Guardrail expansion joint detection device and detection method thereof

Abstract

The invention relates to the technical field of road infrastructure detection, and discloses a guardrail expansion joint detection device and a detection method thereof, and the method comprises the steps: calculating a spatial sampling demand coefficient based on the running speed and acceleration of a mobile vehicle and the offset of a laser scanning instrument from a target; calculating a turning compensation coefficient based on the path curvature and the yaw velocity; calculating a scanning state matching degree according to the point cloud density and the signal-to-noise ratio under the spatial sampling demand coefficient and the turning compensation coefficient; calculating a system health limit coefficient based on the laser temperature and the system calculation load rate; and finally, according to the current scanning trigger frequency, the scanning state matching degree and the system health limit coefficient, dynamically calculating and adjusting the target frequency. Through multi-factor fusion analysis and health perception, self-adaptive regulation and control of the laser scanning frequency are realized, and the problems of under-sampling and over-sampling of point cloud data and equipment health risks caused by fixed frequency or only dependence on vehicle speed linear regulation in the prior art are solved.

Description

Technical Field

[0001] This invention belongs to the field of road infrastructure inspection technology, and in particular relates to a guardrail expansion joint inspection device and its inspection method. Background Technology

[0002] As a critical safety component in bridge and road infrastructure, guardrail expansion joints function to absorb deformation caused by structural thermal expansion and contraction and dynamic loads, ensuring driving stability and structural durability. Deformation, misalignment, or breakage directly jeopardizes the overall stability of the bridge and road safety, potentially leading to serious accidents. Traditional inspection methods rely heavily on manual inspections combined with handheld measuring equipment, resulting in time-consuming and costly operations. Furthermore, these methods heavily depend on the subjective experience of inspectors, making it difficult to guarantee the objectivity and repeatability of results. In addition, manual methods cannot achieve continuous, real-time monitoring and often respond slowly to sudden damage, failing to meet the high-frequency health assessment needs of modern infrastructure. In recent years, inspection technology based on mobile vehicle platforms integrating laser scanners has been gradually promoted. By dynamically acquiring point cloud data to construct a three-dimensional model of the expansion joint, it significantly improves the inspection coverage and data acquisition efficiency. However, existing mobile laser scanning systems still have significant shortcomings in actual operation. The scanning frequency control mechanism generally adopts a fixed threshold or a linear adjustment strategy based solely on vehicle speed, failing to fully incorporate dynamic operating parameters such as changes in vehicle acceleration, path curvature fluctuations, and sensor installation offsets. For example, when a vehicle accelerates, if the scanning frequency is not increased in time, the point cloud sampling interval will widen, resulting in data sparsity and an inability to accurately capture the minute deformations of the expansion joint. Conversely, in deceleration or sharp turning scenarios, a fixed frequency can easily lead to oversampling, generating a large amount of redundant point cloud data, increasing the burden on subsequent processing and reducing detection timeliness. Simultaneously, the system lacks real-time sensing capabilities for laser operating temperature and computing unit load status. Maintaining high-frequency scanning under high temperature or high load conditions not only accelerates equipment aging and increases energy consumption but may also lead to a decline in point cloud quality due to data processing delays or signal drift, and even trigger system failures. These deficiencies cause significant fluctuations in detection data under complex road conditions, failing to consistently meet the accuracy requirements of expansion joint morphology analysis and hindering the long-term reliable operation of the equipment.

[0003] To address the aforementioned issues, existing technologies urgently need improvement. Summary of the Invention

[0004] The purpose of this invention is to provide a guardrail expansion joint detection device and detection method, which aims to solve the above-mentioned problems.

[0005] This invention is implemented as follows: a method for detecting expansion joints in guardrails, comprising: calculating a spatial sampling requirement coefficient based on the driving speed and acceleration of a moving vehicle and the offset of a laser scanning instrument from the target; calculating a turning compensation coefficient based on path curvature and yaw rate; calculating a scanning state matching degree based on the point cloud density and point cloud signal-to-noise ratio under the spatial sampling requirement coefficient and the turning compensation coefficient; calculating a system health limitation coefficient based on the laser temperature and system computing load rate; and calculating a target laser scanning trigger frequency based on the current laser scanning trigger frequency, the scanning state matching degree, and the system health limitation coefficient, and adjusting the current laser scanning trigger frequency to the target laser scanning trigger frequency.

[0006] A further technical solution involves the following process for calculating and obtaining the target laser scanning trigger frequency: First, obtain the current laser scanning trigger frequency, scanning state matching degree, and system health limit coefficient. Second, based on the deviation between the current scanning trigger frequency and the scanning state matching degree, calculate the preliminary target frequency using a proportional adjustment method. That is, adjust the current laser scanning trigger frequency by increasing or decreasing it according to a certain proportional coefficient based on the difference between the current scanning state matching degree and the preset target matching degree, so that the current laser scanning trigger frequency changes towards optimizing the matching degree. Third, based on the system health limit coefficient, perform linear interpolation between the minimum and maximum trigger frequencies allowed by the system to obtain the upper limit of the frequency under the current system health limit coefficient. The higher the system health limit coefficient, the closer the allowed upper limit is to the maximum value; conversely, the lower the coefficient, the closer it is to the minimum value. Fourth, compare the preliminary target frequency with the minimum frequency and the upper limit allowed by the system, and take a reasonable value from the three: first, ensure that it is not lower than the minimum frequency, and then ensure that it does not exceed the upper limit of the frequency, thereby obtaining the final target scanning trigger frequency.

[0007] A further technical solution involves the following process for calculating and obtaining the system health limitation coefficient: obtaining the current laser temperature and system computational load rate; performing maximum-minimum normalization on the current laser temperature to obtain the laser temperature index; performing a ratio calculation on the current system computational load rate and its maximum allowable value to obtain the system computational load rate index; converting the temperature index and load rate index into corresponding health contribution factors, and then using a weighted product to obtain the system health limitation coefficient. The smaller the system health limitation coefficient, the less healthy the system is, and the greater the limitation on the scanning frequency needs to be.

[0008] A further technical solution involves the following process for calculating and obtaining the scanning state matching degree: First, obtain the current spatial sampling requirement coefficient, turning compensation coefficient, point cloud density, and point cloud signal-to-noise ratio (SNR). Second, based on the spatial sampling requirement coefficient and turning compensation coefficient, and combined with the acceptable minimum point cloud density, calculate the desired point cloud density under the current operating conditions. The desired point cloud density increases with the increase of the spatial sampling requirement coefficient and turning compensation coefficient. Third, based on the ratio of the actual point cloud density to the desired point cloud density, normalize using a logarithmic function to obtain the density matching degree. The density matching degree increases with the increase of the actual density relative to the desired density, but the growth rate gradually slows down. Fourth, based on the ratio of the actual point cloud SNR to a reference value determined by the desired density and the minimum density, use an exponentially decaying function to obtain the SNR matching degree. The higher the SNR, the closer the matching degree is to 1; conversely, it approaches 0. Finally, substitute the density matching degree and the SNR matching degree into the formula... Get the matching degree of the scan status , , A higher value indicates a more ideal scanning frequency setting and a better match between data quality and detection requirements. For density matching degree, This represents the signal-to-noise ratio matching degree.

[0009] A further technical solution involves the following process for calculating the spatial sampling demand coefficient: First, the vehicle's speed and acceleration, as well as the offset of the laser scanner from the target, are obtained. Then, the real-time speed of the vehicle and the laser scanner's offset from the target are compared with the system's maximum designed detection speed and the system's maximum allowable offset, respectively, to obtain the speed index and offset index. Next, the vehicle's acceleration is mapped to a speed acceleration index using a centrally symmetric S-shaped function. The speed acceleration index takes the median value when the acceleration is zero, is greater than the median value during acceleration, and is less than the median value during deceleration, thus reflecting the need to increase the sampling frequency during acceleration and decrease the sampling frequency during deceleration. Finally, the speed index, acceleration index, and offset index are linearly weighted and summed according to preset weights to obtain the spatial sampling demand coefficient. This coefficient comprehensively reflects the vehicle's motion state's demand for sampling density; a larger value indicates a higher demand.

