Anti-collision identification early warning method and system for tire crane cart

Abstract

The invention relates to the technical field of harbor machinery safety protection, in particular to an anti-collision recognition early warning method and system for a tire crane cart. Performing calibration correction and calibration combination on laser radar information and monitoring image information in the information to obtain a combined information set; performing plane fitting and plane point projection on the obstacle according to the combined information set and the tire crane cart operation characteristics to obtain obstacle contour map information in the tire crane cart operation process; analyzing the contour map information of the obstacle to determine the center point coordinate of the obstacle, and analyzing the three-dimensional center point coordinate of the obstacle in combination with the joint information set; and constructing a collision risk dynamic evaluation system, and predicting the collision risk of the tire crane cart by combining the operation condition of the tire crane cart, the three-dimensional center point coordinates and the combined information set. The position and the distance between the tire crane cart and the obstacle are dynamically evaluated, anti-collision early warning is achieved, and operation safety is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of port machinery safety protection, in particular to a collision identification and early warning method and system for a rubber-tired crane trolley. BACKGROUND

[0002] In the process of port container loading and unloading operation, the rubber-tired crane trolley is one of the key equipment, which has the advantages of strong mobility and wide operation range. However, the port operation environment is complex, and the rubber-tired crane trolley is prone to collide with other equipment, containers, personnel and other obstacles during operation. Once a collision occurs, it will cause equipment damage, cargo damage, and even casualties and other consequences.

[0003] The existing collision avoidance technology for the rubber-tired crane trolley has many deficiencies: first, a single sensor or data acquisition device is used for environment sensing and data detection, which makes the data easily affected by external environmental factors, and thus leads to poor information quality; second, the target recognition algorithm for obstacles has insufficient accuracy, and it is difficult to effectively and accurately identify the approximate shape and center position of the obstacles; third, the collision risk assessment model is relatively simple, and does not fully consider the effect of the equipment motion state on the risk assessment result, so it cannot accurately predict the dynamic collision risk between the rubber-tired crane trolley and the obstacles. The existing technology cannot take obstacle avoidance measures in time when the collision risk occurs.

[0004] Therefore, a new collision identification and early warning method and system for the rubber-tired crane trolley are needed, which can realize risk prediction and intelligent avoidance between the rubber-tired crane trolley and obstacles, improve the safety of the rubber-tired crane trolley operation, and reduce the occurrence of collision accidents. SUMMARY

[0005] In view of the deficiencies of the existing method and the needs of practical application, in order to guarantee the information data quality in the process of loading and unloading operation, effectively determine the obstacle contour shape and center position, realize the accurate prediction of the collision risk between the tire crane car and the obstacle, and improve the operation safety of the tire crane car. On the one hand, the present application provides a kind of anti-collision identification early warning method for tire crane car, and its method includes: obtaining the original data information in the operation process of tire crane car, calibrating and correcting laser radar information and monitoring image information in original data information and joint calibration, to obtain the joint information set in the operation process of tire crane car;According to the running characteristics of tire crane car and the joint information set, the plane fitting analysis and plane point projection processing of obstacle are carried out, to obtain the obstacle contour graph information in the operation process of tire crane car;According to the center point coordinates of the obstacle determined by the obstacle contour graph information and the joint information set, the three-dimensional center point coordinates of the obstacle are analyzed based on the center point coordinates and the joint information set;Collision risk dynamic evaluation system is constructed, and the collision risk of tire crane car is predicted by combining the operation condition of tire crane car, the three-dimensional center point coordinates, the joint information set and the collision risk dynamic evaluation system, to realize the dynamic identification and anti-collision early warning of tire crane car.

[0006] The present application fuses different data, can give full play to the advantages of each data source, realizes information complementation, and can more completely perceive the surrounding environment through joint information set;Obstacle contour information and three-dimensional coordinate information can more accurately describe the position of obstacle in space, provide decision basis for tire crane car obstacle avoidance;Collision risk dynamic evaluation system can accurately predict the collision risk result, which is beneficial to guarantee the safe operation of port.

[0007] Optionally, the calibration and correction and joint calibration of laser radar information and monitoring image information in original data information to obtain the joint information set in the operation process of tire crane car includes: setting the scanning starting point of laser radar information based on the information acquisition method of laser radar device;According to the installation position of laser radar device and the scanning starting point, the laser radar information acquisition path is divided into a plurality of scanning line segments;According to the plurality of scanning line segments and the installation position of laser radar device, the line segment slope value between adjacent laser radar devices is analyzed.The present application can more comprehensively understand the spatial structure of surrounding environment according to the analysis of line segment slope value between adjacent laser radar devices based on the plurality of scanning line segments and the installation position of laser radar device, to provide more accurate information for subsequent data fusion and obstacle identification.

[0008] Optionally, the calibration correction and joint calibration of the laser radar information and the monitoring image information in the original data information to obtain the joint information set in the operation process of the tire crane car comprises: analyzing the adjacent slope difference between adjacent laser radar devices according to the line segment slope value; setting a dynamic slope threshold and a neighborhood radius threshold according to the installation position of the laser radar device; and correcting the laser radar information in combination with the adjacent slope difference, the dynamic slope threshold and the neighborhood radius threshold to obtain a laser radar data set. The present application can adjust the deviated data in a targeted manner by correcting the laser radar information in combination with the adjacent slope difference, the dynamic slope threshold and the neighborhood radius threshold, so that the information is more in line with the actual situation, effectively eliminates the noise and errors in the data, and improves the accuracy and reliability of the laser radar data.

[0009] Optionally, the calibration correction and joint calibration of the laser radar information and the monitoring image information in the original data information to obtain the joint information set in the operation process of the tire crane car comprises: introducing a standard homogeneous coordinate, establishing a radar-camera coordinate transformation model based on the standard homogeneous coordinate; analyzing and transforming the laser radar data set and the monitoring image information through the radar-camera coordinate transformation model to obtain feature point three-dimensional coordinates in a camera coordinate system; establishing a camera-image coordinate transformation model according to the pinhole camera projection mechanism; and analyzing and converting the feature point three-dimensional coordinates in the camera coordinate system by using the camera-image coordinate transformation model to obtain feature point coordinates in an image coordinate system.

[0010] The present application can more accurately determine the position of the obstacle in the image and its three-dimensional coordinates in the actual space through the coordinate transformation model. In the operation process of the tire crane car, accurately identifying the position and shape of the obstacle plays an important role in avoiding collision.

[0011] Optionally, the calibration correction and joint calibration of the laser radar information and the monitoring image information in the original data information to obtain the joint information set in the operation process of the tire crane car comprises: establishing a pixel coordinate transformation model in combination with the image distortion condition and the resolution correction principle; calibrating and analyzing the feature point coordinates in the image coordinate system by using the pixel coordinate transformation model to obtain the pixel coordinates of the feature points in the monitoring image; and obtaining the joint information set in the operation process of the tire crane car in combination with the pixel coordinates, the laser radar data set and the monitoring image information.

[0012] The pixel coordinate transformation and calibration analysis process of the present application provides an effective data processing method, which can process and fuse data through a model, and reduce data fluctuations caused by environmental changes or equipment errors.

