Human-robot collaboration based robotic perception method and system
By using MEMS galvanometer-driven laser beam reflection technology, a 3D point cloud of AGV cargo is generated and the edge trajectory and protection trajectory are scanned alternately. This solves the problem that traditional AGVs cannot sense obstacles on the outside of the cargo and improves transportation safety.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- BEIJING VOCATIONAL COLLEGE OF ECONOMICS & MANAGEMENT (BEIJING MANAGER COLLEGE)
- Filing Date
- 2026-05-28
- Publication Date
- 2026-07-17
AI Technical Summary
Traditional AGV obstacle avoidance systems cannot effectively detect obstacles outside the cargo, leading to safety hazards and property damage.
A laser beam driven by a MEMS galvanometer is reflected by a ring-shaped curved mirror to scan the cargo, acquire a three-dimensional point cloud, establish a standard outer edge contour, and generate cargo edge trajectory and virtual protection trajectory. The scanning is performed alternately to monitor abnormal movement.
It enables omnidirectional scanning of goods above AGVs, dynamically generates goods edge trajectories and virtual protection trajectories, and provides real-time warnings of potential goods loosening or collision hazards, thereby improving transportation safety.
Smart Images
Figure CN122408657A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of robot perception technology, and particularly relates to robot perception methods and systems based on human-machine collaboration. Background Technology
[0002] With the development of automated logistics and intelligent manufacturing, Automated Guided Vehicles (AGVs) have been widely used in warehousing, production lines, and other scenarios. AGVs are typically equipped with LiDAR, ultrasonic, or vision sensors to detect obstacles around their vehicle and achieve autonomous obstacle avoidance and path planning. However, the protection range of these sensing systems is usually limited to the outer dimensions of the AGV itself. There are significant blind spots in the protection of goods carried on top of the AGV, especially materials that are irregular in size, have varying stacking heights, or are excessively long or wide.
[0003] In actual operation, the goods carried by AGVs may shift, tilt or even become loose during transportation; at the same time, when the shape of the goods exceeds the outline of the vehicle body, traditional obstacle avoidance systems based on the vehicle body cannot detect obstacles outside the goods, which can easily lead to collisions, causing safety hazards and property damage. Summary of the Invention
[0004] The purpose of this invention is to provide a robot perception method based on human-machine collaboration, which aims to solve the problem that traditional obstacle avoidance systems based on the vehicle body cannot perceive obstacles outside the cargo, easily causing collisions, resulting in safety hazards and property damage.
[0005] This invention is implemented as follows: a robot perception method based on human-machine collaboration, the method comprising: A laser beam driven by a MEMS galvanometer is reflected by a ring-shaped curved mirror to scan the cargo, obtain a three-dimensional point cloud, and establish a standard outer edge contour. Based on the standard outer edge profile, a set of laser emission angle sequences and their corresponding illumination paths on the annular curved surface reflector are determined to form the cargo edge trajectory. Based on the standard outer edge contour, a virtual protective line parallel to the edge of the cargo is generated, and the corresponding laser emission angle sequence and curved mirror irradiation path are determined accordingly to form an obstacle protection trajectory. During the AGV's operation, edge trajectory and protection trajectory scanning are performed alternately, and data analysis is conducted simultaneously to determine whether there are any abnormal movements.
[0006] Preferably, the step of determining a set of laser emission angle sequences and their corresponding illumination paths on the annular curved surface mirror based on the standard outer edge contour to form the cargo edge trajectory includes: The standard outer edge contour is discretized into multiple feature points, each feature point corresponding to a spatial direction, which is determined by both the horizontal and vertical angles. Based on the known surface parameters of the annular curved mirror, the spatial direction of each feature point is reversed and mapped to the deflection angle of the MEMS galvanometer and the incident point position of the laser beam on the curved mirror. Arrange the galvanometer deflection angles corresponding to all feature points according to the geometric order of the cargo edge to generate a continuous laser emission angle sequence and its corresponding illumination path on the curved mirror, thus obtaining the cargo edge trajectory.