[0010] A further technical solution involves the following process for obtaining the turning compensation coefficient: obtaining the path curvature and yaw rate; comparing the absolute values ​​of the path curvature and yaw rate with the design maximum curvature and design maximum yaw rate respectively, and then using a min function to limit the ratio to an upper limit of 1 to obtain the path curvature index and yaw rate index; and obtaining the turning compensation coefficient by weighting the path curvature index and yaw rate index using the maximum and average values. The turning compensation coefficient is close to 0 when driving straight, and increases with the increase of either the path curvature index or the yaw rate index when turning. The compensation is stronger when both the path curvature index and the yaw rate index are high.

[0011] A guardrail expansion joint detection device includes a mobile vehicle and a laser scanner installed on the mobile vehicle; a memory for storing a computer program; and a processor for implementing the steps of the guardrail expansion joint detection method described above when executing the computer program.

[0012] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0013] 1. Improve data quality and detection accuracy: By comprehensively considering vehicle speed, acceleration, offset, and turning status, the scanning frequency is dynamically adjusted to ensure that the point cloud density and signal-to-noise ratio meet the detection requirements under different driving conditions. This avoids the loss of details caused by undersampling and the redundant data caused by oversampling, and significantly improves the accuracy of expansion joint deformation measurement.

[0014] 2. Enhance system adaptability and operational stability: Real-time sensing of laser temperature and computational load establishes system health constraint coefficients, ensuring data quality while safely constraining scanning frequency, preventing equipment overload, extending the lifespan of the laser scanner, and reducing the risk of system failure.

[0015] 3. Improved detection efficiency and intelligence: The scanning parameters can be automatically optimized according to real-time working conditions without manual intervention, adapting to complex road conditions and dynamic changes, realizing continuous and intelligent operation of guardrail expansion joint detection, and providing an efficient and reliable technical means for infrastructure health assessment. Attached Figure Description

[0016] Figure 1 A flowchart of a method for detecting expansion joints in guardrails provided by the present invention;

[0017] Figure 2 This is a schematic diagram of the structure of a guardrail expansion joint detection device provided by the present invention.

[0018] In the attached diagram: 1. Mobile vehicle; 2. Laser scanner. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0020] In traditional guardrail expansion joint detection technologies, the scanning frequency adjustment strategy used in mobile laser scanning systems is not designed to adapt to the dynamic movement of vehicles and the health of the system, resulting in unstable point cloud data quality. Specifically, fixed or speed-based linear adjustment methods do not consider factors such as vehicle acceleration, sensor offset, and turning, causing the sampling density to deviate from the ideal value during acceleration, deceleration, or turning, leading to oversampling or undersampling. Furthermore, the system lacks real-time sensing of laser temperature and computational load, making it unable to automatically adjust the operating frequency when the equipment is not in good working order, thus increasing the risk of equipment failure and the probability of data distortion.

[0021] For example, in the actual inspection of bridge expansion joints, when the moving vehicle is traveling in a straight line, a fixed scanning frequency can meet the point cloud density requirements. However, when the vehicle enters a curve and the path curvature increases and the yaw rate increases, if the scanning frequency is not compensated in time, the point cloud density will decrease, resulting in distortion of the expansion joint morphology model. Furthermore, if the laser temperature rises and the system's computational load is too high, continuous high-frequency scanning may cause data processing delays or equipment protective shutdown.

[0022] If the above problems are not addressed, the signal-to-noise ratio and density of point cloud data will not be able to match the detection requirements, and minor deformations or misalignments of expansion joints may not be accurately identified, thus affecting the reliability of structural safety assessments. In addition, continuous operation of the system in an unhealthy state will accelerate the aging of key components, shorten the service life of equipment, and may cause data acquisition failures in important detection tasks, thereby compromising the accuracy of infrastructure maintenance decisions.

[0023] The specific implementation of the present invention will be described in detail below with reference to specific embodiments.

[0024] like Figure 1 As shown, an embodiment of the present invention provides a method for detecting expansion joints in guardrails, comprising:

[0025] Based on the vehicle's speed and acceleration, as well as the laser scanner's distance from the target, a spatial sampling requirement coefficient is calculated. This coefficient is a quantitative indicator that measures the density or precision of point cloud data sampling of the target area under dynamic conditions such as specific speed, acceleration, and laser scanner distance from the target. This coefficient reflects the sampling density level required to ensure data quality under different motion states.

[0026] The turning compensation coefficient is calculated based on the path curvature and yaw rate. The turning compensation coefficient is an adjustment factor introduced to compensate for sampling unevenness or data loss that may occur due to changes in path curvature and yaw rate when the moving vehicle is turning. This coefficient aims to ensure that sufficient point cloud sampling density and quality are maintained during vehicle turning.

[0027] Based on the point cloud density and point cloud signal-to-noise ratio under the spatial sampling requirement coefficient and turning compensation coefficient, the scanning state matching degree is calculated. Point cloud density refers to the number of point cloud data points contained in a unit spatial volume or unit area. This indicator directly reflects the fineness of the point cloud data; the higher the density, the stronger the ability to depict the details of the object surface. Point cloud signal-to-noise ratio refers to the ratio between the effective signal and the noise signal in the point cloud data. This indicator is used to evaluate the quality of the point cloud data; the higher the signal-to-noise ratio, the less the data is affected by noise, and the better its reliability and usability. Scanning state matching degree refers to the degree of conformity between the current laser scanning parameter settings and the actual detection requirements and data quality. This matching degree comprehensively considers factors such as point cloud density and point cloud signal-to-noise ratio to evaluate the effectiveness of the current scanning strategy.

[0028] Based on laser temperature and system computational load rate, a system health limitation coefficient is calculated. Laser temperature refers to the operating temperature of the laser emitting device inside the laser scanning instrument. This temperature is one of the key parameters affecting laser performance, stability, and lifespan; excessively high temperatures may lead to performance degradation or even damage. System computational load rate refers to the proportion of computing resources used by the laser scanning system for data processing, algorithm execution, and other computational tasks. This load rate reflects the current workload of the system; excessively high load rates may lead to slow system response, data processing delays, or system crashes. The system health limitation coefficient is a quantitative indicator used to assess and impose limitations on the overall health of the system based on equipment operating status parameters such as laser temperature and system computational load rate. This coefficient is used to appropriately limit the laser scanning trigger frequency when the system health is poor, in order to protect the equipment and maintain stable system operation.

[0029] Based on the current laser scanning trigger frequency, scanning status matching degree, and system health limitation coefficient, the target laser scanning trigger frequency is calculated and obtained, and the current laser scanning trigger frequency is adjusted to the target laser scanning trigger frequency. The laser scanning trigger frequency refers to the number of laser pulses emitted by the laser scanning instrument per unit time. This frequency directly determines the acquisition rate and spatial sampling density of point cloud data, and is a key parameter affecting detection efficiency and data quality.

[0030] Compared to existing technologies that often use fixed scanning frequencies or simple linear adjustments based on vehicle speed, this embodiment introduces spatial sampling demand coefficients and turning compensation coefficients to achieve refined perception of the dynamic motion state of the moving vehicle. For example, in a scenario where the moving vehicle 1 enters a curve and accelerates, existing technologies may fail to adequately consider factors such as path curvature, yaw rate, acceleration, and offset, leading to undersampling or oversampling of point cloud data, thus affecting detection accuracy. This embodiment, however, can calculate and adjust the spatial sampling demand coefficients and turning compensation coefficients in real time based on changes in these dynamic factors, thereby more accurately assessing the scanning state matching degree and ensuring that ideal point cloud sampling density and quality are maintained even under complex driving conditions.

[0031] Furthermore, existing systems generally lack awareness and adaptive control of the device's own state, continuing to operate at high frequencies even when the laser temperature is too high or the system's computational load is too heavy. This not only increases energy consumption but may also affect the device's lifespan and data stability. This embodiment innovatively introduces a system health limit coefficient, which comprehensively considers key health indicators such as laser temperature and system computational load rate. In the example above, when the laser temperature rises, the system health limit coefficient is calculated to decrease. Even if the data quality requirements are high, the system will limit the upper limit of the target laser scanning trigger frequency, thereby effectively avoiding device overload operation, extending the lifespan of the laser scanning device 2, and ensuring the long-term stability and reliability of the system.

