning module

[0110] Based on the distribution of the credibility scores, if the deviation degree between the credibility score of one sensor and the average value of the credibility scores of other sensors is greater than the preset score deviation threshold, then this sensor is regarded as a sensor to be verified and a verification instruction is generated. For example, the preset score deviation threshold is 20%. When the credibility score of a certain laser profiler deviates from the average value of the credibility scores of other laser profilers and cameras by more than 20%, this laser profiler is regarded as a sensor to be verified, and the system generates a verification instruction to prompt the staff to further check and calibrate this sensor.

[0111] Working principle:

[0112] The operating principle of this system is introduced according to the work process:

[0113] (I) Data acquisition stage

[0114] Data is collected using different types of sensors. In this application, a matrix camera and a laser profiler are specifically used for measurement. The data acquisition module collects the train three-dimensional point cloud data collected by the laser profiler and the train image information collected by the matrix camera in real time. For example, every 1 second, the laser profiler scans the passing train once to generate the three-dimensional point cloud data of the train surface; the matrix camera simultaneously captures the image information of the train.

[0115] (II)Data processing stage

[0116] The data processing module processes the collected data. It unifies the data collected by the area array camera and the 3D laser profiler under the same global coordinate system. Through an accurate calibration process, it determines the conversion relationships between the camera coordinate system, the laser profiler coordinate system, and the global coordinate system to ensure the consistency of the measurement data of different devices in spatial positions. In the area array camera images and the 3D laser profiler point cloud data, representative feature points are respectively extracted, such as the corner points and key contour points of the vehicle. A feature point matching algorithm is used to find the corresponding feature point pairs in the two types of data. When the similarity meets the preset feature similarity threshold (such as 85%), the corresponding key feature points of the point cloud and the key features of the image are used as position correlation features. According to the matched feature point pairs, a position mapping relationship is established between the area array camera images and the 3D laser profiler point cloud data. Through this mapping relationship, the position information measured by the two devices can be correlated with each other, facilitating subsequent anomaly analysis.

[0117] (III) Data comparison stage

[0118] 1. Consistency verification

[0119] Taking a certain laser profiler as the target sensor, a corner point of the train head is selected as the target verification point according to the position correlation features, and its target verification value in the world coordinate system is obtained. Another laser profiler with key position features related to this laser profiler is used as the reference sensor, and the reference verification value of the same corner point of the train head collected by it is obtained. The consistency score is calculated according to the consistency score calculation formula.

[0120] 2. Stability verification

[0121] Key dimensions such as the width, length, and surface curvature of the train carriages are extracted from the 3D point cloud data of the train collected by the laser profiler and compared with the historical dimensions. The stability score is calculated according to the stability score calculation formula.

[0122] 3. Redundant information verification

[0123] The 3D point cloud data is processed to obtain depth data and preprocessed; the train image information is processed to obtain texture images and preprocessed. The coordinates of the depth data are converted to the same coordinate system as the texture images, and the correlation between the depth data and the texture data is calculated to obtain the redundant information verification score.

[0124] (IV) Credibility evaluation stage

[0125] The credibility evaluation module calculates the credibility scores of each sensor according to a preset credibility scoring strategy. At the same time, according to the weight parameter adjustment sub-strategy, the weight parameters of each verification index are adjusted in combination with historical over-limit data. For example, if the number of train over-limit anomalies in the detection range corresponding to a certain laser profiler has been relatively large recently, the weight parameter of its stability score is appropriately reduced.

[0126] (5) Calibration warning stage

[0127] The calibration warning module makes a judgment based on the distribution of the credibility scores. If the deviation degree between the credibility score of a certain sensor and the average value of the credibility scores of other sensors is greater than a preset score deviation threshold (such as 20%), then this sensor is regarded as a sensor to be verified, a verification instruction is generated, and the staff is notified to check and calibrate this sensor.

[0128] A laser profiler can be used to obtain three-dimensional point cloud data, and an area array camera obtains the image information of the train. The three-dimensional point cloud data and the image information can be associated according to position features, such as contour features like edges and corners, and then the credibility score is calculated.

[0129] Through the above steps, the system of the present invention can perform real-time calibration and monitoring on the positions of the sensors, ensuring the accuracy and reliability of the clearance identification.

[0130] The above is only the preferred embodiment of the present invention, and the protection scope of the present invention is not limited to the above embodiments. All technical solutions falling within the idea of the present invention belong to the protection scope of the present invention. It should be noted that for those of ordinary skill in the art, several improvements and refinements made without departing from the principle of the present invention should also be regarded as within the protection scope of the present invention.

