Method for automatically measuring geometric parameters of overhead line system based on laser radar

Abstract

The invention discloses a method for automatically measuring geometric parameters of a contact network based on a laser radar, which comprises the following steps of: scanning and measuring the contact network by using a measuring trolley running on a track through matching a vehicle-mounted single-line laser radar with a precision odometer, a tilt angle sensor and an inertial measurement unit; based on laser point cloud data, through the steps of preprocessing, contact line recognition and extraction, three-dimensional track reconstruction and the like, accurate calculation of geometric parameters such as the guide height and the pull-out value of the contact network is achieved. The measuring precision can reach the millimeter level, and the method has the advantages of being high in automation degree, light and flexible in equipment, high in cost effectiveness and the like, can be widely applied to various rail transit systems such as high-speed railways and urban rail transit which are powered by the contact network, provides scientific data support for contact network state monitoring and maintenance decision making, and has wide application prospects. And the operation safety and the operation and maintenance economy of the rail transit system are improved.

Description

TECHNICAL FIELD

[0001] The present application belongs to the field of rail transit technology, and relates to a catenary geometric parameter automatic measurement method based on a laser radar. The present application is applicable to rail transit systems using catenary power supply, such as high-speed railways, intercity railways, general speed railways, urban rail transit, subways, and tramcars, and performs high-precision, real-time, and automatic measurement and monitoring on key geometric parameters of catenary, such as the catenary height and the catenary pull-out value, to provide reliable data support for the safe operation, state evaluation, and maintenance decision of catenary equipment. BACKGROUND

[0002] The catenary is a key component of the electrified rail transit power supply system, and the accurate measurement of its geometric parameters is of great significance to ensuring the safety of train operation and the reliability of power supply. The catenary geometric parameters mainly include the catenary height, the catenary pull-out value, and the positioner slope, which directly affect the dynamic interaction performance of the pantograph and the catenary.

[0003] Currently, there are mainly the following methods for measuring the geometric parameters of the catenary of rail transit:

[0004] 1. Manual measurement method: The geometric parameters of the catenary are measured by manually carrying measurement tools. This method is simple to operate, but has low measurement efficiency, the precision is greatly affected by human factors, and there are safety hazards.

[0005] 2. Catenary detection vehicle measurement method: A dedicated catenary detection vehicle is used to carry out dynamic measurement with measurement equipment. This method has high measurement efficiency, but the detection vehicle is expensive, has high maintenance cost, occupies line resources, and affects normal operation.

[0006] 3. Image measurement method: The catenary image is collected by a camera device installed on the roof of the vehicle, and the geometric parameters are extracted by image processing technology. This method is safe without direct contact with the catenary, but the image quality is easily affected in complex environments such as tunnels and at night, and the measurement precision is unstable.

[0007] 4. Laser measurement method: The geometric parameters of the catenary are measured by using the principle of laser ranging. This method has high measurement precision, but the equipment cost is high, and the measurement performance will be affected in bad weather conditions.

[0008] The main problems existing in the prior art include: insufficient measurement precision, which is difficult to meet the requirements of high-speed railways for accurate measurement of catenary geometric parameters; low automation level, which requires a large amount of manual participation; large size and heavy weight of the measurement equipment, which is not convenient for flexible deployment; limited data processing capacity, which makes it difficult to realize real-time analysis and early warning; and high measurement cost, which makes it difficult to be widely applied.

[0009] With the rapid development of high-speed railway and urban rail transit, higher requirements are put forward for the precision, efficiency and automation degree of the measurement of the geometric parameters of the catenary. Therefore, developing a catenary geometric parameter measurement method with high precision, high efficiency and high automation degree has important theoretical significance and practical value. SUMMARY

[0010] In order to overcome the shortcomings of the prior art, the present application provides a catenary geometric parameter automatic measurement method based on laser radar, which can measure the geometric parameters such as the catenary height and the pull-out value with high precision and high efficiency, improve the automation degree of measurement, reduce the labor cost, and ensure the safe operation of rail transit.

[0011] In order to achieve the above-mentioned purpose, the present application adopts the following technical solutions:

[0012] A catenary geometric parameter automatic measurement method based on laser radar, comprising the following steps:

[0013] Step 1: configuring a measurement trolley to perform system calibration and calibration on a standard track gauge horizontal calibration platform, the measurement trolley comprising one or more of a laser radar, an odometer, an inclination sensor, an inertial measurement unit, a data acquisition and processing unit, and a wireless data transmission module, the system calibration and calibration comprising zero calibration, size parameter calibration, odometer calibration, system verification, and calibration data storage;

[0014] Step 2: collecting original data with time stamp during the running of the measurement trolley on the track, the original data comprising one or more of polar coordinate parameters, mileage data, inclination data, and inertial measurement data;

[0015] Step 3: preprocessing the original data, at least including noise point removal, point cloud registration, data synchronization, and attitude correction, to obtain processed point cloud data;

[0016] Step 4: based on the point cloud data, using a multi-step feature extraction and pattern recognition method to identify and reconstruct the catenary line space structure in the point cloud data;

[0017] Step 5: based on the catenary line space structure, calculating the catenary line height parameter and the pull-out value parameter;

[0018] Step 6: parameter verification and correction output.