[0013] Optionally, the plane fitting analysis and plane point projection processing of the obstacle according to the tire crane car operation characteristics and the joint information set to obtain the obstacle contour graph information in the tire crane car operation process comprises: randomly extracting reference points based on the tire crane car operation characteristics and the joint information set; obtaining a plane normal vector according to the covariance matrix and the reference points; establishing a plane equation expression and a point-plane distance analysis expression according to the plane normal vector; performing plane fitting analysis on the obstacle through the plane equation expression, the point-plane distance analysis expression and the joint information set, and obtaining an obstacle plane fitting result. The plane fitting result of the obstacle is obtained through plane fitting, and the point is projected onto the corresponding plane through plane point projection processing, so that the contour graph information of the obstacle is obtained. Based on this, the complete contour of the obstacle in the three-dimensional space can be obtained, and more comprehensive environmental information for the operation of the tire crane car is provided.

[0014] Optionally, the plane fitting analysis and plane point projection processing of the obstacle according to the tire crane car operation characteristics and the joint information set to obtain the obstacle contour graph information in the tire crane car operation process comprises: setting a calibration board point cloud; obtaining an obstacle plane projection analysis function based on a space projection method, the joint information set and the calibration board point cloud; obtaining obstacle plane projection information according to the obstacle plane projection analysis function; introducing an image coordinate axis vector; obtaining obstacle pixel coordinate results according to the image coordinate axis vector and the obstacle plane projection information; and integrating the obstacle pixel coordinate results to obtain the obstacle contour graph information in the tire crane car operation process. The laser radar information and the monitoring image information are fused, so that the obstacle plane projection information is more comprehensive and accurate, which is conducive to providing clear and intuitive obstacle contour information subsequently.

[0015] Optionally, the center point coordinate of the obstacle is determined according to the obstacle contour graph information and the joint information set, and the three-dimensional center point coordinate of the obstacle is analyzed based on the center point coordinate and the joint information set, which comprises: obtaining edge adjustment parameters based on the obstacle contour graph information and the joint information set; deriving and establishing a center point coordinate analysis function according to an image coordinate axis vector, the edge adjustment parameters and the obstacle contour graph information; determining the center point coordinate of the obstacle through the center point coordinate analysis function; constructing a three-dimensional coordinate equation set according to a plane equation expression and the center point coordinate of the obstacle; and obtaining the three-dimensional center point coordinate of the obstacle based on the three-dimensional coordinate equation set.

[0016] The plane equation expression and the three-dimensional coordinate equation set of the application can accurately describe the spatial position and attitude of the plane where the obstacle is located, and the center point coordinate is combined with the spatial information, which improves the positioning accuracy of the obstacle in the three-dimensional space and provides protection for the safe operation of the tire crane car.

[0017] Optionally, the construction collision risk dynamic evaluation system combines the tire crane car operation condition, the three-dimensional center point coordinates, the joint information set and the collision risk dynamic evaluation system to predict the collision risk of the tire crane car, so as to realize the dynamic identification and anti-collision warning of the tire crane car, comprising: establishing a relative distance prediction function and a relative speed prediction function in the collision risk dynamic evaluation system; combining the tire crane car operation condition, the three-dimensional center point coordinates, the joint information set, the relative distance prediction function and the relative speed prediction function to analyze the relative distance and the relative speed between the tire crane car and the obstacle; establishing a collision time analysis equation under the motion state in the collision risk dynamic evaluation system according to the tire crane car motion state; predicting the collision time between the tire crane car and the obstacle by using the collision time analysis equation under the motion state; introducing the tire crane car steering angle, setting and distributing the acceleration weight and the steering angle weight based on the tire crane car motion state and the tire crane car steering angle; adjusting and optimizing the collision time by combining the acceleration weight and the steering angle weight, to obtain the dynamic collision time between the tire crane car and the obstacle; and predicting the collision risk between the tire crane car and the obstacle by combining the relative distance, the relative speed and the dynamic collision time, to realize the dynamic identification and anti-collision warning of the tire crane car.

[0018] The present application can predict the collision risk by combining the relative distance, the relative speed and the dynamic collision time, can comprehensively obtain the running conditions between the obstacle and the tire crane car, and can more accurately predict the risk of collision with the obstacle under different motion states, further ensuring the operation safety of the tire crane car.

[0019] In order to efficiently execute the tire crane car anti-collision identification and warning method provided by the present application, the present application further provides a tire crane car anti-collision identification and warning system, which comprises an input device, a processor, an output device and a memory, wherein the input device, the processor, the output device and the memory are connected to each other, the memory comprises the tire crane car anti-collision identification and warning method as described in the first aspect of the present application, the memory is used for storing a computer program, the computer program comprises program instructions, and the processor is configured to call the program instructions. The tire crane car anti-collision identification and warning system provided by the present application has compact structure, strong applicability and greatly improved operation efficiency. BRIEF DESCRIPTION OF DRAWINGS

[0020] Figure 1 The flowchart of the tire crane car anti-collision identification and warning method of the present application; Figure 2The structural diagram of the anti-collision identification and early warning system for the tire crane trolley. DETAILED DESCRIPTION

[0021] Specific embodiments of the present application will now be described in detail with reference to the figures. Like numbers in different figures represent the same or similar elements. The implementation of the application will be described with reference to the following examples. Other elements, features and aspects of the application are apparent from the following description.

[0022] Throughout this specification, reference has been made to "one embodiment", "an embodiment", "one example", or "an example" meaning that a particular feature, structure, or characteristic described in connection with the embodiment or example is included in at least one embodiment of the application. Therefore, the appearance of the phrases "in one embodiment", "in an embodiment", "one example", or "an example" in various places throughout this specification are not necessarily all referring to the same embodiment or example. Furthermore, the particular features, structures, or characteristics can be combined in any suitable manner in one or more embodiments or examples. In addition, the skilled person will appreciate that the drawings provided herein are for illustrative purposes only and are not necessarily drawn to scale.

[0023] Reference will now be made to Figure 1 In order to improve the information accuracy of the loading and unloading process, effectively determine the spatial distribution and center position of the obstacle, realize the accurate prediction of the collision risk between the tire crane trolley and the obstacle, and improve the operation safety level of the tire crane trolley, the present application provides an anti-collision identification and early warning method for a tire crane trolley, which comprises the following steps: S1, obtaining the original data information in the operation process of the tire crane trolley, calibrating and correcting the laser radar information and monitoring image information in the original data information, and jointly calibrating to obtain a joint information set in the operation process of the tire crane trolley, the specific setting steps and implementation contents are as follows: First, the original data information in the operation process of the tire crane trolley is obtained.

[0024] In order to effectively obtain the relevant data information of the tire crane during the operation process, various types of sensors are installed on the tire crane, and a data acquisition scheme is designed, which specifically relates to laser radar, millimeter wave radar, ultrasonic sensor, camera and other equipment. Among them, the laser radar is installed on the top and around the tire crane, which is mainly used to collect three-dimensional point cloud data of the surrounding environment; the millimeter wave radar is installed at the front and rear of the vehicle, which is mainly responsible for detecting long-distance obstacles; the ultrasonic sensor is installed at the bottom and both sides of the vehicle body, which is used for detecting short-distance obstacles; the camera is installed at the front, rear and both sides of the vehicle body, which is used to obtain visual images of the surrounding environment. The above-mentioned sensors will collect the environmental information around the tire crane in real time, and transmit the collected data to the data storage system.