[0032] In summary, this embodiment achieves intelligent dynamic adjustment of the laser scanning frequency through multi-factor fusion analysis (including spatial sampling requirements, turning compensation, and scanning state matching) and health perception (system health limitations). This adaptive control mechanism not only significantly improves the point cloud data quality and detection efficiency of guardrail expansion joint detection, but also effectively balances system health and operational safety, thereby comprehensively improving the reliability, accuracy, and adaptability of the guardrail expansion joint detection device and solving key technical problems existing in the prior art.

[0033] This application further proposes the following procedure for calculating and obtaining the target laser scanning trigger frequency:

[0034] The system acquires the current laser scan trigger frequency, scan status matching degree, and system health limitation coefficient. The current laser scan trigger frequency represents the system's current operating state; the scan status matching degree reflects the degree to which the current scan data quality meets the detection requirements; and the system health limitation coefficient quantifies the system's current operational health and tolerable load. These parameters can be obtained through real-time sensor acquisition, internal system status monitoring, or preliminary calculation steps. For example, the current laser scan trigger frequency can be directly read from the laser scanner's control module; the scan status matching degree can be calculated by a dedicated data processing module based on a point cloud data quality assessment algorithm; and the system health limitation coefficient can be obtained by the system health monitoring module through a comprehensive evaluation based on parameters such as laser temperature and computational load rate.

[0035] Based on the deviation between the current scan trigger frequency and the scan state matching degree, a preliminary target frequency is calculated using a proportional adjustment method. Specifically, the current laser scan trigger frequency is adjusted by increasing or decreasing it according to a certain proportional coefficient based on the difference between the current scan state matching degree and the preset target matching degree, so that the current laser scan trigger frequency changes towards the direction of optimized matching degree. The specific calculation method is as follows: substitute the current laser scan trigger frequency and the scan state matching degree into the formula... Obtain the initial target frequency This step aims to initially adjust the laser scanning trigger frequency based on the gap between the current scan data quality and the desired target. The formula employs a proportional (P) control concept, where... This is the current laser scanning trigger frequency. This is a proportionality coefficient used to control the sensitivity or intensity of the adjustment. The target value for the scanning state matching degree, The current scan state matching degree; when the current scan state matching degree Below the target value of scan state matching At that time, the initial target frequency It will increase to improve sampling density and data quality; conversely, when Higher than hour, This will be reduced to avoid oversampling and wasted resources. Scale factor It can be set empirically or dynamically adjusted through adaptive algorithms based on the actual application scenario and system response characteristics.

[0036] Based on the system health limit coefficient, a linear interpolation is performed between the minimum and maximum allowed trigger frequencies to obtain the upper limit of the frequency under the current system health limit coefficient. The higher the system health limit coefficient, the closer the allowed upper limit is to the maximum value; conversely, the lower the coefficient, the closer it is to the minimum value. The specific calculation method is as follows: substitute the system health limit coefficient into the formula. Get the frequency limit This step is used to dynamically set a frequency upper limit based on the system's health status to protect the system and ensure its stable operation. The minimum triggering frequency allowed by the system. The maximum triggering frequency allowed by the system. and It is an inherent constraint of the system hardware and software. The system health constraint coefficient is a value between 0 and 1, reflecting the health level of the system. When the system is in good health (…),… When approaching 1), the upper frequency limit Approaching This allows the system to run at a higher frequency; when the system health is poor ( When it is close to 0, Approaching This limits the system's operation under high loads, preventing overload or damage. This dynamic upper limit setting allows frequency adjustment to balance data quality with system security and stability.

[0037] The initial target frequency is compared with the system's minimum and upper frequency limits, and a reasonable value among the three is selected: first, ensure it is not lower than the minimum frequency, then ensure it does not exceed the upper frequency limit, thus obtaining the final target scan trigger frequency; the specific calculation method is: substitute the initial target frequency and the upper frequency limit into the formula. Obtain the target laser scanning trigger frequency ,in, For the initial target frequency, The minimum triggering frequency allowed by the system. This is the upper limit of the frequency; this step is crucial for ultimately determining the target laser scanning trigger frequency. It ensures the final target frequency through nested min / max functions. It will not fall below the minimum trigger frequency allowed by the system. It will not exceed the dynamic frequency limit determined by the system's health status. .first, Ensure the initial target frequency Even if the calculated result is too low, it will not fall below the minimum frequency at which the system can function normally. Then, This result was further compared with the upper limit of dynamic frequency. The two values ​​are compared, and the smaller value is taken to ensure that the final target frequency is within the range allowed by system health. This step-by-step restriction strategy allows frequency adjustments to both respond to data quality requirements and strictly adhere to the safety boundaries of system operation.

[0038] This application's solution achieves intelligent and adaptive control of the laser scanning trigger frequency through a multi-stage frequency adjustment logic. First, the system acquires the current laser scanning trigger frequency, scanning status matching degree, and system health limit coefficient in real time. These parameters comprehensively reflect the current operating status, data quality requirements, and system capacity. Next, based on the current laser scanning trigger frequency and scanning status matching degree, a preliminary target frequency is calculated through a proportional control mechanism. This preliminary target frequency is designed to adjust according to the deviation between the current scan data quality and the desired target, quickly responding to data quality demands. Simultaneously, the system dynamically calculates a frequency upper limit using the system health limit coefficient. This upper limit is adjusted in real time based on health indicators such as laser temperature and system computational load rate, ensuring that while pursuing data quality, the system is not overloaded or damaged. Finally, the preliminary target frequency is comprehensively judged against the system's minimum allowable trigger frequency and the dynamically calculated frequency upper limit. By taking the maximum and minimum values, a target laser scanning trigger frequency that satisfies data quality requirements while strictly adhering to the system's safe operating boundaries is ultimately determined. This layered, multi-factor comprehensive adjustment mechanism makes the adjustment process of laser scanning trigger frequency more stable, safe and efficient, avoiding the risk of system instability or equipment damage that may be caused by adjusting a single factor, thus continuously providing high-quality guardrail expansion joint detection data in complex and ever-changing environments.

[0039] The following is a concrete example. During the detection of guardrail expansion joints, the processor can periodically execute a frequency adjustment algorithm. For example, every second, the processor first reads the current laser scanning trigger frequency from the control interface of laser scanning device 2, obtains the currently calculated scan state matching degree from the data processing module, and obtains the current system health limitation coefficient from the system health monitoring module. Assume the current laser scanning trigger frequency is 100kHz, the scan state matching degree is 0.7, and the system health limitation coefficient is 0.8. The preset target value for the scan state matching degree... The ratio is 0.9, the proportionality coefficient. The minimum allowed trigger frequency is 0.5. 50kHz, maximum trigger frequency The frequency is 200kHz. The processor substitutes the current laser scan trigger frequency of 100kHz and the scan state matching degree of 0.7 into the formula. The calculated frequency is 110kHz, thus obtaining the initial target frequency. The frequency is 110kHz. Next, the processor substitutes the system health limit factor of 0.8 into the formula. The calculation yields 170kHz, thus providing the upper frequency limit. The frequency is 170kHz. Finally, the processor substitutes the initial target frequency of 110kHz, the minimum allowed trigger frequency of 50kHz, and the upper frequency limit of 170kHz into the formula. The calculated frequency was 110 kHz, which was then used to determine the target laser scanning trigger frequency. The system will then adjust the trigger frequency of the laser scanner 2 to 110kHz. In this way, the system can dynamically and safely adjust the scanning frequency based on real-time data quality requirements and its own health status.

[0040] Through the above technical solution, this application provides a more refined, adaptive, and secure laser scanning trigger frequency adjustment mechanism. This mechanism not only dynamically adjusts the frequency based on the matching degree between scan data quality and detection requirements, ensuring high-quality point cloud data in different detection environments, but also, by introducing a system health constraint coefficient, can assess the operating status of the system (such as the laser) in real time and dynamically set the frequency upper limit accordingly. This effectively avoids problems such as system overload, equipment damage, or operational instability caused by pursuing high data quality. This frequency adjustment strategy, which comprehensively considers data quality requirements and system health status, makes the guardrail expansion joint detection process more stable and reliable, extends equipment lifespan, and improves overall detection efficiency and safety.