[0007] Preferably, the step of generating a virtual protective line parallel to the edge of the cargo based on the standard outer contour, and determining the corresponding laser emission angle sequence and curved mirror irradiation path accordingly to form an obstacle protection trajectory includes: Based on the standard outer edge contour, a fixed safety distance is preset on the outside of the cargo, and a virtual protective line is generated that is parallel to the edge of the cargo and maintains a safe distance from the edge. The virtual protective line is composed of a series of spatial points. The orientation of each spatial point on the virtual protective line is reverse-mapped to the deflection angle of the MEMS galvanometer and the position of the incident point of the laser beam on the curved surface mirror based on the surface parameters of the annular surface mirror. The mirror deflection angles corresponding to all points on the virtual protective line are sorted according to the geometric direction of the protective line to generate corresponding illumination paths, thus obtaining the obstacle protection trajectory.
[0008] Preferably, the step of alternately scanning the edge trajectory and the protection trajectory during AGV operation, and simultaneously performing data analysis to determine whether there is any movement anomaly, includes: During the journey, the MEMS galvanometer is cyclically controlled to perform laser scanning of the cargo edge trajectory and obstacle protection trajectory in a point-to-point alternation manner. The point cloud data of the cargo surface obtained by edge trajectory scanning and the obstacle ranging data obtained by protective trajectory scanning are recorded separately, and the two types of data are identified and classified according to trajectory type and stored separately. The edge trajectory point cloud data is compared with the standard outer edge contour, and the protection trajectory ranging data is compared with the preset safe distance threshold, and the judgment result is output.
[0009] Preferably, the distance between the virtual protective line and the edge of the goods is 5-10cm.
[0010] Another object of the present invention is to provide a robot perception system based on human-machine collaboration, the system comprising: The contour detection module is used to scan the cargo by driving a laser beam through a MEMS galvanometer and reflecting it through a ring-shaped curved mirror, thereby acquiring a three-dimensional point cloud and establishing a standard outer edge contour. The edge trajectory construction module is used to determine a set of laser emission angle sequences and their corresponding illumination paths on the annular curved surface mirror based on the standard outer edge contour, thereby forming the cargo edge trajectory; The protection trajectory construction module is used to generate a virtual protection line parallel to the edge of the cargo based on the standard outer edge contour, thereby determining the corresponding laser emission angle sequence and the curved mirror irradiation path to form an obstacle protection trajectory. The movement protection module is used to alternately scan the edge trajectory and the protection trajectory during the AGV's movement, and simultaneously perform data analysis to determine whether there are any movement anomalies.
[0011] Preferably, the edge trajectory construction module includes: The feature point discrete unit is used to discretize the standard outer edge contour into multiple feature points. Each feature point corresponds to a spatial direction, which is determined by the horizontal angle and the vertical angle. The first route mapping unit is used to reverse map the spatial direction of each feature point to the deflection angle of the MEMS galvanometer and the incident point position of the laser beam on the curved surface mirror based on the known surface parameters of the annular curved surface mirror. The first trajectory construction unit is used to arrange the galvanometer deflection angles corresponding to all feature points according to the geometric order of the cargo edge, generate a continuous laser emission angle sequence and its corresponding illumination path on the curved reflector, and obtain the cargo edge trajectory.
[0012] Preferably, the protection trajectory construction module includes: The protective line construction unit is used to generate a virtual protective line that is parallel to the edge of the cargo and maintains a safe distance from the edge, based on a standard outer edge contour and a fixed safety distance from the outer edge of the cargo. The virtual protective line is composed of a series of spatial points. The second route mapping unit is used to reverse map the direction of each spatial point on the virtual protection line to the deflection angle of the MEMS galvanometer and the position of the incident point of the laser beam on the curved surface mirror according to the surface parameters of the annular curved surface mirror. The second trajectory construction unit is used to sort the galvanometer deflection angles corresponding to all points on the virtual protective line according to the geometric direction of the protective line, generate the corresponding illumination path, and obtain the obstacle protection trajectory.
[0013] Preferably, the mobile protection module includes: The alternating scanning unit is used to cyclically control the MEMS galvanometer to perform laser scanning of the cargo edge trajectory and obstacle protection trajectory in a point-by-point alternating manner during the driving process; The scanning data storage unit is used to record the cargo surface point cloud data obtained by edge trajectory scanning and the obstacle ranging data obtained by protective trajectory scanning, and to identify and classify the two types of data according to trajectory type for storage; The protection determination unit is used to compare the edge trajectory point cloud data with the standard outer edge contour, compare the protection trajectory ranging data with the preset safety distance threshold, and output the determination result.