[0025] In this embodiment, the obstacles can be different monitoring targets during the operation process of the tire crane, and the obstacles can be carrying vehicles, workers, moving goods and the like during the operation process. Real-time master of the position and state of the obstacles (different monitoring targets) can ensure the safe operation of the tire crane, improve the operation efficiency and maintain the order of the whole operation site.

[0026] Meanwhile, when installing the sensors, the data acquisition scheme should be followed. The laser radar selects a high-precision and high-resolution device to ensure that its scanning angle can cover the range around the tire crane The detection distance of the millimeter wave radar should be more than 200m; the detection distance of the ultrasonic sensor can be selected within to meet the short-distance detection requirements of the tire crane; the camera adopts a high-definition camera with wide-angle view and good night vision capability. The above-mentioned sensors and devices can be connected to the data storage system through wired or wireless mode, and based on this, the real-time transmission of the tire crane related data is realized, so as to quickly and effectively obtain the original data information of the tire crane during the operation process.

[0027] In actual application scenarios, due to the restriction of external environmental factors and data acquisition conditions, information noise points and outliers will be mixed in the original collected data information. The above-mentioned abnormal data points will affect the data quality. In order to ensure the accuracy and effectiveness of the data information, data registration and information segmentation work need to be carried out, and the original data information needs to be adjusted and optimized, so as to obtain the laser radar data set of the tire crane during the operation process.

[0028] Considering that the laser radar is installed on the top and around the tire crane trolley, the embodiment distinguishes different data collection points based on the slope difference value information. On this basis, a data adaptive adjustment method based on the slope value is further proposed, which can adjust the data according to the characteristics and variation law of the data itself, so that the data information is more in line with the actual situation, which is helpful to efficiently and accurately obtain the laser radar data set of the tire crane trolley during operation. The specific implementation steps are as follows: The first step is to set the scanning starting point of the laser radar information based on the information collection method of the laser radar device. That is, the first point on each scanning line is determined as the starting point, which is used as the reference for subsequent information collection and processing.

[0029] The second step is to divide the laser radar information collection path into multiple scanning line segments according to the installation position of the laser radar device and the scanning starting point. Based on the installation position of the laser radar device and the set scanning starting point, the information collection path of the laser radar is divided, and all points in the scanning line are processed in turn, so that each scanning line is divided into multiple scanning line segments, laying a foundation for the relationship analysis between adjacent line segments.

[0030] The third step is to analyze the line segment slope value between adjacent laser radar devices according to the multiple scanning line segments and the installation position of the laser radar device.

[0031] In the embodiment, the slope and slope are calculated based on the scanning line method. For a small line segment composed of two adjacent points, the calculation formula of the slope value satisfies the following relationship: Wherein, represents the slope value of the small line segment between adjacent laser radar installation points, represents the three-dimensional coordinates of different laser radar installation positions, represents the number of different laser radars.

[0032] Wherein is the point coordinate on the continuous scanning line. While processing the laser radar information, the distance data collected by the millimeter wave radar and ultrasonic sensor are calibrated and corrected, which helps to eliminate sensor errors and ensure the accuracy and consistency of various sensor data.

[0033] The fourth step is to analyze the adjacent slope difference between adjacent laser radar devices according to the line segment slope value.

[0034] The point cloud filtering algorithm based on the slope value calculates and analyzes the slope difference value, which helps to separate the monitoring points on the top and around the tire crane trolley, and provides technical support for subsequent data processing and key information extraction. In the embodiment, the slope difference value is mainly used to detect the distribution relationship of adjacent points in the scanning line profile.

[0035] After obtaining the above slope value , the adjacent slope difference between adjacent laser radar devices is further calculated , and the calculation formula is as follows: , wherein, represents the adjacent slope difference between adjacent laser radar devices, represents the small line segment slope value between the laser radar installation point and the laser radar installation point , represents the small line segment slope value between the laser radar installation point and the laser radar installation point , and represents the small line segment slope value between the laser radar installation point and the laser radar installation point , and represents the number of different laser radars.

[0036] The fifth step is to set a dynamic slope threshold and a neighborhood radius threshold according to the installation position of the laser radar device. According to the installation position of the laser radar device, a dynamic slope threshold and a neighborhood radius threshold are set, and adaptive slope filtering operation and information optimization are performed using the two thresholds, that is, adaptive adjustment is performed on the original data information according to the actual situation to adapt to different monitoring scenes and data characteristics.

[0037] The sixth step is to calibrate and correct the laser radar information in combination with the adjacent slope difference, the dynamic slope threshold and the neighborhood radius threshold to obtain a laser radar data set.

[0038] After the adjacent slope difference between adjacent laser radar devices is obtained by the above formula, slope value filtering operation is performed on the original data information and the original point cloud during the operation of the tire-mounted crane. If and the number of points in the neighborhood , the point is marked as a boundary point, and in the embodiment, the neighborhood radius threshold is introduced as an adjustable parameter, which can make the data adaptive adjustment method based on the slope value better adapt to the data processing needs in different scenes.

[0039] In the above process, the neighborhood radius size and the slope threshold need to be reasonably selected according to the shape of the tire-mounted crane and the actual environmental conditions to effectively separate the monitoring points on the top and around the tire-mounted crane. In an alternative embodiment, for shapes with a larger slope threshold, a smaller neighborhood radius threshold and a larger dynamic slope threshold are set; for other shapes, a smaller dynamic slope threshold and a larger neighborhood radius threshold are set. Under the condition that the remaining parameters remain unchanged, adaptive adjustment of the dynamic slope threshold and the neighborhood radius threshold can effectively improve the accuracy of the tire-mounted crane operation information.

[0040] After the above processing and adjustment, the laser radar data set in the operation process of the tire crane trolley can be effectively obtained, providing an information basis for subsequent data analysis and tire crane trolley collision identification.

[0041] After completing the calibration and correction of the laser radar data set, the laser radar data set and the monitoring image information in the operation process of the tire crane trolley need to be calibrated and combined to obtain the combined information set in the operation process of the tire crane trolley.

[0042] In the process of collision identification and early warning of the tire crane trolley, the fusion of camera information and laser radar data can effectively improve the accuracy of collision prediction results. Since the laser radar data set and the monitoring image information in the operation process of the tire crane trolley come from different devices, direct use exists mismatching problems, and the original data has noise and sequence disorder problems, etc. Therefore, the monitoring image information collected by the camera device needs to be denoised, enhanced and inspected, and then the calibration and combined analysis of the laser radar data set and the monitoring image information is carried out. The specific process and principle are as follows: The calibration and combined analysis of the laser radar data set and the monitoring image information belongs to fusion technology, and the information calibration result directly determines the actual effect of subsequent collision prediction, identification and positioning. In the process of collision identification and early warning of the tire crane trolley, the camera and the laser radar can quickly detect the position and actual condition of the spreader, container or other target objects. In order to realize the effective combination of data collected by different devices, calibration targets need to be pre-set on the tire crane trolley according to the actual use situation on site, and then subsequent calibration and combined analysis is carried out based on this, and then different data calibration and combined analysis is continued.

[0043] The first step is to introduce standard homogeneous coordinates, and to establish a radar-camera coordinate transformation model based on standard homogeneous coordinates.