[0019] In summary, the present application has the following advantages:

[0020] 1. High precision measurement: High precision single line laser radar is used in combination with precise odometer, inclination sensor and inertial measurement unit, and through multi-sensor data fusion technology, the measurement precision can reach millimeter level, the height measurement precision is better than ±5mm, the pull-out value measurement precision is better than ±5mm, which meets the strict requirements of high-speed railway and other contact network geometry parameter measurement.

[0021] 2. High degree of automation: The whole measurement process is automatically completed, from data acquisition, processing to result output without manual intervention, which greatly reduces the labor cost and work intensity, improves the operation safety, and avoids the high-altitude operation risk in traditional manual measurement.

[0022] 3. Portable and flexible equipment: The measurement trolley is small in size and light in weight, easy to carry and deploy, can be quickly put into use, and has little impact on normal operation of rail transit, without occupying a large amount of line resources.

[0023] 4. All-weather working ability: Laser radar technology is used, which has low dependence on environmental light conditions, can work stably in daytime and nighttime, and the measurement result is less affected by environmental factors, which expands the application range and working time window of the equipment.

[0024] 5. Real-time data processing capability: Through the vehicle-mounted data processing unit and wireless transmission technology, real-time processing and remote transmission of measurement data are realized, which supports timely discovery and processing of contact network geometry parameter abnormal conditions, and improves the operation efficiency and fault response speed.

[0025] 6. Strong systematic error correction capability: Through the inclination sensor, inertial measurement unit and advanced calibration algorithm, the influence of attitude change of the measurement trolley during driving on the measurement precision is effectively eliminated, which improves the reliability of the measurement result, especially the measurement precision in the track irregularity section.

[0026] 7. High cost-effectiveness: Compared with the traditional large detection vehicle, the measurement method and equipment of the present application have low cost, simple maintenance, and are economical and practical, which makes small and medium-sized railway operation units also affordable for high-precision contact network measurement equipment.

[0027] 8. Data value improvement: The measurement result can be directly associated with the line mileage, which is convenient for establishing a space-time distribution database of contact network geometry parameters, providing scientific basis for contact network state evaluation, trend analysis and maintenance decision, and supporting intelligent maintenance and predictive maintenance of railway contact network system.

[0028] 9. Wide applicability: The method is not only suitable for high-speed railway contact network measurement, but also can be applied to various types of rail transit systems using contact network power supply, such as ordinary speed railway, urban rail transit, subway and tram, which can adapt to the measurement needs of different types of contact network through simple parameter adjustment. BRIEF DESCRIPTION OF DRAWINGS

[0029] Figure 1 The structural block diagram of the laser radar-based overhead line geometry parameter automatic measurement method of the application.

[0030] Figure 2 The overall structure schematic diagram of the measurement trolley.

[0031] Figure 3 The working scene schematic diagram of the measurement trolley. DETAILED DESCRIPTION

[0032] The embodiments of the application are illustrated by specific examples below, and other advantages and effects of the application can be easily understood by those skilled in the art from the disclosure. The application can also be implemented or applied by other different embodiments, and various modifications or changes can be made to the details in the specification based on different views and applications without departing from the spirit of the application. It should be noted that the following examples and features in the examples can be combined with each other without conflict.

[0033] The application provides a laser radar-based overhead line geometry parameter automatic measurement method, comprising the following steps:

[0034] Step 1: device construction

[0035] A measurement trolley capable of freely traveling on a rail is provided, the bottom of the trolley is provided with a wheel set matched with the rail, and a high-precision single-line laser radar scanner scanning upward is installed on the top of the trolley; a precise odometer, a three-axis tilt sensor, an inertial measurement unit, a data acquisition and processing unit, and a wireless data transmission module are further arranged on the trolley; the odometer is used to record the travel mileage and position information of the measurement trolley; the tilt sensor is used to measure the longitudinal tilt angle, lateral tilt angle and rotation angle of the trolley; the inertial measurement unit is used to measure the acceleration and angular velocity of the trolley; the data acquisition and processing unit is used to acquire and process the data of the laser radar, the odometer, the tilt sensor and the inertial measurement unit; the wireless data transmission module is used to transmit the processed data to a remote server or a mobile terminal; wherein the laser radar adopts a single-line vertical scanning mode, the scanning frequency is adjustable, the scanning angle range covers the position of the overhead line, and the distance measurement accuracy is not less than ±5mm.