[0041] This application further proposes the following process for calculating and obtaining the system health limitation coefficient:

[0042] The system acquires the current laser temperature and system compute load. Laser temperature is a key parameter reflecting the laser's operating status; excessively high temperatures can lead to performance degradation, shortened lifespan, or even permanent damage. System compute load reflects the processor's workload; excessive load can cause data processing delays, reduced real-time performance, or system crashes. These parameters can be acquired in real time using sensors integrated within the laser scanning device 2 or a system monitoring module. For example, laser temperature can be measured using a thermistor or infrared temperature sensor, while system compute load can be obtained through operating system APIs or hardware monitoring interfaces to determine CPU, GPU, or memory usage.

[0043] The current laser temperature is then normalized to its maximum and minimum values ​​to obtain the laser temperature index. This step aims to map the raw temperature value to a standardized, dimensionless range for comprehensive evaluation with other health indicators. For example, the laser temperature can be linearly normalized between its minimum permissible operating temperature and maximum safe operating temperature, or a nonlinear function can be used for mapping to more accurately reflect the impact of temperature on system health.

[0044] The system compute load factor index is obtained by comparing the current system compute load rate with its maximum allowable load. Its purpose is to quantify the proportion of the current compute load relative to the system's limits. This allows for standardized comparisons of load conditions across different systems or time periods. For example, if the system's maximum allowable compute load rate is 100% and the current load rate is 80%, then the load factor index is 0.8.

[0045] The temperature index and load rate index are converted into corresponding health contribution factors, and then a weighted product is used to obtain the system health limitation coefficient. The smaller the system health limitation coefficient, the less healthy the system is, and the more severely the scanning frequency needs to be limited. The specific calculation method is as follows: substitute the current laser temperature index and the calculated system load rate index into the formula. Obtain the system health limit coefficient , , The smaller the value, the less healthy the system, requiring greater frequency limitation. and All are health restriction weights ranging from 0 to 1, and , and This allows for adjusting the relative importance of temperature and computational load to system health limits based on the specific application scenario and system characteristics. For example, in temperature-sensitive lasers, this can be achieved by... Higher weight, and Specifically, values ​​can be assigned through preset strategies or dynamic algorithms; This refers to the laser temperature index. Calculate the load rate index for the system. and The index represents the reciprocal of the health level; that is, the higher the index, the lower the health level, and the smaller the value of 1 - index.

[0046] The proposed solution acquires the laser temperature and system computed load rate in real time, and standardizes them to obtain the laser temperature index and system computed load rate index. Then, these two indices are substituted into a weighted product formula to comprehensively calculate the system health limitation coefficient. This coefficient directly reflects the current health status of the system; the smaller the value, the less healthy the system, thus requiring stricter limits on the laser scanning trigger frequency. This calculation method allows the system to dynamically assess its own operating status and use it as an important basis for adjusting the scanning frequency. By using this system health limitation coefficient... The target laser scanning trigger frequency is determined by factors such as the matching degree of the scanning state, ensuring that the system always operates within a safe and stable range while meeting the data acquisition requirements, effectively avoiding performance degradation or equipment damage caused by overload or overheating.

[0047] The following is a concrete example. Assume that during the inspection of guardrail expansion joints, the system monitors the current laser temperature at 70℃, and the system's computational load rate is 75%. The maximum allowable safe temperature for the laser is 80℃, and the minimum operating temperature is 0℃. The system's maximum allowable computational load rate is 100%. To calculate the system's health limit coefficient, the current laser temperature is first normalized to its maximum and minimum values ​​to obtain the laser temperature index. The value is 0.875. Next, the current system compute load rate is compared to its maximum allowable value to obtain the system compute load rate index. The value is 0.75. Assume the health restriction weight is... Set to 0.6, Set it to 0.4. Substitute these indices and weights into the formula. The calculation yields 0.29. The result will provide a specific... The value, for example, is approximately 0.29. This calculated value... The value (0.29) will be used as a quantitative indicator of the system's health status and will be used to calculate the trigger frequency for subsequent target laser scanning, thereby guiding the system to adjust the frequency under the current health status.

[0048] Through the above technical solution, this application provides an accurate and dynamic system health assessment mechanism. This mechanism comprehensively considers laser temperature and system computational load rate, and uses a weighted product formula to generate system health limit coefficients, enabling the system to perceive its own operating status in real time. This ensures that when adjusting the laser scanning trigger frequency, not only data acquisition requirements are considered, but also the system's tolerance is fully taken into account, thereby effectively avoiding the risks of performance degradation, data quality deterioration, or equipment damage caused by system overload or overheating, and significantly improving the long-term stability and reliability of the guardrail expansion joint detection system.

[0049] This application further proposes the following process for calculating and obtaining the matching degree of the scan state:

[0050] The system acquires the current spatial sampling demand coefficient, turning compensation coefficient, point cloud density, and point cloud signal-to-noise ratio. The spatial sampling demand coefficient reflects the dynamic demand of the moving vehicle's motion state (such as speed, acceleration, and offset) on the spatial sampling density of the point cloud. It can be obtained from vehicle motion data acquired by onboard sensors (e.g., GPS, IMU) and comprehensively evaluated in conjunction with the relative position information of the laser scanner and the target object, or it can be achieved by preset sampling demand levels for different motion states and matching corresponding coefficients. The turning compensation coefficient reflects the additional demand for point cloud sampling density when the vehicle turns. It can be calculated by the path planning system or vehicle dynamic perception system based on the vehicle's path curvature and yaw rate. For example, it can be estimated in real time by analyzing the vehicle's steering angle, wheel speed difference, or IMU data, or it can be activated based on preset turning radius or angular velocity thresholds. Point cloud density refers to the number of points in a unit volume or area. It is an important indicator for measuring the spatial resolution of point cloud data. It can be calculated by processing point cloud data acquired by a laser scanner in real time. For example, it can be calculated by counting the number of points in a local area and dividing by the volume or area of ​​that area, or indirectly by analyzing the average distance between adjacent points in the point cloud data. Point cloud signal-to-noise ratio (SNR) measures the ratio of effective signal to noise in point cloud data. It is a key indicator for evaluating the quality of point cloud data. It can be calculated by analyzing the original laser echo signal. For example, it can be calculated by comparing the peak intensity of the echo signal with the background noise level, or indirectly by counting the proportion of outliers or anomalies in the point cloud data.

[0051] Based on the spatial sampling demand coefficient and the turning compensation coefficient, and combined with the acceptable minimum point cloud density, the desired point cloud density under the current operating conditions is calculated. The desired point cloud density increases with the increase of the spatial sampling demand coefficient and the turning compensation coefficient. The specific calculation method is as follows: substitute the current spatial sampling demand coefficient and the turning compensation coefficient into the formula. Obtain the desired point cloud density ,in, For the lowest acceptable point cloud density, This is the spatial sampling requirement coefficient. This step aims to dynamically determine an ideal point cloud density target based on the current motion state and environmental requirements, using a turning compensation coefficient. The formula combines a weighted combination of the spatial sampling requirement coefficient and the turning compensation coefficient with the acceptable minimum point cloud density. These factors are combined to ensure that a reasonable and adaptive desired density can be set under different operating conditions. For example, when the vehicle is traveling at high speed or making a sharp turn, these two coefficients will increase, thereby increasing the desired point cloud density to ensure a sufficient sampling rate.

[0052] The density matching degree is obtained by normalizing the ratio of the actual point cloud density to the desired point cloud density using a logarithmic function. The density matching degree increases with the increase of the actual density relative to the desired density, but the rate of increase gradually slows down. The specific calculation method is as follows: substitute the current point cloud density and the desired point cloud density into the formula. Obtain density matching degree ,in, Point cloud density, The desired point cloud density is defined here; this step quantifies the degree of matching between the current actual point cloud density and the desired point cloud density. This logarithmic function maps the density ratio to a matching degree value, such that the matching degree approaches 1 when the actual density is close to or exceeds the desired density, and approaches 0 when the actual density is significantly lower than the desired density. This helps assess whether the current scan frequency provides sufficient spatial sampling.