[0014] Preferably, the distance between the virtual protective line and the edge of the goods is 5-10cm.
[0015] The robot perception method based on human-machine collaboration provided by this invention achieves omnidirectional scanning of goods above AGV through the cooperation of a single MEMS galvanometer and a ring-shaped curved surface reflector. It can dynamically generate the edge trajectory of goods and virtual protection trajectory without adding multiple sets of sensors, and simultaneously monitors the displacement of goods and external obstacles through a cross-scanning mechanism. This solves the blind spot problem of traditional AGVs being unable to perceive the risks on the outside of oversized goods, and can also provide real-time warnings of loose goods or collision hazards, greatly improving transportation safety. Attached Figure Description
[0016] Figure 1 A flowchart illustrating a robot perception method based on human-machine collaboration provided in an embodiment of the present invention; Figure 2 This is an architecture diagram of a robot perception system based on human-machine collaboration, provided in an embodiment of the present invention. Detailed Implementation
[0017] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.
[0018] like Figure 1 The diagram shows a flowchart of a robot perception method based on human-machine collaboration provided in an embodiment of the present invention. The method includes: The S100 uses a MEMS galvanometer to drive a laser beam, which is reflected by a ring-shaped curved mirror to scan the cargo, acquire a three-dimensional point cloud, and establish a standard outer edge contour.
[0019] In this step, a laser rangefinder is installed at the center of the bottom of the AGV robot. The laser rangefinder is equipped with a MEMS galvanometer. The emitted light from the laser rangefinder is reflected by the MEMS galvanometer, which can adjust the angle in the horizontal plane and the vertical plane. A ring-shaped curved mirror is installed at the bottom of the robot. The detection light is reflected by the ring-shaped curved surface, which can adjust the angle, thereby realizing the distance detection from bottom to top. When the goods are loaded onto the AGV, the laser beam driven by the MEMS galvanometer is reflected by the ring-shaped curved mirror to scan the goods, thereby obtaining a series of three-dimensional point clouds. Based on the three-dimensional point cloud, the outline of the goods can be extracted, which is the standard outer edge outline.
[0020] S200, based on the standard outer edge contour, determines a set of laser emission angle sequences and their corresponding irradiation paths on the annular curved surface reflector, forming the cargo edge trajectory.
[0021] In this step, after obtaining the standard outer edge contour model of the cargo, the standard outer edge contour model is discretized into a series of spatial feature points. Each feature point corresponds to a laser emission direction determined by a horizontal angle and a vertical angle. That is, the detection light is emitted along the horizontal angle and the vertical angle to illuminate the corresponding spatial feature point. According to the known surface parameters of the annular curved mirror, which are obtained through actual measurement and calibration, the spatial direction of each feature point is reverse-mapped to the deflection angle that the MEMS galvanometer needs to perform, as well as the incident point position of the laser beam on the curved mirror. The galvanometer deflection angles corresponding to all feature points are arranged in clockwise order according to the edge of the cargo to generate a continuous laser emission angle sequence and its corresponding illumination path on the curved mirror. This path is the cargo edge trajectory.
[0022] S300 generates a virtual protective line parallel to the edge of the cargo based on the standard outer contour, and determines the corresponding laser emission angle sequence and curved mirror irradiation path accordingly to form an obstacle protection trajectory.
[0023] In this step, based on the established standard outer edge contour, a fixed safety distance, such as 10cm, is preset on the outside of the cargo. This distance is adjusted according to the actual working conditions to generate a virtual protective line parallel to the edge of the cargo and maintaining a constant distance from the edge. This virtual protective line consists of a series of spatial points. The closest distance between the virtual protective line and the edge of the cargo is exactly equal to the preset safety distance. Using the same optical reverse mapping method as the previous step, the direction of each spatial point on the virtual protective line is converted into the deflection angle of the MEMS galvanometer and the incident point position on the curved mirror. These angles are sorted clockwise according to the protective line to obtain the obstacle protection trajectory.
[0024] During the AGV's operation, the S400 alternately scans the edge trajectory and the protection trajectory, and simultaneously performs data analysis to determine whether there are any abnormal movements.