[0036] The measurement trolley comprises a chassis module, an electronic equipment cabin and a three-axis fine adjustment mounting platform.

[0037] The chassis module mainly comprises a wheel set system, a travel driving system and a posture sensing system.

[0038] The wheel set system is installed at the bottom of the car body of the measuring trolley, adopts a "4-point contact" layout, and includes four wheels, two of which are driving wheels connected with the driving system of the measuring trolley, and the other two are driven wheels for balancing and supporting. The wheels are made of special polymer composite material, the surface hardness reaches Rockwell hardness HRC 55-60, and the wear resistance is good, so that the wheels can adapt to various rail surface conditions. The wheel rim angle is 75°±1°, which meets the requirements of the national standard GB2585-2007 "Railway Vehicle Wheel Tread and Rim Size and Wheelset Inner Distance", so as to ensure the stability and safety of the trolley running on the track. The wheel set is equipped with an elastic suspension system, which can adapt to a track gauge error of ±3mm, and ensure stable contact between the four wheels and the track.

[0039] The driving system adopts a direct-current brushless motor, and the motor power is transmitted to the driving wheel through a planetary gear reducer. An electromagnetic power-off brake is provided, which can automatically lock the wheel in the case of power failure or emergency.

[0040] The inertial measurement unit and the tilt sensor of the attitude sensing system are installed on the special fixed seat at the center of the car body, and are strictly aligned with the car body coordinate system. The odometer adopts a high-precision incremental optical encoder, which is directly connected to the driven wheel shaft. The driven wheel connected with the encoder is made of high-friction coefficient material, and appropriate wheel pressure is applied, so as to minimize the slipping phenomenon and improve the accuracy of the mileage measurement.

[0041] The electronic equipment cabin is used for installing the main controller, the data acquisition unit, the wireless data transmission module and the power supply system.

[0042] The main controller adopts a high-performance embedded computing platform, uses a customized embedded Linux system, and is optimized for real-time data processing. The maximum delay of the key data processing process is not more than 10ms, which meets the real-time control and data processing requirements.

[0043] The data acquisition unit adopts a 32-bit microcontroller, which is responsible for real-time acquisition and preprocessing of sensor data such as encoders, tilt sensors and inertial measurement units. A high-precision time synchronization mechanism is adopted, and all collected data is attached with a 1ms resolution timestamp, and the synchronization accuracy is better than 10ms.

[0044] The wireless data transmission module supports 4G / 5G and Wi-Fi communication, which is used for remote data transmission and remote control instructions.

[0045] The power management system is equipped with a high-performance lithium ion battery pack and a built-in BMS battery management system. In the standard working mode, the trolley can work continuously for about 5 hours.

[0046] The three-axis fine adjustment mounting platform is arranged on the top of the vehicle body of the measuring trolley and is used for fixing and fine adjusting the laser radar. The platform is provided with a three-axis fine adjustment mechanism, the adjustable range is ±10 mm, and the adjustment accuracy is 0.02 mm. A damping device is arranged between the platform and the vehicle body, is made of damping silica gel material, can effectively absorb the vibration generated by the vehicle driving, and ensures the stable work of the sensor. A precisely processed calibration reference point is arranged on the platform and is used for accurately positioning and calibrating the installation position of the sensor.

[0047] The measurement system hardware composition design provided by the embodiment has the characteristics of compact structure, high integration, high precision and simple operation, and can meet the technical requirements of automatic measurement of the geometric parameters of the overhead line system of rail transit, and provides a reliable hardware foundation for subsequent data acquisition and parameter calculation.

[0048] Step 2: System calibration

[0049] Before the measurement of the geometric parameters of the overhead line system, the measuring trolley is placed on the horizontal calibration platform of the standard track gauge to perform system inspection and system calibration, so as to ensure the measurement accuracy and reliability.

[0050] The system inspection process mainly includes hardware self-checking and software initialization. The hardware self-checking refers to automatically executing a hardware self-checking program after the system is started, and the checking items include battery power state checking (the remaining power is required to be greater than or equal to 30%), sensor connection state checking (all sensors communicate normally), storage space checking (the available storage space is greater than or equal to 10 GB), and drive system checking (the motor and wheel set are running normally). The software initialization includes loading system configuration parameters, initializing the data acquisition module, establishing the sensor data stream and activating the real-time monitoring system.

[0051] The system calibration process includes zero calibration, size parameter calibration, odometer calibration, system verification and calibration data storage.