[0053] The signal-to-noise ratio (SNR) of the actual point cloud is compared to a reference value determined by the desired density and the minimum density. An exponential decay function is used to obtain the SNR matching degree; the higher the SNR, the closer the matching degree is to 1; conversely, it approaches 0. The specific calculation method is as follows: substitute the current point cloud SNR and the desired point cloud density into the formula. Obtain the signal-to-noise ratio matching degree ,in, For point cloud signal-to-noise ratio, For the desired point cloud density, This step determines the minimum acceptable point cloud density and assesses the match between the quality of the point cloud data (signal-to-noise ratio) and the detection requirements (desired point cloud density). The exponential function formula maps the ratio of signal-to-noise ratio to desired density to a matching score. A matching score approaches 1 when the signal-to-noise ratio is high and the desired density requirement is met, and approaches 0 otherwise. This helps ensure that data quality meets requirements while spatial sampling is satisfied.

[0054] Substitute density matching degree and signal-to-noise ratio matching degree into the formula Get the matching degree of the scan status , , A higher value indicates a more ideal scanning frequency setting and a better match between data quality and detection requirements. For density matching degree, This represents the signal-to-noise ratio matching degree. This step is crucial for a comprehensive evaluation of the scan status. The formula incorporates density matching degree. and signal-to-noise ratio matching degree The final scan state matching degree is calculated using geometric mean and exponential decay term. Geometric mean ensures that when both matches are highly accurate... The high density and signal-to-noise ratio (SNR) matching factors penalize situations where the difference between two matching degrees is too large; that is, when density and SNR matching degrees are unbalanced, the overall matching degree will decrease. This ensures that the evaluation of the scan status is comprehensive and balanced, avoiding focusing on only a single indicator while ignoring other important factors.

[0055] The proposed solution provides crucial input for the dynamic adjustment of the laser scanning trigger frequency by calculating the scanning state matching degree in detail. The calculation process first acquires real-time data such as spatial sampling requirement coefficient, turning compensation coefficient, point cloud density, and point cloud signal-to-noise ratio (SNR). This data comprehensively reflects the motion state of the moving vehicle, environmental complexity, and the quality of the current point cloud data. Next, based on the spatial sampling requirement coefficient and turning compensation coefficient, the desired point cloud density is dynamically calculated, enabling the system to adjust the point cloud density requirements according to actual working conditions. Subsequently, the density matching degree between the current point cloud density and the desired point cloud density, as well as the SNR matching degree between the point cloud density and the desired point cloud density, are calculated separately, quantifying the suitability of the current scanning state from two dimensions: spatial resolution and data quality. Finally, a comprehensive formula is used to fuse the density matching degree and the SNR matching degree, and a penalty mechanism for the difference between the two is introduced to obtain a comprehensive, balanced, and robust scanning state matching degree. This precise matching degree of the scanning state serves as an important basis for calculating the trigger frequency of the subsequent target laser scan, enabling the system to adjust the laser scan trigger frequency more intelligently and accurately, ensuring that the data acquisition efficiency and quality are optimized while meeting the requirements for guardrail expansion joint detection.

[0056] As a specific implementation method, it is assumed that during the inspection of guardrail expansion joints, the current spatial sampling demand coefficient of the mobile vehicle 1 is... The turning compensation coefficient is 0.6. The value is 0.2. This represents the real-time point cloud density. The point cloud signal-to-noise ratio is 900 points / square meter. The minimum acceptable point cloud density is 10 dB. The value is 500 points per square meter. First, according to the formula... Calculate the desired point cloud density Substituting the numerical values, we obtain 900 points per square meter. Next, we calculate the density matching degree. According to the formula Substituting the values, we get 1. Then, we calculate the signal-to-noise ratio matching degree. According to the formula Substituting the values, we obtain 0.9962. Finally, we calculate the scan state matching degree. According to the formula Substituting the values, we get 0.9962. The final scan state matching degree is... The value of 0.9962 indicates that the current scanning status is highly matched with the detection requirements.

[0057] By comprehensively considering spatial sampling requirement coefficients, turning compensation coefficients, point cloud density, and point cloud signal-to-noise ratio (SNR), this application can dynamically calculate the desired point cloud density and further evaluate the density matching degree and SNR matching degree. Finally, by fusing these matching degrees, a comprehensive and balanced scanning state matching degree is obtained. This allows the adjustment of the laser scanning trigger frequency to no longer rely solely on a single or limited set of indicators, but rather to accurately reflect the complexity of the current detection environment, the motion state of the moving vehicle, and the quality of the actual point cloud data. Therefore, the system can more intelligently and adaptively adjust the scanning frequency, ensuring that high-quality point cloud data meeting the requirements for guardrail expansion joint detection can be acquired with optimal efficiency under different working conditions. This avoids resource waste caused by oversampling or insufficient data caused by undersampling, significantly improving the accuracy and reliability of the detection.

[0058] This application further proposes the following process for calculating the spatial sampling demand coefficient:

[0059] The system acquires the speed and acceleration of the mobile vehicle, as well as the offset of the laser scanner from the target. The speed of the mobile vehicle 1 refers to its instantaneous speed during the detection process, which directly affects the distribution density of laser points on the target surface per unit time. Acceleration refers to the rate of change of vehicle speed; when the vehicle accelerates or decelerates, the scanning frequency needs to be proactively adjusted to maintain sampling uniformity. The offset of the laser scanner 2 from the target refers to the lateral distance between the scanning centerline of the laser scanner 2 and the target (such as a guardrail expansion joint). This offset affects the projection shape and density of the laser points on the target, especially at the edges of the scanning range. Acquiring these parameters is fundamental to the subsequent calculation of the spatial sampling requirement coefficient, ensuring a comprehensive understanding of the vehicle's motion state and the target's relative position. These parameters can be acquired in various ways. For example, the speed and acceleration of the mobile vehicle 1 can be acquired in real time via an onboard inertial measurement unit (IMU) or global positioning system (GPS) module, or read through the vehicle's CAN bus interface. The offset of the laser scanner 2 from the target can be calculated in real time using the echo data of the laser scanner 2 itself, for example, by analyzing the lateral position of the target features in the point cloud data, or by measuring it using an additional visual sensor (such as a camera) combined with image processing technology.

[0060] The vehicle's speed and the real-time offset of the laser scanner from the target are compared with the system's designed maximum detection speed and the system's maximum allowable offset, respectively, to obtain the speed index and offset index. This step aims to normalize the original physical quantities (speed and offset) into dimensionless exponents for subsequent comprehensive calculations. By comparing these with the system's preset maximum values, parameters of different dimensions can be unified to a range of 0 to 1, reflecting their relative levels within the system's capabilities. The speed index represents the proportion of the current speed to the system's designed maximum detection speed, reflecting the contribution of speed to sampling requirements. The offset index represents the proportion of the current offset to the system's maximum allowable offset, reflecting the contribution of offset to sampling requirements. This normalization process allows different physical quantities to be weighted and combined within a unified framework, thereby more rationally assessing spatial sampling requirements. The ratio processing can use simple division operations, such as dividing the current speed by the system's designed maximum detection speed and dividing the current offset by the system's maximum allowable offset. In practical applications, to avoid division by zero or to handle outliers, the maximum value can be set appropriately, and the calculation results can be limited to ensure that the exponent value is within a reasonable range, such as between 0 and 1.