[0044] The introduction of standard homogeneous coordinates is conducive to subsequent matrix operations. The rotation matrix is set to satisfy the following relationship: wherein, R represents the rotation matrix, Ri and Rj represent the elements of the rotation matrix.

[0045] In the equation, Ri and Rj are the elements of the rotation matrix, which describe the rotation relationship between the laser radar coordinate system and the camera coordinate system, and can reflect the relative rotation angle and direction of different coordinate systems in space.

[0046] The translation vector is set to satisfy the following relationship: wherein, T represents the translation vector, 、 、 respectively represent the translation distance in the axis direction.

[0047] 、 、 Also refers to the component of the translation vector, the translation distance of the origin of the laser radar coordinate system relative to the origin of the camera coordinate system in the axis direction.

[0048] Based on the above standard homogeneous coordinates, a radar-camera coordinate transformation model is further established, and satisfies the following relationship; wherein, represents the output coordinates of the radar coordinate system transformed into the camera coordinate coordinate system, represents a rotation matrix, represents a translation vector, represents the coordinates of the feature point in the laser radar.

[0049] The output coordinate information of the radar coordinate system after being transformed into the camera coordinate system, i.e. the three-dimensional coordinates of the feature point in the camera coordinate system, can be quickly obtained by analyzing and transforming the laser radar data set and the monitoring image information through the radar-camera coordinate transformation model.

[0050] The second step is to establish a camera-image coordinate transformation model according to the pinhole camera projection mechanism.

[0051] The camera-image coordinate transformation model is constructed according to the pinhole camera model projection mechanism, and satisfies the following relationship; wherein, represents the pixel coordinates of the feature point in the monitoring image, represents the focal length of the camera, represents the output coordinates of the radar coordinate system transformed into the camera coordinate coordinate system.

[0052] The camera focal length is an important parameter inside the camera, which determines the monitoring image magnification and imaging clarity. The feature point coordinates in the image coordinate system can be effectively obtained by analyzing and converting the feature point three-dimensional coordinates in the camera coordinate system using the camera-image coordinate transformation model.

[0053] The third step is to establish a pixel coordinate transformation model combined with the image distortion and resolution correction principle.

[0054] A pixel coordinate transformation model is set in combination with the image distortion condition and the resolution correction principle, and meets the following relationship: wherein, represents the pixel coordinates of the feature point in the monitoring image, represents the pixel size in the direction, represents the pixel size in the direction, represents the pixel size in the direction, represents the pixel size in the direction, represents the main point offset of the image coordinate system origin in the pixel coordinate system axis direction, represents the main point offset of the image coordinate system origin in the pixel coordinate system axis direction, represents the radial and tangential distortion terms.

[0055] The pixel coordinate system is the most commonly used coordinate system in image processing, which is helpful for accurately describing the position of the feature point in the image; the main point offset of the image coordinate system origin in the pixel coordinate system axis direction, that is, the pixel number of the pixel coordinate system origin relative to the image coordinate system origin in the axis direction; similarly, the main point offset of the image coordinate system origin in the pixel coordinate system axis direction, that is, the pixel number of the pixel coordinate system origin relative to the image coordinate system origin in the axis direction.

[0056] The pixel coordinate transformation model is used to calibrate and analyze the feature point coordinates in the image coordinate system, so as to obtain the pixel coordinates of the feature point in the monitoring image.

[0057] In the fourth step, the joint information set of the tire crane during operation is obtained in combination with the pixel coordinates, the laser radar data set and the monitoring image information.

[0058] Through the coordinate transformation and parameter calculation of the above three steps, the coordinate joint calibration of the laser radar and the camera can be effectively realized, the feature point data is transformed into the monitoring image, and the calibration fusion between different sensor data is realized. In combination with the pixel coordinates, the laser radar data set and the monitoring image information, the joint information set of the tire crane during operation can be obtained, which provides more accurate and comprehensive data support for the anti-collision identification and early warning of the tire crane.

[0059] The laser radar data set and the monitoring image information can be optimized and fused during the operation of the tire crane, the coordinate transformation and parameter calculation are completed by using the above coordinate transformation model, which is helpful for subsequent effective monitoring and marking analysis of the position and state change of the obstacle, and then the adaptability and practicality of the anti-collision identification and warning method for the tire crane can be improved.

[0060] S2, according to the running characteristics of the tire crane and the joint information set, the plane fitting analysis and plane point projection processing of the obstacle are carried out to obtain the obstacle contour information in the operation process of the tire crane, and the specific steps and implementation contents are as follows: The first step is to extract the obstacle calibration plate data and the contour preliminary extraction. According to the running characteristics of the tire crane, the calibration plate data of the obstacle is extracted from the joint information set, and the related data is helpful for subsequent determination of the plane equation where the calibration plate is located.

[0061] The laser radar data of the obstacle in the joint information set is projected onto a two-dimensional plane, and the two-dimensional plane points are converted into corresponding obstacle images. According to the energy value of each point in the monitoring image, a binary operation is performed, which is helpful for accurately extracting the obstacle contour data information. Through the above binary operation, the obstacle in the monitoring image can be separated from the environment background, and the obstacle contour feature is further highlighted.

[0062] The second step is to select reference points. Based on the running characteristics of the tire crane, reference points are randomly selected from the top and around the tire crane to form a set , wherein , represents the number of reference points, and the random selection of reference points can more comprehensively reflect the spatial characteristics around the tire crane, and improve the accuracy of the plane fitting result.

[0063] The third step is to calculate the plane normal vector. According to the covariance matrix and the reference points, the plane normal vector can be obtained. The plane normal vector is set as , that is, the eigenvector corresponding to the maximum eigenvalue of the covariance matrix composed of the set .

[0064] The calculation formula of the covariance matrix is as follows: , wherein represents the calculation formula of the covariance matrix, represents the number of reference points in the set, represents any one reference point in the set, represents the center point of the set, represents the transpose vector of

[0065] For column vectors Perform a transpose operation to obtain its transpose vector. The transpose operation swaps the rows and columns of a column vector, allowing you to... Transform the column vector into The row vector.

[0066] in The central point of the set satisfies the following relationship: For the above covariance matrix Eigenvalue decomposition yields the eigenvector corresponding to the largest eigenvalue, which can then be used as the plane normal vector. .

[0067] The fourth step established the plane equation expression and the point-plane distance analysis expression based on the plane normal vector.

[0068] Further settings The distance from the origin of the coordinate system to the plane can be represented by a set. center point It is calculated by dot product with the normal vector.

[0069] The distance from the origin of the coordinate system to the plane The following relationship must be satisfied: Based on plane normal vector The distances from the reference point and the origin of the coordinate system to the plane are used to fit the obstacle plane, and the plane equation expression is established to satisfy the following relationship; in, Represents the plane normal vector, i.e., the set The eigenvector corresponding to the largest eigenvalue of the covariance matrix. This represents the distance from the origin of the coordinate system to the plane. Represents three-dimensional coordinates.

[0070] Therefore, any point can be calculated. Distance to the plane The point-to-surface distance analysis expression is: in, Represent any point Distance to the plane, Represents the three-dimensional coordinates of different lidar installation locations. This indicates the serial number of different lidar units. This represents the plane normal vector.