[0052] (1) Zero calibration. The zero calibration is a calibration work of placing the measuring trolley on the special calibration platform, including laser radar zero calibration and inclination sensor zero calibration. Among them, the laser radar zero calibration is to use a precise level to ensure that the verticality error of the scanning plane of the laser radar and the vertical plane of the track is not more than 0.1°, and if it exceeds the range, the adjustment is performed through the fine adjustment mechanism until the requirements are met; the inclination sensor zero calibration is to collect 10 times of inclination sensor reading data of the trolley on the standard horizontal plane, take the average value to eliminate random errors, and obtain the reference value (α0, β0, γ0) of the subsequent measurement inclination sensor, and the calculation formula is:

[0053] α0=(1 / n)·∑α i

[0054] β0=(1 / n)·∑β i

[0055] Y0 = (1 / n) * ∑Yi i

[0056] where n = 10, i = 1, 2, 3,..., 10

[0057] (2) Dimensional parameter calibration. The dimensional parameter calibration includes measuring and recording the installation height h0 of the laser radar from the rail surface, measuring and recording the lateral offset d0 of the laser radar relative to the center of the track, and recording the wheel diameter value D. The measurement of the laser radar installation height is to measure the vertical distance from the laser radar emission center to the rail surface using a high-precision digital caliper, and the average value is taken for three measurements with an accuracy of ±0.1 mm; the measurement of the lateral offset of the laser radar is to measure the horizontal distance from the laser radar center to the center line of the track using a standard reference ruler, and the average value is taken for three measurements with an accuracy of ±0.1 mm; the measurement of the wheel diameter value is to measure the diameter of the encoder connected to the wheel using a precision micrometer, and the average value is taken for six point measurements with an accuracy of ±0.01 mm.

[0058] (3) Odometer calibration. The odometer is calibrated on a calibration track with a standard length of 100±0.001m, the encoder count value N is recorded by measuring the trolley passing through the calibration track at a constant speed, and then the odometer correction coefficient K1 is calculated.

[0059] K1 = L / N * π * D

[0060] where L is the actual length of the calibration section (m), D is the wheel diameter (m), and N is the encoder count value.

[0061] (4) System verification. The geometric parameters of the known parameter catenary are measured on a standard test section, the measurement results are compared with the standard values, the error distribution is calculated, and then based on the error analysis, the height calibration coefficient k_H, the height zero correction value b_H, the pull-out value calibration coefficient k_D, and the pull-out value zero correction value b_D are obtained, and then the calibration coefficient matrix is established with these five parameters.

[0062] k_H = S_std / S_meas

[0063] b_H = H_std - k_H * H_meas

[0064] k_D = D_std / D_meas

[0065] b_D = D_std - k_D * D_meas

[0066] where the suffix std is the standard value, and the suffix meas is the measured value.

[0067] (5) Calibration data storage. Store all calibration values of zero position parameters (a0, b0, g0), size parameters (h0, d0), mileage correction coefficient K1 and calibration coefficient matrix (k_H, b_H, k_D, b_D) into system configuration file, and record information such as calibration time, environmental temperature, operator, etc. for traceability.

[0068] Step 3: Data acquisition

[0069] The measurement trolley is driven to travel along the track, and the scanning frequency of the laser radar is set to f Hz, which can be adaptively adjusted according to the speed of the trolley, generally 25 Hz-100 Hz. During the travel of the trolley, the laser radar continuously scans upward, and each scan obtains a point cloud data set in a vertical plane, recording the polar coordinate parameters of each scan point: distance value r and angle value θ. At the same time, the odometer records the cumulative distance s traveled by the trolley and the corresponding timestamp t. The tilt sensor records the three-axis tilt data of the trolley during travel: longitudinal tilt angle a, lateral tilt angle b, and rotation angle g. The inertial measurement unit records the three-axis acceleration and three-axis angular velocity data of the trolley. All data are provided with accurate timestamps for subsequent time synchronization and data fusion.

[0070] (1) Initial positioning. Place the measurement trolley at the starting point of the track to be measured, confirm the starting position coordinates and mileage by GPS or mileage post, and then input the starting mileage value and line number in the system;

[0071] (2) Measurement start. After selecting the appropriate inspection speed on the operation interface, start the data acquisition system and control the system to drive the measurement trolley to start traveling uniformly along the track;

[0072] (3) Real-time data acquisition. Set the scanning frequency of the laser radar to f Hz, which can be adaptively adjusted according to the speed of the trolley, preferably 25 Hz-100 Hz. The laser radar continuously scans upward, and each scan obtains a point cloud data in a plane, recording the polar coordinate parameters of each scan point: distance value r and angle value θ. At the same time, the odometer records the cumulative distance s traveled by the trolley and the corresponding timestamp t. The tilt sensor records the three-axis tilt data of the trolley during travel: longitudinal tilt angle a, lateral tilt angle b, and rotation angle g. The inertial measurement unit records the three-axis acceleration and three-axis angular velocity data of the trolley. All data are provided with accurate timestamps for subsequent time synchronization and data fusion.