[0061] The driving acceleration is mapped to a driving acceleration exponent using a centrally symmetric S-shaped function. The driving acceleration exponent takes the median value when the driving acceleration is zero, is greater than the median value during acceleration, and is less than the median value during deceleration. This reflects the requirement to increase the sampling frequency during acceleration and decrease the sampling frequency during deceleration. The specific calculation method is as follows: substitute the driving acceleration of the moving vehicle into the formula. Obtain the driving acceleration index , , This indicates that the vehicle is traveling at a constant speed (with zero acceleration) and is in a state meeting basic requirements. This indicates that the vehicle is accelerating, and the scanning frequency needs to be increased in advance to avoid increasing the sampling interval. This indicates that the vehicle is decelerating; the scanning frequency can be appropriately reduced to avoid oversampling. The acceleration of the moving vehicle. This serves as a reference value for driving acceleration; this step aims to convert the driving acceleration of the moving vehicle 1 into an index that reflects its impact on spatial sampling requirements. This formula utilizes the properties of the hyperbolic tangent function (tanh) to express acceleration. Mapped to the range (0,1), with 0.5 as the reference point. When the vehicle travels at a constant speed (acceleration...) When it is 0), A value of 0.5 indicates that the system is in a baseline sampling requirement state. When the vehicle accelerates, A value greater than 0.5 indicates a need to increase the scanning frequency to avoid increasing the sampling interval and thus maintain sampling density. When the vehicle decelerates, A value less than 0.5 indicates that the scan frequency can be appropriately reduced to avoid oversampling. This nonlinear mapping can more precisely capture the impact of acceleration on sampling requirements, especially when acceleration changes significantly. This serves as a reference value for driving acceleration and can be calibrated based on actual application scenarios and system performance. For example, experimental testing or simulation analysis can be used to determine at what acceleration levels a significant adjustment to the scanning frequency is necessary, thereby setting an appropriate value. The formula can be implemented using floating-point operations performed by the processor, where the hyperbolic tangent function can be implemented through mathematical library function calls or numerical approximation methods.

[0062] The driving speed index, driving acceleration index, and offset index are linearly weighted and summed according to preset weights to obtain the spatial sampling demand coefficient. The spatial sampling demand coefficient comprehensively reflects the vehicle's motion state's requirement for sampling density; a larger value indicates a higher demand. The specific calculation method is as follows: substitute the driving speed index, driving acceleration index, and offset index into the formula. Obtain the spatial sampling requirement coefficient , , This timeframe indicates that the vehicle is stationary, with no acceleration and no deviation, requiring the lowest sampling rate. The time represents the vehicle's maximum speed, maximum positive acceleration, and maximum offset, with the highest sampling requirement. , and All are weighting coefficients with values ​​ranging from 0 to 1, and , The driving speed index, The acceleration index is the driving speed index. This is the offset exponent. This step aims to comprehensively consider the combined effects of driving speed, driving acceleration, and offset on spatial sampling requirements through weighted summation, thereby obtaining the final spatial sampling requirement coefficient. This coefficient is a value between 0 and 1, intuitively reflecting the current demand for laser scanning frequency. When When the value is close to 0, it indicates that the vehicle is stationary, has no acceleration, and no deviation, requiring the lowest sampling rate; when... When the value is close to 1, it indicates that the vehicle is traveling at its highest speed, maximum positive acceleration, and maximum deviation, indicating the highest sampling requirement. This is achieved by assigning weighting coefficients to different indices. , and The weighting coefficients can be flexibly adjusted to influence the total sampling requirements based on the actual testing scenario and the sensitivity to different factors. , and The setting needs to satisfy the condition that its value range is 0-1 and the sum is 1, to ensure The rationality of these weighting coefficients can be ensured through expert experience, historical data analysis, or machine learning methods, so that the calculated... It can most accurately reflect the actual sampling requirements. The calculation of this formula can be completed by the processor performing simple multiplication and addition operations.

[0063] In the method for detecting expansion joints in guardrails, accurate assessment of spatial sampling requirements under the current environment is necessary to achieve intelligent adjustment of the laser scanning trigger frequency. This solution constructs a dynamic spatial sampling requirement coefficient by comprehensively considering the driving speed and acceleration of the moving vehicle 1, as well as the offset of the laser scanning instrument 2 from the target. Specifically, firstly, the system acquires in real-time the speed and acceleration of the moving vehicle 1, as well as the offset of the laser scanner 2 from the target. This raw data is crucial information reflecting the vehicle's motion state and the target's relative position. To unify these physical quantities with different dimensions for comprehensive evaluation, the real-time speed of the moving vehicle 1 and the real-time offset of the laser scanner 2 from the target are compared with the system's designed maximum detection speed and the system's maximum allowable offset, respectively, to obtain a dimensionless speed index. and offset index This normalization process allows the influence of different factors to be compared and weighted on a uniform scale. Simultaneously, the acceleration of moving vehicle 1... Substituted into a nonlinear function To obtain the driving acceleration index This function cleverly maps acceleration to the (0,1) interval and uses 0.5 as the baseline for constant speed travel. This causes the exponent to increase during acceleration (indicating a need to increase the frequency) and decrease during deceleration (indicating a need to decrease the frequency), thus proactively addressing the impact of vehicle speed changes on the sampling interval. Finally, the resulting speed exponent is... Acceleration Index and offset index Substitute the weighted sum into the formula Calculate the final spatial sampling requirement coefficient. Among them, the weighting coefficient , and The system allows for flexible configuration of the importance of different factors based on the actual application scenario. Through the above process, this solution can transform the dynamic motion state of the mobile vehicle 1 and the relative positional change between the laser scanning device 2 and the target into a quantified spatial sampling demand coefficient. This coefficient was then used to calculate the desired point cloud density. This, in turn, affects the matching degree of the scanning state. The calculation. This dynamic, multi-factor comprehensive evaluation method allows the scanning state matching degree to more accurately reflect the actual sampling requirements, thereby providing a basis for the subsequent target laser scanning trigger frequency. The calculation and adjustment provide more precise input. Compared to methods that rely solely on a single or static parameter to assess sampling requirements, this solution, through a comprehensive consideration of speed, acceleration, and offset, enables the system to more intelligently and precisely adjust the laser scanning frequency, ensuring high-quality and efficient point cloud data under various driving conditions.

[0064] As a specific implementation method, during the inspection of guardrail expansion joints, a high-precision inertial navigation system (such as a GNSS / IMU integrated navigation module) can be configured on the mobile vehicle 1 to acquire the vehicle's speed and acceleration in real time. For example, this system can output the vehicle's three-dimensional speed and acceleration data at a frequency of 100Hz. Simultaneously, the laser scanning device 2 can be equipped with a built-in distance measurement function, or it can calculate the lateral offset between the scanning centerline of the laser scanning device 2 and the guardrail expansion joint target by processing the scanned point cloud data in real time. For example, by identifying the guardrail edge features and calculating its distance from the vehicle's centerline. Assuming the system's maximum designed detection speed is 100km / h and the maximum allowable offset is 1 meter, when the mobile vehicle 1 travels at 50km / h and the laser scanning device 2's offset from the target is 0.5 meters, the speed index... The offset exponent can be calculated to be 0.5. The value can be calculated to be 0.5. For the driving acceleration, assume that moving car 1 is currently moving at a speed of 0.5... The acceleration is accelerated, and the preset driving acceleration reference value is used. 1 At this time, the vehicle acceleration index According to the formula Perform the calculations. Finally, set the weighting coefficients. 0.4 For 0.3 and The calculated driving speed index is 0.3 (satisfying the condition that the sum is 1). Acceleration Index and offset index Substitute into the formula This allows us to obtain the spatial sampling requirement coefficient at the current moment. For example, if It is 0.5. It is 0.73. If it is 0.5, then The calculated value is 0.569. This calculated value... The value will serve as an important input for subsequent calculations of the desired point cloud density and the matching degree of the scanning state, thereby guiding the dynamic adjustment of the laser scanning trigger frequency.

[0065] Through the above technical solution, this application can accurately obtain the driving speed and acceleration of the mobile vehicle 1, as well as the offset of the laser scanning device 2 from the target, and convert them into unified, physically meaningful indices. Based on these indices, the spatial sampling requirement coefficient is calculated by weighted summation. This coefficient comprehensively and dynamically reflects the actual needs for point cloud sampling under the current detection environment. This method overcomes the problem of inaccurate assessment of sampling needs that may exist in traditional methods, and avoids undersampling or oversampling caused by changes in vehicle motion and target relative position. Therefore, this solution can ensure that the system maintains optimal point cloud data quality and detection efficiency under various complex driving conditions, significantly improving the adaptability and reliability of guardrail expansion joint detection.