[0071] The fifth step is to perform plane fitting analysis on the obstacle by using the plane equation expression, the point-to-plane distance analysis expression and the joint information set, and to obtain the obstacle plane fitting result.

[0072] A distance threshold is further set If the distance of any one point to the plane is less than the distance threshold , the point is an inlier, otherwise, the point is an outlier. By setting the threshold, the points meeting the plane fitting condition can be screened, and the accuracy of the obstacle plane fitting result is improved.

[0073] The plane equation expression, the point-to-plane distance analysis expression and the joint information set are used to perform plane fitting analysis on the obstacle. In the embodiment, the points are selected from the two-dimensional plane of the obstacle according to the energy values of the laser points of the obstacle, and the plane fitting is performed by using the random sample consensus algorithm. The random sample consensus algorithm can effectively exclude the interference of outliers, and ensure that the obtained obstacle contour data information is accurate and effective. By using the above expressions and threshold parameters, the obstacle plane fitting result can be obtained, which provides a reference basis for subsequent obstacle projection and positioning analysis.

[0074] The sixth step is to set the calibration plate point cloud and related parameter information.

[0075] According to the above obstacle plane fitting result, the coordinate information of any point on the plane can be obtained. Based on this, the coordinate of any point on the plane is set as , the calibration plate point cloud is set as , is the projection point on the plane, and the obstacle plane normal vector is . The parameters are provided for subsequent projection calculation.

[0076] The seventh step is to obtain the obstacle plane projection analysis function based on the spatial projection method, the joint information set and the calibration plate point cloud.

[0077] Since the three-dimensional space points of the obstacle are not completely in the same plane, the three-dimensional space points of the obstacle need to be projected into two-dimensional coordinate points. Therefore, the obstacle plane projection analysis function is established as follows: wherein, represents the projection point of on the plane, represents the coordinate of any point on the obstacle fitting plane, represents the calibration plate point cloud, represents the obstacle plane normal vector, represents ​​the module of

[0078] According to the obstacle plane projection information calculated by the above obstacle plane projection analysis function, three-dimensional space information can be converted into two-dimensional plane information, which is convenient for subsequent collision prediction analysis between the tire crane and the obstacle.

[0079] The eighth step introduces the image coordinate axis vector and related parameter information.

[0080] The vector related to the image coordinate axis is set, and two vectors are and respectively, and and are the modules of vectors and respectively; and are the components of and directions respectively, and the related parameters are mainly used for subsequent pixel coordinate calculation and analysis.

[0081] The ninth step obtains the obstacle pixel coordinate result according to the image coordinate axis vector and the obstacle plane projection information.

[0082] Based on the projection analysis result of any point on the obstacle fitting plane, the pixel coordinates of any point on the obstacle fitting plane can be quickly obtained. Based on the projection analysis result of any point on the obstacle fitting plane, a pixel coordinate analysis function of the obstacle fitting plane is established, and the following relationship is satisfied: wherein, represents the pixel coordinate analysis function, represents the pixel coordinates of any point on the obstacle fitting plane, represents the coordinates of any point on the obstacle fitting plane, represents the point cloud of the calibration board, represents a vector related to the image coordinate axis, represents another vector related to the image coordinate axis, represents the module of , represents the module of , represents the component of direction, represents the component of direction. According to the above pixel coordinate analysis function, the pixel coordinates of any point on the obstacle fitting plane are obtained, and the conversion between the plane projection information and the image pixel information is realized.

[0083] The tenth step integrates the obstacle pixel coordinate results to obtain the obstacle contour graph information in the tire crane trolley operation process.

[0084] The energy value of the laser scanning point is obtained according to the joint information set in the tire crane trolley operation process, and the pixel coordinates of any point on the obstacle fitting plane are combined The obstacle image is subjected to binaryzation processing, and the embodiment adopts an adaptive threshold binaryzation method, which automatically adjusts the threshold according to the gray scale distribution of the local region of the obstacle image, so as to better adapt to the obstacle image under different light conditions. The above adaptive threshold binaryzation can improve the binaryzation quality of the obstacle image, so that the extraction and analysis results of the obstacle contour information are more accurate.

[0085] The obstacle contour graph information in the tire crane trolley operation process is finally obtained by integrating the obstacle pixel coordinate results through the above implementation steps, which provides accurate and comprehensive data support for the anti-collision recognition and early warning of the tire crane trolley. The above steps can systematically and accurately obtain the obstacle contour graph information in the tire crane trolley operation process, which is beneficial to improve the operation safety and reliability of the tire crane trolley.

[0086] S3, according to the obstacle contour graph information and the joint information set, the center point coordinates of the obstacle are determined, and based on the center point coordinates and the joint information set, the three-dimensional center point coordinates of the obstacle are analyzed, and the specific implementation contents are as follows: The first step obtains the edge fitting adjustment parameters based on the obstacle contour graph information and the joint information set. In the embodiment, the adjustment parameters related to the edge fitting are determined, which are respectively represented by , wherein corresponds to the edge fitting adjustment parameter of the axis, corresponds to the edge fitting adjustment parameter of the axis.

[0087] The second step derives and establishes the center point coordinate analysis function according to the image coordinate axis vector, the edge fitting adjustment parameter and the obstacle contour graph information.

[0088] In the embodiment, the edge extraction operation of the obstacle is mainly performed according to the image coordinate axis vector, the edge fitting adjustment parameter and the obstacle contour graph information, and then the center of the ellipse is obtained by fitting the edge, and the center point coordinate analysis function is derived and established, and the function expression is as follows: wherein, the center point coordinate analysis function is represented by the center pixel coordinates of the obstacle fitting plane are respectively represented by the pixel coordinates of any point on the obstacle fitting plane are respectively represented by the pixel coordinates of any point on the obstacle fitting plane are respectively represented by axis and adjustment parameters related to axis edge fitting, denotes a vector related to the image coordinate axis, denotes another vector related to the image coordinate axis, denotes the modulus of the modulus of the modulus of the modulus of the component in the direction, denotes the component in the direction.

[0089] By the above center point coordinate analysis function, the center point coordinate of the obstacle can be determined.

[0090] In the third step, a three-dimensional coordinate equation set is constructed according to the plane equation expression and the center point coordinate of the obstacle.

[0091] Substituting the obtained center point into the plane equation expression, the three-dimensional center point coordinate of the obstacle can be calculated, and based on the embodiment, the plane equation expression is as follows: Let the three-dimensional center point coordinate be Since the center point satisfies the above plane equation expression on the plane, and the three-dimensional coordinate equation set can be established according to the mapping relationship between the pixel coordinate and the actual coordinate .

[0092] At the same time, considering that there is a scaling ratio and an offset between the image coordinate system and the actual coordinate system, and the height of the center point of the obstacle from the ground plane is known , the following three-dimensional coordinate equation set can be established: wherein, denotes the three-dimensional coordinate of the center point of the obstacle, denotes the coordinate of the center point of the obstacle in the actual coordinate system, and denote the scaling ratios in the and directions, respectively, and denote the offsets in the and directions, respectively, denotes the height of the center point of the obstacle from the ground plane.

[0093] By solving the above three-dimensional coordinate equation set, the three-dimensional center point coordinates of the obstacle can be obtained , and the three-dimensional center point coordinates of the obstacle can be finally obtained based on the three-dimensional coordinate equation set through the above implementation steps.