[0073] The original data format collected by the laser radar is:

[0074] {t_i, (r_ij, θ_ij) | j = 1, 2,..., m}

[0075] where t_i is the timestamp of the i-th scan, r_ij and θ_ij are the distance value and angle value of the j-th point in the scan, and m is the number of points in a single scan.

[0076] The data format recorded by the odometer is:

[0077] {t_k, s_k | k = 1, 2,...}

[0078] where t_k is the timestamp and s_k is the cumulative distance traveled at the corresponding time.

[0079] The data format recorded by the tilt sensor is:

[0080] {t_l, α_l, β_l, γ_l | l = 1, 2,...}

[0081] where t_l is the timestamp, and α_l, β_l, γ_l are the longitudinal tilt angle, lateral tilt angle, and rotation angle at the corresponding time.

[0082] The data format recorded by the inertial measurement system is:

[0083] {t_p, (a_x, a_y, a_z, ω_x, ω_y, ω_z)_p | p = 1, 2,...}

[0084] where t_p is the timestamp, a_x, a_y, a_z are the three-axis acceleration values, and ω_x, ω_y, ω_z are the three-axis angular velocity values.

[0085] Step 4: Data preprocessing

[0086] The collected raw data is preprocessed, including:

[0087] 4.1. Noise point removal: Based on distance threshold method, neighborhood density analysis and statistical outlier detection algorithm, the noise points and outliers in the point cloud are removed;

[0088] Specifically, the noise point removal algorithm includes three methods of distance threshold filtering, statistical outlier detection and neighborhood density analysis. The distance threshold filtering is to set the minimum effective distance r_min=0.5 m and the maximum effective distance r_max=10 m, and exclude the points whose distance values are not in the range of [r_min, r_max]; the statistical outlier detection is to apply the improved z-score algorithm to the point set in each scanning frame, calculate the mean value μ and the standard deviation σ of the distance values, and if the distance value r of a point satisfies |r-μ|>λ·σ, mark it as a noise point, wherein λ is a threshold coefficient, and λ=2.5 in the embodiment; the neighborhood density analysis is to calculate the k-neighborhood point set N_k(P) of each point P, and k=5 in the embodiment, calculate the average distance d_avg of P to each point in N_k(P), and if d_avg exceeds the preset threshold d_th, d_th=0.1 m in the embodiment, mark P as a noise point.

[0089] 4.2, Point cloud registration: based on the accurate timestamp, the point cloud data acquired at different times is spatially registered, based on the mileage data, the point cloud data acquired at different times is converted to the same coordinate system, the car driving trajectory curve T(t) is established, the point cloud in each scanning frame is transformed to the global coordinate system based on the car pose at the acquisition time, and the three-dimensional point cloud data along the track direction is generated;

[0090] 4.3, Data synchronization: the linear interpolation method is used to synchronize the lidar point cloud data, the mileage data, the inclination data and the inertial measurement data based on the timestamp, and a unified data structure is established;

[0091] 4.4, Pose correction: the Kalman filter algorithm is used to fuse the inclination sensor and inertial measurement unit data, estimate the accurate pose of the car during driving, and compensate the pose of the point cloud data, so as to eliminate the influence of the car motion on the measurement results.

[0092] Specifically, the pose correction algorithm includes two parts of pose data fusion and point cloud pose compensation. The pose data fusion is to fuse the inclination sensor and inertial measurement system data by using the extended Kalman filter (EKF) algorithm, the state vector is defined as X=[α,β,γ,ω_x,ω_y,ω_z]^T, and the observation vector is defined as Z=[α_meas,β_meas,γ_meas,ω_x_meas,ω_y_meas,ω_z_meas]^T; the point cloud pose compensation is to construct a pose compensation matrix according to the fused pose data, and apply the compensation matrix to correct the point cloud data.