[0066] This application further proposes the following procedure for the turning compensation coefficient:

[0067] Acquiring path curvature and yaw rate is crucial for assessing a vehicle's turning behavior during guardrail expansion joint inspection. Path curvature characterizes the degree of bending in the vehicle's path, with a higher value indicating a more curved path. Yaw rate reflects the angular velocity of the vehicle's rotation around its vertical axis, i.e., how quickly the vehicle turns. Accurate acquisition of path curvature and yaw rate is essential for evaluating a vehicle's turning state during guardrail expansion joint inspection. These values ​​can be obtained from the vehicle's own sensors (e.g., inertial measurement unit (IMU), GNSS receiver) or through post-processing calculations of the vehicle's trajectory. For example, path curvature can be estimated by the ratio of vehicle speed to yaw rate, or calculated by fitting vehicle position data; yaw rate can be directly output from the IMU sensor.

[0068] The absolute values ​​of path curvature and yaw rate are compared with the design maximum curvature and design maximum yaw rate, respectively. A min function is used to limit the ratio to an upper limit of 1, yielding the path curvature index and yaw rate index. The design maximum curvature and design maximum yaw rate are preset reference values ​​used for normalization. They represent the maximum turning degree and maximum turning speed that the vehicle might achieve, as considered during the system's design. These values ​​can be set according to the actual application scenario, vehicle type, and the requirements of the detection task. For example, for highway detection vehicles, the design maximum curvature and design maximum yaw rate can be set as the maximum safe turning parameters allowed at high speeds; for urban road detection vehicles, they can be set as the parameters for sharp turns common on urban roads. The min function limits the upper limit of the ratio to 1 to ensure that the normalized index value does not exceed 1. This means that even if the actual path curvature or yaw rate exceeds the design maximum value, the corresponding index value will be limited to 1, thus avoiding the introduction of excessive weights or unreasonable compensations in subsequent calculations. This approach ensures the effective range of the index and the stability of the calculation. The path curvature index and yaw rate index are the results of normalizing the path curvature and yaw rate, respectively, and their values ​​typically range from 0 to 1. These indices can uniformly measure the influence of different physical quantities on the turning compensation requirements. For example, when the path curvature or yaw rate reaches its design maximum value, the corresponding index is 1; when it is 0, the index is also 0.

[0069] The turning compensation coefficient is obtained by weighting the path curvature index and yaw rate index using their maximum and average values. The turning compensation coefficient is close to 0 when driving straight, and increases with either the path curvature index or the yaw rate index when turning. The compensation is stronger when both the path curvature index and the yaw rate index are high. The specific calculation method is as follows: substitute the path curvature index and the yaw rate index into the formula... Obtain the turning compensation coefficient , When going straight When turning, a higher curvature or yaw rate will trigger compensation; when both are high, the compensation is stronger. This is a weighting factor ranging from 0 to 1, used to balance the relative importance of the path curvature index and yaw rate index when calculating the turning compensation coefficient. When the value is close to 1, the system tends to compensate by taking the maximum of the two factors; that is, if either factor is higher, it will cause a larger compensation. When the value is close to 0, the system tends to compensate by averaging the two values; that is, both values ​​need to be relatively high to trigger a significant compensation. This can be achieved by adjusting... It can flexibly control the sensitivity of cornering compensation according to actual detection needs and vehicle motion characteristics. Specifically, values ​​can be assigned through preset strategies or dynamic algorithms; The path curvature index. This is the yaw rate index.

[0070] This application's solution obtains the vehicle's path curvature and yaw rate, two key parameters that comprehensively reflect the vehicle's turning state. To transform these physical quantities into unified indices usable for compensation calculations, the absolute values ​​of path curvature and yaw rate are first compared to their corresponding design maximum values. A min function is then used to limit the upper limit of these ratios to 1, resulting in the path curvature exponent and yaw rate exponent. This normalization process ensures the comparability and stability of different physical quantities in subsequent calculations. Subsequently, these two exponents are substituted into a weighted formula combining the maximum and average values, and a tradeoff factor is introduced. Adjustments were made, and the final turning compensation coefficient was calculated. This coefficient can sensitively reflect the degree of a vehicle's turning; when the vehicle is traveling straight, A value close to 0 indicates that no additional compensation is needed; however, when a vehicle turns, whether the curvature is high, the yaw rate is high, or both are high, it will lead to... This increases the desired point cloud density, thereby promoting higher density in subsequent scan state matching calculations. This mechanism allows the system to dynamically adjust the point cloud density requirement based on the actual turning conditions of the vehicle, thereby affecting the adjustment of the laser scanning trigger frequency. This ensures that high-quality guardrail expansion joint detection data can be obtained even during turning, effectively avoiding the problem of insufficient sampling caused by turning.

[0071] As a specific implementation method, during the inspection of guardrail expansion joints, the mobile vehicle 1 can be equipped with an inertial measurement unit (IMU) and a Global Navigation Satellite System (GNSS) receiver. The IMU can provide the vehicle's yaw rate in real time, while the GNSS data, combined with the vehicle's motion model, can be used to calculate the path curvature. For example, when the vehicle travels at a certain speed, the IMU measures the yaw rate, and the GNSS data calculates the path curvature. Based on the system's preset maximum design curvature and maximum design yaw rate, the path curvature index is first calculated. and yaw rate index Assuming a tradeoff factor If set to 0.7, then the turning compensation coefficient... According to the formula Perform calculations. For example, if It is 0.6 and If it is 0.8, then , Substituting into the formula, we can obtain It is 0.77. This is the calculated value. The value will then be used to calculate the scan state matching degree, thereby guiding the laser scanner 2 to adjust its trigger frequency.

[0072] Through the above technical solution, this application can accurately quantify the dynamic characteristics of a vehicle during turning and convert them into a turning compensation coefficient. The introduction of this coefficient allows the system to fully consider the additional sampling density requirements of vehicle turning when calculating the scanning state matching degree. When a vehicle enters a curve, the turning compensation coefficient increases according to changes in path curvature and yaw rate, thereby prompting the system to increase the desired point cloud density and correspondingly increase the laser scanning trigger frequency. This ensures that when a vehicle is turning, especially during sharp or high-speed turns, the laser scanning device 2 can collect data at a higher frequency, effectively avoiding the loss of detailed information about guardrail expansion joints or a decrease in detection accuracy due to insufficient sampling, and significantly improving the integrity and reliability of guardrail expansion joint detection data under complex road conditions.

[0073] like Figure 2 As shown, in some other embodiments, this application proposes a guardrail expansion joint detection device, including a mobile vehicle 1 and a laser scanner 2 installed on the mobile vehicle 1; a memory for storing a computer program; and a processor for implementing the steps of the guardrail expansion joint detection method described above when executing the computer program.

[0074] The core innovation of this embodiment lies in the integration of the mobile vehicle 1 and the laser scanning device 2, and the introduction of an adaptive frequency control mechanism based on multi-factor fusion analysis and health perception. This enables real-time adaptation to changes in vehicle motion status and detection scenario, while also taking into account system health and operational safety, thereby improving the reliability, accuracy, and adaptability of guardrail expansion joint detection.

[0075] Specifically, the mobile vehicle 1 serves as a detection platform, carrying a laser scanning device 2 that travels along the expansion joint of the guardrail. The computer program stored in the memory contains instructions for implementing the aforementioned guardrail expansion joint detection method. When the processor executes these instructions, it can calculate the spatial sampling requirement coefficient based on the mobile vehicle 1's speed and acceleration, and the offset of the laser scanning device 2 from the target; calculate the turning compensation coefficient based on the path curvature and yaw rate; calculate the scanning state matching degree based on the point cloud density and point cloud signal-to-noise ratio under the spatial sampling requirement coefficient and the turning compensation coefficient; calculate the system health limitation coefficient based on the laser temperature and system computing load rate; and calculate the target laser scanning trigger frequency based on the current laser scanning trigger frequency, the scanning state matching degree, and the system health limitation coefficient, and adjust the current laser scanning trigger frequency to the target laser scanning trigger frequency.