[0094] In the embodiment, the edge fitting parameters are obtained based on the obstacle contour information and the joint information set, and a center point coordinate analysis function is further derived to determine the two-dimensional center point coordinates of the obstacle. Various factors in the image can be fully considered to affect the edge fitting, and then the center of the ellipse fitted by the edge is calculated, so that the center point positioning is more accurate, and a reliable foundation is provided for subsequent three-dimensional positioning and collision avoidance judgment.

[0095] In another aspect, the three-dimensional coordinate equation set is constructed by combining the plane equation expression and the mapping relationship between the pixel coordinates and the actual coordinates, which can accurately calculate the center point coordinates of the obstacle in three-dimensional space, and has an important role in accurately identifying the position of the obstacle during the operation of the tire crane, and can effectively avoid collision accidents caused by inaccurate positioning, and improve the operation safety.

[0096] S4, a collision risk dynamic evaluation system is constructed, and the collision risk of the tire crane is predicted by combining the operation status of the tire crane, the three-dimensional center point coordinates, the joint information set and the collision risk dynamic evaluation system, to realize dynamic identification and collision warning of the tire crane, and the specific implementation contents are as follows: After successfully obtaining the three-dimensional center point coordinates of the obstacle , in order to further protect the operation safety of the tire crane, the motion state of the obstacle and the tire crane (including speed, acceleration, steering angle, etc.) and the position, speed and three-dimensional center point coordinates of the obstacle are further analyzed based on the motion state monitoring equipment and sensors. Through the related information, the risk of collision between the tire crane and the obstacle can be evaluated in real time and accurately.

[0097] According to various motion state sensors installed on the tire crane, the encoder and the gyroscope can obtain the motion data of the tire crane in real time. The encoder can be used to measure the running speed and acceleration of the tire crane. By recording the change of pulse number output by the encoder in unit time, and combining the resolution and transmission ratio of the encoder, the speed and acceleration of the tire crane can be accurately calculated. The gyroscope can measure the steering angle of the tire crane , which can sensitively perceive the rotational motion of the crane in the horizontal plane, and output the angle change in the form of electrical signal, and after processing, the accurate steering angle data can be obtained.

[0098] The position information of the obstacle can be obtained by using sensors such as laser radar and camera. Meanwhile, the position information of the obstacle at different time is continuously tracked and analyzed, and the speed of the obstacle can be calculated .

[0099] Suppose at two continuous time and , the three-dimensional center point coordinates of the obstacle are and , so the speed components of the obstacle in three coordinate axis directions (x, y, z) are respectively: The combined speed of the obstacle can be calculated by the following formula: The relative distance prediction function and the relative speed prediction function are established in the collision risk dynamic evaluation system.

[0100] According to the position coordinate information of the tire crane and the obstacle, the relative distance between the two can be calculated .

[0101] Suppose at any time , the three-dimensional center point coordinates of the tire crane are , and the three-dimensional center point coordinates of the obstacle are The relative distance between the two can be calculated by the distance formula between two points in three-dimensional space, that is, the relative distance prediction function satisfies the following relationship: To calculate the relative speed between the tire crane and the obstacle , the relative speed components of the two in three coordinate axis directions need to be determined first. Suppose the speed components of the tire crane in the x, y, z axis directions are respectively , , , , and the speed components of the obstacle in the x, y, z axis directions are respectively , , , , then the components of the relative speed in three axis directions are respectively: The relative speed can be calculated by the following formula, that is, the relative speed prediction function is as follows: ​By combining the operational status of the tire crane, the three-dimensional center point coordinates, the joint information set, the relative distance prediction function, and the relative speed prediction function, the relative distance and relative speed between the obstacle and the tire crane are analyzed, further providing technical support and reference for the collision avoidance warning of the tire crane.

[0102] In the dynamic assessment system for collision risk of tire cranes, in order to more accurately predict the collision time between obstacles and tire cranes, the embodiment constructs a collision time analysis equation under motion state, and comprehensively considers the impact of factors such as changes in motion state and steering angle on collision time.

[0103] Construct collision time analysis equations under motion conditions.

[0104] Due to the acceleration of the tire crane trolley This will change its speed, thus affecting the collision time. Assuming the collision time is predicted... Inside, the acceleration of the tire crane trolley remains constant, and the collision time can be corrected according to the displacement formula of uniformly accelerated linear motion.

[0105] Let the elapsed time be... Afterwards, the displacement of the tire crane trolley is The displacement of the obstacle is If the two collide, then the following condition is met: in, This represents the distance between the two at the initial moment, i.e., the relative distance. .

[0106] Displacement of the tire crane trolley The following relationship must be satisfied: .

[0107] Displacement of obstacles The following relationship must be satisfied: .

[0108] Substituting the above two equations into From this, we can obtain the collision time analysis equation under motion conditions, which satisfies the following relationship: Further analysis reveals: When solving the above equations, positive real solutions are taken as the collision time after considering the effect of acceleration. ,when hour, .

[0109] Further, the turning angle of the tire crane is introduced to set and distribute acceleration weight and turning angle weight based on the motion state of the tire crane and the turning angle of the tire crane.

[0110] Turning angle of the tire crane The tire crane will change its driving direction, thereby affecting the relative position relationship with the obstacle and the collision time. In order to consider the influence of the turning angle, the motion of the tire crane is decomposed into two components along the current driving direction and perpendicular to the current driving direction.

[0111] The turning radius of the tire crane is set as According to the turning angle and the wheelbase of the tire crane , the turning radius can be calculated.

[0112] In the case of considering turning, the motion trajectory of the tire crane is a circular arc. According to the geometric relationship and kinematics formula of the circular arc motion, combined with the motion information of the obstacle, the collision time is corrected by numerical simulation or iterative calculation method to obtain the collision time after considering the influence of the turning angle .

[0113] In order to comprehensively consider the influence of acceleration and turning angle on the collision time, in the embodiment, the weighted average or other reasonable method is adopted to fuse and to obtain a more accurate dynamic collision time . In the embodiment, different weights and are distributed to and according to the size of acceleration and turning angle, wherein .

[0114] The acceleration weight and the turning angle weight are combined to adjust and optimize the collision time to obtain the dynamic collision time between the obstacle and the tire crane.

[0115] Further adjust and optimize the collision time to obtain the dynamic collision time, in the embodiment, the acceleration weight and the turning angle weight are combined to adjust and optimize the collision time, thereby obtaining the dynamic collision time between the obstacle and the tire crane , and the calculation formula is: In actual application, the above weights can be adjusted according to actual situation. When the acceleration has greater influence on the motion state, the can be increased; when the turning angle has greater influence on the motion state, the can be increased.In this way, the collision time between the obstacle and the tire crane can be more accurately predicted, providing technical support for collision warning and safety management.

[0116] The relative distance, relative speed, and dynamic collision time are combined to predict the collision risk between the obstacle and the tire crane, so as to realize dynamic identification and anti-collision warning of the tire crane.

[0117] According to the calculated dynamic collision time and the preset risk threshold , , the collision risk is divided into three risk levels: low, medium, and high, and the specific division standards are as follows: When , it indicates that the tire crane has sufficient time to collide with the obstacle, and the collision risk is low, which is determined as a low risk level.