[0093] Step 5: Contact line identification and extraction

[0094] Based on the processed point cloud data, a multi-step feature extraction and pattern recognition method is used to identify the contact line structure in the point cloud:

[0095] 5.1, Spatial clustering: improved density-based spatial clustering of applications with noise (DBSCAN) is adopted to classify the point cloud, including:

[0096] The neighborhood radius, i.e., the proximity distance threshold between points, is defined as ε = 0.05 m, the minimum number of points required to form a cluster is defined as MinPts = 10, and the maximum gap between adjacent points in the same cluster is defined as MaxGap = 0.2 m. The steps of the clustering algorithm include initializing the state of all points as "unvisited", randomly selecting an unvisited point p for processing, finding the set of all points N(p) in the ε neighborhood of p, if |N(p)| < MinPts, marking p as "noise"; otherwise, creating a new cluster C and adding p to C, then for each point p' in N(p): if p' state is "unvisited", mark it as "visited" and find the ε neighborhood N(p') of p'; if |N(p')| ≥ MinPts, add the points in N(p') to N(p) for further processing, when all points in N(p) have been processed, the cluster C is formed, and the above steps are repeated until all points are visited. The improvements include introducing spatial continuity constraints, using kd-tree to accelerate neighborhood search, and increasing the density adaptive mechanism.

[0097] 5.2, Contact line feature extraction and recognition, including height range screening, geometric feature analysis, and cross-sectional feature analysis: calculate the spatial distribution characteristics of each cluster, including height distribution, linearity, cross-sectional feature, etc., and screen the candidate point set that meets the contact line feature;

[0098] Height range screening is based on prior knowledge that the overhead contact line height is generally in the range of 4.0 m to 7.0 m. The height distribution of each cluster C is calculated to obtain the height median h_median(C). If h_median(C) is not in the range of [4.0 m, 7.0 m], the cluster is excluded.

[0099] Geometric feature analysis is to calculate the principal component analysis (PCA) eigenvalue of each cluster C, define λ1≥λ2≥λ3, and then calculate the linearity index L(C) = (λ1-λ2) / λ1. If L(C) < 0.8, the cluster is excluded because the contact line should have high linearity. If the cluster C passes the linearity test, record its principal direction vector v1, which corresponds to the eigenvalue λ1.

[0100] Cross-sectional feature analysis is to construct a plane P perpendicular to the principal direction v1 for each candidate cluster C, project the points in C onto P to obtain the projected point set C_proj, apply the circular fitting algorithm to C_proj to obtain the fitting circle radius r and fitting error e. If r is in the range of the contact line radius [4 mm, 8 mm] and e < 2 mm, then C is determined as the point set of the contact line.

[0101] 5.3, Model fitting: Apply cylinder model fitting algorithm to the candidate point set, extract the accurate spatial position and diameter information of the contact line;

[0102] Adopt cross-section circle center positioning, along the main direction of the contact line, every Δs, Δs = 0.1m in this embodiment, select a cross-section, at each cross-section, apply least squares circle fitting algorithm to determine the center position of the contact line cross-section circle, the circle fitting algorithm adopts the method of combining algebraic fitting and geometric optimization to improve the positioning accuracy.

[0103] 5.4, Trajectory reconstruction: Along the track direction, connect the discrete contact line position points through the cubic spline interpolation algorithm to reconstruct the continuous spatial trajectory of the contact line.

[0104] Specifically, arrange all cross-section center points in order of mileage to form a discrete contact line trajectory point set {(s_i, x_i, y_i, z_i)}, apply the cubic spline interpolation algorithm to fit x(s), y(s), and z(s) respectively to obtain a continuous contact line spatial curve, the cubic spline interpolation ensures the C 2 continuity of the reconstructed trajectory, which conforms to the physical characteristics of the contact line.

[0105] It also includes trajectory smoothing, applying the Savitzky-Golay filter to the reconstructed contact line trajectory to remove high-frequency noise, the filter parameters are window size w = 21 and polynomial order p = 3, which ensures that the smoothing process does not change the overall shape characteristics of the trajectory.

[0106] Step 6: Accurate measurement of catenary geometric parameters

[0107] Step 6.1: Guide height calculation method

[0108] Guide height is defined as the vertical distance from the contact line to the track surface, its calculation adopts the following steps:

[0109] 6.1.1, For each position point s of the track, extract the corresponding laser radar measurement original data: distance value r and scanning angle θ from the reconstructed contact line trajectory;

[0110] 6.1.2, Convert polar coordinate data to three-dimensional coordinates in Cartesian coordinate system:

[0111] x raw = 0

[0112] y raw = r cos θ

[0113] z raw = r sin θ

[0114] wherein the laser radar scanning plane is perpendicular to the longitudinal axis of the track, scanning in the YZ plane, x raw is 0; θ is the elevation angle from the transverse horizontal (Y-axis positive direction) upward;

[0115] 6.1.3, apply the posture compensation matrix to convert the measurement coordinates to the standard reference coordinate system, eliminating the influence of the posture change of the trolley:

[0116]

[0117] wherein R is the posture rotation matrix constructed by the three-axis inclination (α, β, γ) measured by the inclination sensor and the calibration reference angle (α0, β0, γ0):

[0118] R = R_z (γ - γ0) · R_y (β - β0) · R_x (α - α0)