[0076] Through the above technical solution, the device can intelligently and dynamically adjust the laser scanning frequency to adapt to changes in vehicle movement and detection scenarios, while taking into account the system's health status. This effectively solves the technical problems of low efficiency and strong subjectivity in traditional detection methods, as well as the oversampling or undersampling of point cloud data, lack of equipment status perception and adaptive control in existing mobile laser scanning detection solutions under complex driving conditions.

[0077] In some embodiments described above in this application, a structural implementation of a guardrail expansion joint detection device is proposed. However, during its implementation, it is necessary to ensure the coordinated operation of the various components to achieve the adaptive frequency control function. To this end, this application further proposes specific steps for implementing the above method by having a processor execute a computer program stored in memory.

[0078] The solution proposed in this application constructs a complete adaptive detection system by integrating a mobile vehicle 1, a laser scanning device 2, a memory, and a processor. The mobile vehicle 1 provides a mobile platform, the laser scanning device 2 is responsible for collecting point cloud data, the memory stores the control program, and the processor dynamically calculates and adjusts the laser scanning trigger frequency based on the real-time collected vehicle status and system health data.

[0079] Through the above technical solution, the device can sense changes in motion status (such as acceleration, turning, and deviation) and system health status (such as laser temperature and computing load rate) in real time during vehicle operation, and intelligently adjust the scanning frequency accordingly. This ensures the quality and detection accuracy of point cloud data, avoids overload operation of the equipment, and extends the service life of the equipment.

[0080] The following is a concrete example: When the mobile vehicle 1 travels at a constant speed on a straight road, the spatial sampling demand coefficient and turning compensation coefficient are both low, the scan state matching degree is high, and the system health limit coefficient is also at a normal level. At this time, the laser scanning trigger frequency remains at the baseline value. When the mobile vehicle 1 enters a curve and accelerates, the turning compensation coefficient and spatial sampling demand coefficient increase. If the scan state matching degree decreases, the processor will calculate a higher target frequency to increase the sampling density. At the same time, if the laser temperature rises and the system health limit coefficient decreases, the processor will limit the upper limit of the frequency to prevent the equipment from overheating. This dynamic adjustment mechanism ensures that high-quality point cloud data can be obtained under various driving conditions, while protecting the equipment safety.

[0081] In summary, this embodiment achieves intelligent and adaptive detection of guardrail expansion joints through the organic combination of hardware devices and software algorithms, significantly improving detection efficiency, data quality, and system reliability.

[0082] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for detecting expansion joints in guardrails, characterized in that, include: Based on the vehicle's speed and acceleration, as well as the offset of the laser scanner from the target, the spatial sampling requirement coefficient is calculated. The turning compensation coefficient is calculated based on the path curvature and yaw rate. Based on the point cloud density and point cloud signal-to-noise ratio under the spatial sampling demand coefficient and the turning compensation coefficient, the scanning state matching degree is calculated and obtained; Based on the laser temperature and the system's calculated load rate, the system health limitation coefficient is calculated and obtained. Based on the current laser scanning trigger frequency, scanning status matching degree, and system health limitation coefficient, the target laser scanning trigger frequency is calculated and obtained, and the current laser scanning trigger frequency is adjusted to the target laser scanning trigger frequency.

2. The method for detecting expansion joints in guardrails according to claim 1, characterized in that, The process for calculating and obtaining the target laser scanning trigger frequency is as follows: Obtain the current laser scan trigger frequency, scan status matching degree, and system health limitation coefficient; Based on the deviation between the current scanning trigger frequency and the matching degree of the scanning state, a preliminary target frequency is calculated by proportional adjustment. That is, according to the difference between the current scanning state matching degree and the preset target matching degree, the current laser scanning trigger frequency is increased or decreased by a certain proportional coefficient to make the current laser scanning trigger frequency change in the direction of optimizing the matching degree. Based on the system health limit coefficient, a linear interpolation is performed between the minimum and maximum allowed trigger frequencies to obtain the upper limit of the frequency under the current system health limit coefficient. The higher the system health limit coefficient, the closer the allowed upper limit is to the maximum value; conversely, the lower the coefficient, the closer it is to the minimum value. The initial target frequency is compared with the minimum frequency and the upper limit of the frequency allowed by the system, and a reasonable value among the three is selected: first, ensure that it is not lower than the minimum frequency, then ensure that it does not exceed the upper limit of the frequency, so as to obtain the final target scanning trigger frequency.

3. The method for detecting expansion joints in guardrails according to claim 2, characterized in that, The process for calculating and obtaining the system health limitation coefficient is as follows: Obtain the current laser temperature and system load rate; The laser temperature is obtained by performing maximum-minimum normalization on the current laser temperature. The system computing load rate index is obtained by comparing the current system computing load rate with its maximum allowed value. The temperature index and load rate index are converted into corresponding health contribution factors, and a weighted product is used to obtain the system health limitation coefficient. The smaller the system health limitation coefficient, the less healthy the system is, and the more restrictive the scanning frequency needs to be.

4. The method for detecting expansion joints in guardrails according to claim 2, characterized in that, The process for calculating and obtaining the scanning state matching degree is as follows: Obtain the current spatial sampling demand coefficient, turning compensation coefficient, point cloud density, and point cloud signal-to-noise ratio; Based on the spatial sampling demand coefficient and the turning compensation coefficient, and combined with the acceptable minimum point cloud density, the expected point cloud density under the current working condition is calculated. The expected point cloud density increases with the increase of the spatial sampling demand coefficient and the turning compensation coefficient. Based on the ratio of the actual point cloud density to the desired point cloud density, the density matching degree is obtained by normalizing it using a logarithmic function. The density matching degree increases as the actual density increases relative to the desired density, but the growth rate gradually slows down. The signal-to-noise ratio (SNR) is compared to a reference value determined by the desired density and the minimum density. An exponential decay function is used to obtain the SNR matching degree. The higher the SNR, the closer the matching degree is to 1; conversely, it approaches 0. Substitute density matching degree and signal-to-noise ratio matching degree into the formula Get the matching degree of the scan status , , A higher value indicates a more ideal scanning frequency setting and a better match between data quality and detection requirements. For density matching degree, This represents the signal-to-noise ratio matching degree.

5. The method for detecting guardrail expansion joints according to claim 4, characterized in that, The process for calculating and obtaining the spatial sampling demand coefficient is as follows: Obtain the moving vehicle's speed and acceleration, as well as the offset of the laser scanner from the target; The driving speed of the mobile vehicle and the real-time offset of the laser scanner from the target are compared with the maximum detection speed designed by the system and the maximum offset allowed by the system, respectively, to obtain the driving speed index and the offset index. The driving acceleration is mapped to the driving acceleration exponent by a centrally symmetric S-shaped function. The driving acceleration exponent takes the middle value when the driving acceleration is zero, is greater than the middle value during acceleration, and is less than the middle value during deceleration, thus reflecting the need to increase the sampling frequency for acceleration and decrease the sampling frequency for deceleration. The driving speed index, driving acceleration index, and offset index are linearly weighted and summed according to preset weights to obtain the spatial sampling demand coefficient. The spatial sampling demand coefficient comprehensively reflects the vehicle's motion state's demand for sampling density, and the larger the value, the higher the demand.

6. The method for detecting expansion joints in guardrails according to claim 4, characterized in that, The process for obtaining the turning compensation coefficient is as follows: Obtain path curvature and yaw rate; The absolute values ​​of the path curvature and yaw rate are compared with the design maximum curvature and design maximum yaw rate, respectively. After the min function limits the upper limit of the ratio to 1, the path curvature index and yaw rate index are obtained. The turning compensation coefficient is obtained by weighting the path curvature index and the yaw rate index by the maximum and average values. The turning compensation coefficient is close to 0 when going straight, and increases with the increase of either the path curvature index or the yaw rate index when turning. The compensation is stronger when both the path curvature index and the yaw rate index are high.

7. A device for detecting expansion joints in guardrails, characterized in that, This includes the mobile vehicle and the laser scanning device installed on it; Memory, used to store computer programs; A processor, configured to, when executing a computer program, implement the steps of the guardrail expansion joint detection method as described in any one of claims 1-6.

Images