[0118] When , it means that the collision time is gradually approaching, and there is a certain possibility of collision, which is classified as a medium risk level.

[0119] When , it means that the collision is about to occur, and the collision risk is extremely high, which is classified as a high risk level.

[0120] Among them, and are the upper and lower limits of the preset risk threshold respectively and .

[0121] In actual operation, the preset threshold can be flexibly adjusted according to different working environments and safety requirements. In an optional embodiment, , .

[0122] Further, the program code of the dynamic collision risk assessment model is as follows: import math # Preset risk threshold η_t1 = 9 # Low risk threshold (seconds) η_t2 = 3 # High risk threshold (seconds) # Simulate tire crane motion state data v_carx = 1 # Car x-axis speed component (m / s) v_cary = 0 # Car y-axis speed component (m / s) v_carz = 0 # Car z-axis speed component (m / s) a_carx = 0 # Car x-axis acceleration component (m / s²) a_cary = 0 # Big car y-axis acceleration component (m / s²) a_carz = 0 # Big car z-axis acceleration component (m / s²) theta_car = 0 # Big car steering angle (radians) # Simulate obstacle information X0 = 10 # Obstacle x-coordinate (m) Y0 = 5 # Obstacle y-coordinate (m) Z0 = 0 # Obstacle z-coordinate (m) v_obx = -0.5 # Obstacle x-axis velocity component (m / s) v_oby = 0 # Obstacle y-axis velocity component (m / s) v_obz = 0 # Obstacle z-axis velocity component (m / s) # Simulate tire crane car position X_car = 0 # Big car x-coordinate (m) Y_car = 0 # Big car y-coordinate (m) Z_car = 0 # Big car z-coordinate (m) # Calculate relative distance d_oc=√((X_car-X_0 )^2+(Y_car-Y_0 )^2+(Z_car-Z_0 )^2 ) # Calculate relative velocity components v_oc-x=(v_car-x)-(v_ob-x) v_oc-y=(v_car-y)-(v_ob-y) v_oc-Z=(v_car-z)-(v_ob-z) # Calculate relative velocity v_oc=√((v_oc-x)^2+(v_oc-y)^2+(v_oc-Z)^2 ) # Preliminary prediction of collision time if v!= 0: t = d / v else: t = float('inf') # Relative speed is 0, consider no collision # Consider acceleration influence to correct collision time (assuming only x-axis acceleration) # Quadratic equation coefficients a = 0.5 * a_carx b = v_carx - v_obx c = d # Solve a quadratic equation discriminant = b**2 - 4 * a * c if discriminant >= 0: η_t1 = (-b + math.sqrt(discriminant)) / (2 * a) η_t2= (-b - math.sqrt(discriminant)) / (2 * a) t_a = min(η_t1, η_t2) elifη_t1> 0: t_a =η_t1 elifη_t2> 0: t_a =η_t2 else: t_a = t else: t_a = t # Consider the impact of steering angle on corrected collision time # Assuming that steering has some impact on collision time, multiply by a coefficient (random simulation between 0.8 and 1.2). import random if theta_car != 0: t_theta = t_a * random.uniform(0.8, 1.2) else: t_theta = t_a # Dynamic collision time obtained through comprehensive correction t_dyn = (t_a + t_theta) / 2 # Assess collision risk level if t_dyn > η_t1: risk_level = "Low risk" elifη_t2 < t_dyn <= η_t1: risk_level = "Medium risk" else: risk_level = "High risk" print(f"Relative distance: {d:.2f} meters") print(f"Relative speed: {v:.2f} m / s") print(f"Initial predicted collision time: {t:.2f} s") print(f"Collision time with acceleration correction: {t_a:.2f} s") print(f"Collision time with steering angle correction: {t_theta:.2f} s") print(f"Dynamic collision time: {t_dyn:.2f} s") print(f"Collision risk level: {risk_level}") The anti-collision identification and warning method for the tire crane also includes a warning and automatic obstacle avoidance mechanism. Different levels of warning signals can be issued according to the collision risk level, and the tire crane control system is linked to realize automatic obstacle avoidance. The specific steps are as follows: Low risk level warning: When the collision risk is at a low risk level, a warning signal is sent to the operator through sound prompts and display screen flashing, reminding the operator to pay attention to the surrounding environment.

[0123] Medium risk level warning: When the collision risk is at a medium risk level, in addition to sound prompts and display screen flashing, the tire crane's speed reduction device is started to reduce the speed of the crane.

[0124] High risk level warning: When the collision risk is at a high risk level, a strong alarm sound is immediately emitted, and the tire crane's emergency brake device is automatically triggered to stop the crane from running, and an emergency help signal is sent to the port dispatching system.

[0125] In the embodiment, the warning system uses a sound and light alarm, which is installed in the cab of the tire crane and in a conspicuous position outside. The warning methods under different risk levels are as follows: Low risk level: The sound and light alarm emits intermittent beeping sounds, and the display screen in the cab flashes.

[0126] Medium risk level: The beeping sound frequency is accelerated, the display screen displays a speed reduction prompt, and the tire crane's speed reduction device is started through a control relay.

[0127] High risk level: The sound and light alarm emits a continuous strong alarm sound, the display screen displays an emergency brake prompt, and a control signal is sent to the emergency brake interface of the tire crane control system to trigger the emergency brake device to stop the crane from running, and an emergency help signal is sent to the port dispatching system through the wireless communication module.

[0128] The anti-collision identification and early warning method for the tire crane vehicle fully gives play to the data acquisition advantages of different sensors, effectively overcomes the problem of single sensor interference by environmental factors, and can effectively improve the accuracy and reliability of the operation monitoring information of the tire crane vehicle through plane fitting analysis and plane point projection processing methods.

[0129] The method for determining the obstacle center point coordinates can accurately identify the actual situation and specific position of various obstacles around the tire crane vehicle, and also has good identification effect for complex-shaped or seriously shielded obstacles.

[0130] The dynamic collision risk assessment model can analyze the motion state changes between the tire crane vehicle and the obstacles in real time, which helps to accurately assess the collision risk and provides a reliable reference basis for the tire crane vehicle warning and automatic obstacle avoidance. At the same time, different levels of warning signals are sent according to the collision risk level, and automatic obstacle avoidance is realized by linking with the tire crane control system, which can effectively avoid collision accidents and improve the safety of tire crane vehicle operation.

[0131] See Figure 2 In an optional embodiment, in order to efficiently implement the anti-collision identification and early warning method for the tire crane vehicle provided by the present application, the present application further provides an anti-collision identification and early warning system for the tire crane vehicle, which comprises a processor, an input device, an output device and a memory, and the above-mentioned processor, input device, output device and memory are connected with each other, wherein the memory is used to store a computer program, the computer program comprises program instructions, the processor is configured to call the program instructions, and the specific steps of the anti-collision identification and early warning method for the tire crane vehicle and related embodiments provided by the present application are executed. The anti-collision identification and early warning system for the tire crane vehicle of the present application has a complete structure and is objective and stable.

[0132] Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present application, and not to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that: it can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacement for part or all of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present application, and they should be covered in the scope of the claims and the description of the present application.