[0119] 6.1.4, apply the head calculation formula:

[0120] H = z corrected -h0

[0121] wherein z corrected is the contact line height coordinate after posture compensation, h0 is the calibration height offset of the laser radar installation position relative to the rail surface;

[0122] 6.1.5, apply the calibration coefficient to compensate for system error:

[0123] H_calibrated = k_H · H + b_H

[0124] wherein k_H is the calibration proportion coefficient of the head, b_H is the zero point correction value of the head, both of which are obtained by standard test section calibration;

[0125] 6.1.6, apply temperature compensation to correct environmental impact:

[0126] H_final = H_calibrated + α_T · (T - T0)

[0127] wherein α_T is the temperature compensation coefficient (typical value is 0.05 mm / ℃), T is the current environmental temperature, and T0 is the reference temperature (20℃) at the time of calibration.

[0128] Step 6.2: Pull-out value calculation method

[0129] The pull-out value is defined as the horizontal offset distance of the contact line relative to the track center line, and its calculation uses the following steps:

[0130] 6.2.1, for each position point s of the track, extract the corresponding laser radar measurement raw data: distance value r and scanning angle θ from the reconstructed contact line trajectory;

[0131] 6.2.2, Convert polar coordinate data to three-dimensional coordinates in Cartesian coordinate system:

[0132] x raw = 0

[0133] y raw = r cos θ

[0134] z raw = r sin θ

[0135] Wherein, the laser radar scanning plane is perpendicular to the longitudinal axis of the track, scanning in the YZ plane, x raw is 0; θ is the elevation angle from the horizontal (Y axis positive direction) upward; y raw is the lateral offset component of the contact line;

[0136] 6.2.3, Apply pose compensation matrix to convert measurement coordinates to standard reference coordinate system, eliminate the influence of trolley pose change:

[0137]

[0138] Wherein, R is the pose rotation matrix constructed by the three-axis inclination angle (α, β, γ) measured by the inclination sensor and the calibration reference angle (α0, β0, γ0):

[0139] R = R_z (γ - γ0) · R_y (β - β0) · R_x (α - α0)

[0140] 6.2.4, Apply pull-out value calculation formula:

[0141] D = y corrected - d0

[0142] Wherein, y corrected is the lateral coordinate of the contact line after pose compensation, d0 is the calibration lateral offset of the laser radar installation position relative to the track center line;

[0143] 6.2.5, Apply calibration coefficient to compensate system error:

[0144] D_calibrated = k_D · D + b_D

[0145] Wherein, k_D is the pull-out value calibration proportion coefficient, b_D is the pull-out value zero point correction value, both of which are obtained by standard test section calibration;

[0146] 6.2.6, For curve section, apply design value correction:

[0147] D_final = D_calibrated - D_design (s)

[0148] where D design(s) is the design pull-out value at this milepost, provided by the line parameter database. For straight segments, D design(s) = 0.

[0149] Step 7: Parameter verification and correction

[0150] 7.1, Multi-sensor data fusion: Through Kalman filtering algorithm, fuse the data of laser radar, inclination sensor and inertial measurement unit, improve the measurement stability and accuracy;

[0151] 7.2, Error compensation: Establish a comprehensive error model including proportional coefficient error, zero drift error and nonlinear error, and compensate for the original measurement results;

[0152] 7.3, Data smoothing: Adopt adaptive weighted sliding window filtering algorithm to smooth the calculation results and suppress the influence of random errors;

[0153] 7.4, Outlier detection: Apply outlier detection algorithm based on statistical analysis to identify and mark possible measurement outliers.

[0154] Step 8: Result output and evaluation

[0155] 8.1, Compare the measurement results with the design standard value and the allowable deviation value, and automatically identify the out-of-limit points;

[0156] 8.2, Generate detailed measurement report containing measurement position, lead value, pull-out value and evaluation results;

[0157] 8.3, Present the distribution of catenary geometric parameters through visual interface, including plan view, longitudinal section view and three-dimensional model, etc.

[0158] 8.4, According to the evaluation results, mark the abnormal sections and give maintenance suggestions;

[0159] 8.5, Store the measurement data into the catenary geometric parameter database, support historical comparison and trend prediction.

[0160] Obviously, the described embodiments are only part of the embodiments of the present application, not all. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor should belong to the scope of protection of the present application.