[0025] In this step, during the AGV's movement, the system cyclically scans the cargo edge trajectory and obstacle protection trajectory according to a preset point-to-point alternation method. This involves merging the two sets of trajectories: first scanning one or more spatial points in the cargo edge trajectory, then scanning one or more spatial points in the obstacle protection trajectory, repeating this process alternately to address both monitoring tasks. The distance measurement data obtained from each scan is collected in real time and stored in a categorized manner. The edge trajectory data reflects the current distance of each point on the cargo surface, while the protection trajectory data reflects the distance of each point on the virtual safety boundary. On one hand, the current edge trajectory data is compared with the initial standard contour. If the deviation at any point exceeds a preset offset threshold, it is determined that the cargo has shifted or become loose, and an alarm signal is immediately issued. On the other hand, the distance measured by the protection trajectory is compared with a safety distance threshold. If the distance at any point is found to be less than the safety threshold, it is determined that there is a collision risk in that direction, and a deceleration or stop command is immediately issued to the AGV control system.
[0026] In a preferred embodiment of the present invention, the step of determining a set of laser emission angle sequences and their corresponding illumination paths on the annular curved surface mirror based on the standard outer edge contour to form the cargo edge trajectory includes: S201 discretizes the standard outer edge contour into multiple feature points, each feature point corresponding to a spatial direction, which is determined by the horizontal angle and the vertical angle.
[0027] In this step, after obtaining the standard outer contour of the goods, multiple feature points are extracted from it. The standard outer contour is essentially composed of a series of three-dimensional coordinate points. Feature point extraction can be performed based on a preset edge precision. For example, according to a preset angular resolution, a sampling point is set every 1°, and multiple feature points are extracted evenly on the contour. Each feature point contains its coordinate information in three-dimensional space and the angle information of the detection light. The angle information consists of the horizontal azimuth angle and the vertical pitch angle. The horizontal azimuth angle is the yaw angle of the light relative to the AGV vehicle coordinate system, and the vertical pitch angle is the elevation angle or depression angle of the light relative to the horizontal plane.
[0028] S202, based on the known surface parameters of the annular curved mirror, reverse the spatial direction of each feature point to the deflection angle of the MEMS galvanometer and the incident point position of the laser beam on the curved mirror.
[0029] In this step, for each feature point, reverse ray tracing calculation is performed based on the known surface parameters of the toroidal surface mirror. The known surface parameters include the curvature and normal direction of each point on the surface. Specifically, it is known that after the laser beam is reflected by the surface mirror, it needs to exit along the spatial direction specified by the feature point. By solving the inverse problem of the law of reflection, the incident laser direction corresponding to this exit direction is deduced, thereby determining the angle to which the MEMS galvanometer needs to be deflected, as well as the position of the incident point of the laser beam on the surface reflection.
[0030] S203, arrange the mirror deflection angles corresponding to all feature points in clockwise order according to the edge of the cargo to generate a continuous laser emission angle sequence and the corresponding illumination path on the curved mirror, thus obtaining the cargo edge trajectory. Sort the MEMS mirror deflection angles corresponding to all feature points in clockwise order according to the edge of the cargo to form a control command sequence. Each command in the control command sequence contains the target deflection angle of the mirror. Based on the control command sequence, the feature points on the edge of the cargo can be scanned.
[0031] In a preferred embodiment of the present invention, the step of generating a virtual protective line parallel to the edge of the cargo based on the standard outer edge contour, and determining the corresponding laser emission angle sequence and curved mirror irradiation path accordingly to form an obstacle protection trajectory includes: S301, based on the standard outer edge contour, a fixed safety distance is preset on the outside of the cargo, and a virtual protective line is generated that is parallel to the edge of the cargo and maintains a safe distance from the edge. The virtual protective line is composed of a series of spatial points.
[0032] In this step, a fixed safety distance is preset on the outside of the goods according to the standard outer edge contour. The safety distance is set according to the requirements. The larger the safety distance, the higher the protection level. For example, if the safety distance is set to 5cm, a virtual protection line is generated. The virtual protection line is parallel to the edge of the goods, and the distance between the two is equal to the safety distance. Similarly, the virtual protection line is also composed of multiple points.
[0033] S302 maps the orientation of each spatial point on the virtual protective line to the deflection angle of the MEMS galvanometer and the position of the laser beam incident on the curved mirror based on the surface parameters of the annular surface mirror.