Claims

1. A collision avoidance identification and warning method for a tyre crane, characterized in that, The method comprises: acquiring original data information in the operation process of the tire crane, calibrating and correcting the laser radar information and the monitoring image information in the original data information, and jointly calibrating to obtain a joint information set in the operation process of the tire crane; according to the running characteristics of the tire crane and the joint information set, performing plane fitting analysis and plane point projection processing on the obstacles to obtain obstacle contour graph information in the operation process of the tire crane; determining the center point coordinates of the obstacles according to the obstacle contour graph information and the joint information set, and analyzing the three-dimensional center point coordinates of the obstacles based on the center point coordinates and the joint information set; constructing a collision risk dynamic evaluation system, combining the operation status of the tire crane, the three-dimensional center point coordinates, the joint information set, and the collision risk dynamic evaluation system to predict the collision risk of the tire crane, so as to realize dynamic identification and anti-collision warning of the tire crane.

2. The anti-collision identification and early warning method for the tyre crane according to claim 1, characterized in that, The calibration and correction of the laser radar information and the monitoring image information in the original data information to obtain the joint information set in the operation process of the tire crane comprises: setting a scanning starting point of the laser radar information based on an information acquisition method of the laser radar device; dividing the laser radar information acquisition path into a plurality of scanning line segments according to the installation position of the laser radar device and the scanning starting point; analyzing the line segment slope value between adjacent laser radar devices according to the plurality of scanning line segments and the installation position of the laser radar device.

3. The anti-collision identification and early warning method for the tyre crane according to claim 2, characterized in that, The calibration and correction of the laser radar information and the monitoring image information in the original data information to obtain the joint information set in the operation process of the tire crane comprises: analyzing the adjacent slope difference between adjacent laser radar devices according to the line segment slope value; setting a dynamic slope threshold and a neighborhood radius threshold according to the installation position of the laser radar device; calibrating and correcting the laser radar information to obtain a laser radar data set in combination with the adjacent slope difference, the dynamic slope threshold, and the neighborhood radius threshold.

4. The anti-collision identification and early warning method for the tyre crane according to claim 1, characterized in that, The calibration and correction of the laser radar information and the monitoring image information in the original data information to obtain the joint information set in the operation process of the tire crane comprises: introducing a standard homogeneous coordinate, and establishing a radar-camera coordinate transformation model based on the standard homogeneous coordinate; analyzing and transforming the laser radar data set and the monitoring image information through the radar-camera coordinate transformation model to obtain feature point three-dimensional coordinates in the camera coordinate system; establishing a camera-image coordinate transformation model according to the pinhole camera projection mechanism; using the camera-image coordinate transformation model to analyze and convert the feature point three-dimensional coordinates in the camera coordinate system to obtain feature point coordinates in the image coordinate system.

5. The anti-collision identification and early warning method for the tyre crane according to claim 4, characterized in that, The calibration and correction of the laser radar information and the monitoring image information in the original data information to obtain the joint information set in the operation process of the tire crane comprises: establishing a pixel coordinate transformation model in combination with the image distortion condition and the resolution correction principle; Calibrating and analyzing the feature point coordinates in the image coordinate system by using the pixel coordinate transformation model to obtain pixel coordinates of the feature points in the monitoring image; Obtaining a joint information set in the tire-mounted crane operation process by combining the pixel coordinates, the laser radar data set and the monitoring image information.

6. The anti-collision identification and early warning method for the tyre crane according to claim 1, characterized in that, The plane fitting analysis and the plane point projection processing of the obstacle according to the tire-mounted crane running characteristics and the joint information set to obtain the obstacle contour graph information in the tire-mounted crane operation process includes: Randomly extracting a reference point based on the tire-mounted crane running characteristics and the joint information set; Obtaining a plane normal vector according to the covariance matrix and the reference point; Establishing a plane equation expression and a point-plane distance analysis expression according to the plane normal vector; Performing the plane fitting analysis of the obstacle by using the plane equation expression, the point-plane distance analysis expression and the joint information set, and obtaining the obstacle plane fitting result.

7. The anti-collision identification and early warning method for the tyre crane according to claim 6, characterized in that, The plane fitting analysis and the plane point projection processing of the obstacle according to the tire-mounted crane running characteristics and the joint information set to obtain the obstacle contour graph information in the tire-mounted crane operation process includes: Setting a calibration board point cloud; Obtaining an obstacle plane projection analysis function based on the space projection method, the joint information set and the calibration board point cloud; Obtaining obstacle plane projection information according to the obstacle plane projection analysis function; Introducing an image coordinate axis vector; Obtaining an obstacle pixel coordinate result according to the image coordinate axis vector and the obstacle plane projection information; Integrating the obstacle pixel coordinate result to obtain the obstacle contour graph information in the tire-mounted crane operation process.

8. The anti-collision identification and early warning method for the tyre crane according to claim 1, characterized in that, The center point coordinates of the obstacle are determined according to the obstacle contour graph information and the joint information set, and the three-dimensional center point coordinates of the obstacle are analyzed based on the center point coordinates and the joint information set, which includes: Obtaining edge adjustment parameters based on the obstacle contour graph information and the joint information set; Deriving and establishing a center point coordinate analysis function according to an image coordinate axis vector, the edge adjustment parameters and the obstacle contour graph information; Determining the center point coordinates of the obstacle by using the center point coordinate analysis function; Constructing a three-dimensional coordinate equation set according to a plane equation expression and the center point coordinates of the obstacle; Obtaining the three-dimensional center point coordinates of the obstacle based on the three-dimensional coordinate equation set.

9. The anti-collision identification and warning method for the tyre crane according to claim 1, characterized in that, The collision risk dynamic evaluation system is constructed, the collision risk of the tire-mounted crane is predicted by combining the tire-mounted crane operation condition, the three-dimensional center point coordinates, the joint information set and the collision risk dynamic evaluation system, so as to realize the dynamic identification and anti-collision warning of the tire-mounted crane, which includes: Establishing a relative distance prediction function and a relative speed prediction function in the collision risk dynamic evaluation system; Analyzing the relative distance and the relative speed between the obstacle and the tire-mounted crane by combining the tire-mounted crane operation condition, the three-dimensional center point coordinates, the joint information set, the relative distance prediction function and the relative speed prediction function; Establishing a collision time analysis equation in the motion state in the collision risk dynamic evaluation system according to the tire-mounted crane motion state. The collision time analysis equation in the motion state is used to predict the collision time between the obstacle and the tire crane trolley; The steering angle of the tire crane trolley is introduced, and the acceleration weight and the steering angle weight are set and distributed based on the motion state of the tire crane trolley and the steering angle of the tire crane trolley; The collision time is adjusted and optimized in combination with the acceleration weight and the steering angle weight to obtain the dynamic collision time between the obstacle and the tire crane trolley; The collision risk between the obstacle and the tire crane trolley is predicted in combination with the relative distance, the relative speed and the dynamic collision time to realize the dynamic identification and anti-collision warning of the tire crane trolley.

10. A collision avoidance identification warning system for a tyre crane, characterised in that, The system comprises a processor, an input device, an output device and a memory, which are connected with each other, wherein the memory is used to store a computer program, the computer program comprises program instructions, and the processor is configured to call the program instructions to execute the anti-collision identification and warning method for the tire crane trolley according to any one of claims 1-9.

Images