Claims

1. An automatic measurement method for the geometric parameters of a contact wire based on lidar, characterized in that, Includes the following steps: Step 1: Configure the measuring trolley to perform system calibration and verification on the standard gauge horizontal calibration platform. The measuring trolley includes at least one or more of the following: lidar, odometer, tilt sensor, inertial measurement unit, data acquisition and processing unit, and wireless data transmission module. System calibration and verification includes zero-point calibration, dimensional parameter calibration, odometer calibration, system verification, and calibration data storage. Step 2: Collect raw data with timestamps during the movement of the measuring trolley on the track. The raw data includes one or more of the following: polar coordinate parameters, mileage data, tilt angle data, and inertial measurement data. Step 3: Preprocess the raw data, including at least noise point removal, point cloud registration, data synchronization, and attitude correction, to obtain processed point cloud data; Step 4: Based on point cloud data, a multi-step feature extraction and pattern recognition method is used to identify and reconstruct the spatial structure of the contact line in the point cloud data; Step 5: Based on the spatial structure of the contact wire, calculate the contact wire guide height parameters and pull-out value parameters; Step 6: Output after parameter verification and correction.

2. The automatic measurement method for contact wire geometric parameters based on lidar according to claim 1, characterized in that, In step 1, zero-position calibration includes zero-position calibration of the lidar and zero-position calibration of the tilt sensor; dimensional parameter calibration includes calibration of the lidar installation height above the rail surface, calibration of the lidar lateral offset relative to the track center, and calibration of the wheel diameter of the measuring trolley; odometer calibration includes calibration of the mileage correction coefficient when the measuring trolley passes through the calibration section of the track at a constant speed; and system verification includes obtaining the parameters of the guide height calibration coefficient, guide height zero-point correction value, pull-out value calibration coefficient, and pull-out value zero-point correction value after calculating the error distribution of the measured values ​​of the geometric parameters of the known contact network, and establishing a calibration coefficient matrix.

3. The automatic measurement method for contact wire geometric parameters based on lidar according to claim 1, characterized in that, In step 2, the polar coordinate parameters include the distance and angle values ​​of the lidar acquisition points, the mileage data includes the cumulative distance traveled by the vehicle and the corresponding timestamp, the tilt angle data includes the longitudinal tilt angle, lateral tilt angle and rotation angle of the vehicle, and the inertial measurement data includes the triaxial acceleration and triaxial angular velocity of the vehicle.

4. The automatic measurement method for contact wire geometric parameters based on lidar according to claim 1, characterized in that, In step 3, noise points and outliers in the point cloud are removed based on the distance threshold method, neighborhood density analysis, and statistical outlier detection algorithm.

5. The automatic measurement method for contact wire geometric parameters based on lidar according to claim 4, characterized in that, In step 3, point cloud registration includes: spatially registering point cloud data acquired at different times based on precise timestamps; transforming point clouds acquired at different times to the same coordinate system based on mileage data to establish the vehicle's driving trajectory curve; and transforming the point cloud in each scan frame to the global coordinate system based on the vehicle's pose at the acquisition time to generate three-dimensional point cloud data along the track direction.

6. The automatic measurement method for contact wire geometric parameters based on lidar according to claim 5, characterized in that, In step 3, data synchronization includes: using linear interpolation to synchronize polar coordinate parameters, mileage data, tilt data, and inertial measurement data based on timestamps to establish a unified data structure.

7. The automatic measurement method for contact wire geometric parameters based on lidar according to claim 6, characterized in that, In step 3, attitude correction includes: using a Kalman filter algorithm to fuse data from the tilt sensor and the inertial measurement unit to estimate the precise attitude of the measurement vehicle during its movement, and performing attitude compensation on the point cloud data to eliminate the influence of the measurement vehicle's motion on the measurement results.

8. The automatic measurement method for contact wire geometric parameters based on lidar according to claim 1, characterized in that, In step 4, the identification and reconstruction of the contact line spatial structure includes: using a density-based spatial clustering algorithm to perform preliminary classification of point cloud data; extracting and identifying contact line feature point sets from height range, geometric features, and cross-sectional features; applying a cylinder model fitting algorithm to the point set to extract the precise spatial position and diameter information of the contact line, forming a discrete contact line trajectory point set; and connecting the trajectory position points using a cubic spline interpolation algorithm to reconstruct the continuous spatial trajectory of the contact line.

9. The automatic measurement method for contact wire geometric parameters based on lidar according to claim 1, characterized in that, In step 1, the measuring trolley includes a chassis module, an electronic equipment compartment, and a three-axis fine-tuning mounting platform. The chassis module includes a wheel system located at the bottom of the trolley's body, a driving system for driving the wheel system, and an attitude sensing system. The attitude sensing system includes an inclination sensor and an inertial measurement unit installed in the middle of the trolley's body, an odometer installed in the wheel system, a data acquisition and processing unit, and a wireless data transmission module located in the electronic equipment compartment. The electronic equipment compartment also includes a main controller and a power system. The three-axis fine-tuning mounting platform is located on the top of the trolley's body and is used to install a lidar.

Images