Claims

1. A sensor position calibration system for clearance recognition, characterized in that: Including: A data acquisition module that obtains position information data collected by each sensor; A data processing module that obtains the position information data and respectively identifies position - associated features, where the position - associated features reflect features that conform to a corresponding relationship in the position information data collected by different sensors; A data comparison module that performs consistency verification, stability verification, and redundant information verification based on the position - associated features, and sends the verification result data to a credibility evaluation module; A credibility evaluation module that calculates the credibility scores of each sensor according to a preset credibility scoring strategy, where the lower the credibility score, the higher the probability that the corresponding sensor is abnormal; A calibration warning module that makes a judgment based on the distribution of credibility scores. If the deviation degree between the credibility score of one sensor and the average value of the credibility scores of other sensors is greater than a preset scoring deviation threshold, then this sensor is regarded as a sensor to be verified and a verification instruction is generated; The consistency verification step includes: Defining the current sensor as a target sensor, and obtaining at least one target verification point based on the position - associated features; Obtaining the value of the target verification point in the world coordinate system to define it as a target verification value; Defining the sensor with a position - key feature related to the target sensor as a reference sensor, and obtaining a corresponding reference verification point based on the target verification point; Obtaining the value of the reference verification point in the world coordinate system to define it as a reference verification value; The stability verification step includes: Obtaining position information data and extracting key dimensions, where the key dimensions include width, length, and surface curvature; Obtaining historical dimensions and comparing them with the key dimensions; The redundant information verification step includes: Obtaining depth data based on three - dimensional point cloud data and performing data pre - processing on the depth data; Obtaining a texture image based on train image information and performing image pre - processing; Converting the coordinates of the depth data to the same coordinate system as the texture image so that they have a corresponding relationship; Calculating the correlation between the depth data and the texture data. The greater the correlation, the higher the degree of fit and the higher the redundant information verification score.

2. The sensor position calibration system for clearance recognition according to claim 1, characterized in that: The sensor includes several laser profilers and several area array cameras. The laser profilers obtain three - dimensional point cloud data of the train, and the area array cameras obtain train image information of the train. The data processing module identifies point cloud key features and point cloud feature parameters in the three - dimensional point cloud data. The data processing module identifies image key features and image feature parameters in the train image information, and judges the similarity between the point cloud feature parameters and the image feature parameters. If the similarity meets a preset feature similarity threshold, then the corresponding point cloud key feature and image key feature / several point cloud key features / several image key features are used as the position - associated features. The position - associated features include corner points and contour points of the vehicle.

3. The sensor position calibration system for clearance recognition according to claim 1, characterized in that: The specific calculation formula of the credibility scoring strategy is: ， Where S represents the credibility evaluation score, and A represents the consistency score. The weight parameter representing the consistency score, B represents the stability score. The weight parameter representing the stability score, C represents the redundant information verification score. The weight parameter representing the redundant information verification score.

4. The sensor position calibration system for clearance identification according to claim 1, characterized in that: The specific calculation formula of the consistency score is: ， ， ， D represents the total consistency deviation of the target sensor, and m represents the number of reference sensors. Among them, the feature associated with the j-th position has verification points in the i-th measurement. E is the k-th target verification value of the key feature at the j-th position of the target sensor in the i-th measurement, F is the reference verification value of the l-th reference sensor compared with the target sensor under the same measurement conditions, W represents the weight of the number of verifications, represents the total number of verifications of all reference sensor devices in the corresponding measurement.

5. The sensor position calibration system for clearance identification according to claim 1, characterized in that: The specific calculation formula of the stability score is: ， ， Where V represents the variance between the critical dimension and the historical dimension, and q - 1 represents the number of historical dimensions, represents the parameter value of the historical dimension, represents the parameter value of the corresponding critical dimension.

6. The sensor position calibration system for clearance recognition according to claim 3, characterized in that: The credibility evaluation module is configured with a weight dynamic adjustment strategy, and the formula of the weight dynamic adjustment strategy is specifically: ， ， Among them In it, v takes values of 1, 2, and 3, representing the weight parameters of the consistency score, the weight parameters of the stability score, and the weight parameters of the redundant information verification score respectively represents the environmental parameter vector represents the basic weight represents the influence coefficient of the environmental parameter on the weight represents the quantization function of the influence of the environmental parameter on the weight 7. The sensor position calibration system for clearance identification according to claim 3, characterized in that: The credibility evaluation module is configured with a stability weight parameter adjustment sub-strategy, including obtaining historical over-limit data. The historical over-limit data includes the train over-limit position. It judges the train over-limit position and the detection range of the sensor. If the number of train over-limit anomalies that occur within the detection range corresponding to the target sensor is larger, the basic weight of the stability score weight parameter is smaller. If the number of times the target sensor is verified is larger, the weight parameter of the consistency score is larger. If the number of sensors with position correlation characteristics with the target sensor is smaller, the basic weight of the redundancy information verification score weight parameter is larger.

Images