[0034] In this step, for each spatial point on the virtual protective line, the laser emission direction corresponding to that point is first determined, including the horizontal and vertical angles. Then, based on the curvature, normal direction, and reflection characteristics of each point on the toroidal surface mirror, reverse ray tracing calculation is performed. Taking the point on the virtual protective line as the target emission direction, the solution is reversed to derive the required incident laser direction and determine the angle to which the MEMS galvanometer needs to be deflected. Based on this, the MEMS galvanometer control parameters required for detecting that spatial point can be determined.
[0035] S303: Sort the galvanometer deflection angles corresponding to all points on the virtual protective line according to the geometric direction of the protective line, generate the corresponding illumination path, and obtain the obstacle protection trajectory.
[0036] In this step, the control parameters of the MEMS galvanometer corresponding to the spatial points on the virtual protective line are sorted according to the geometric direction of the protective line. For example, starting from a certain reference point, the entire protective line is traversed clockwise to form a continuous sequence of control commands. The sequence of control commands includes the target deflection angle of the galvanometer. When the galvanometer executes these deflection angles in sequence, the laser beam will form a series of continuous illumination points on the curved mirror. The line connecting these illumination points constitutes the corresponding illumination path on the curved mirror, and the scanning trajectory of the emitted laser beam in space travels along the virtual protective line.
[0037] As a preferred embodiment of the present invention, the step of alternately scanning the edge trajectory and the protection trajectory during the AGV's operation, and simultaneously performing data analysis to determine whether there is any movement anomaly, includes: S401, during the driving process, the MEMS galvanometer is cyclically controlled to perform laser scanning of the cargo edge trajectory and obstacle protection trajectory in a point-to-point alternating manner.
[0038] In this step, during the AGV's movement, the edge trajectory and the protection trajectory are scanned in a point-by-point alternation manner. Point-by-point alternation means that within a complete scanning cycle, after scanning a point on the edge trajectory, the system immediately switches to the corresponding point on the protection trajectory, and then switches back to the next point on the edge trajectory. This alternation continues until all points on both trajectories have been scanned. This alternation strategy allows the two types of monitoring data to be highly intertwined in time, thereby obtaining information on the edge of the cargo and the outer safety boundary almost simultaneously, effectively avoiding the risk of missed detection due to time differences.
[0039] S402 records the cargo surface point cloud data obtained from edge trajectory scanning and the obstacle ranging data obtained from protection trajectory scanning, and identifies and classifies the two types of data according to trajectory type.
[0040] In this step, during the cross-scanning process, the distance data measured at each scan point is collected in real time and recorded together with the trajectory type, spatial direction information, and timestamp of the current scan point. All point cloud data collected from the edge trajectory are grouped together to represent the current outer edge shape of the cargo; all distance data collected from the protection trajectory are grouped together to represent the obstacle distance at each point on the virtual protection line. The two types of data are stored in different memory areas.
[0041] S403 compares the edge trajectory point cloud data with the standard outer edge contour, compares the protection trajectory ranging data with the preset safe distance threshold, and outputs the judgment result.
[0042] In this step, the two types of stored data are processed simultaneously. The real-time point cloud data of the edge trajectory is compared point by point with the initially established standard outer edge contour model. The radial deviation at each feature point is calculated. If the deviation of a feature point exceeds the preset offset threshold, it is determined that the goods have been displaced or loosened, and a displacement alarm signal is immediately generated. The distance measurement data of the protective trajectory is compared with the preset safe distance threshold. If the measured distance of a point is found to be less than the threshold, it is determined that there is a risk of obstacle intrusion in the corresponding direction, and a collision warning signal is immediately generated. The equipment executes the corresponding safety action according to the type of instruction received, such as issuing an audible and visual alarm, actively decelerating to a safe speed, or stopping immediately, thereby ensuring the safety of the transportation process.
[0043] like Figure 2 As shown, this is a robot perception system based on human-machine collaboration provided in an embodiment of the present invention. The system includes: The contour detection module 100 is used to scan the cargo by driving a laser beam through a MEMS galvanometer and reflecting it through a ring-shaped curved mirror, thereby acquiring a three-dimensional point cloud and establishing a standard outer edge contour.
[0044] The edge trajectory construction module 200 is used to determine a set of laser emission angle sequences and their corresponding illumination paths on the annular curved surface mirror based on the standard outer edge contour, thereby forming the edge trajectory of the cargo.
[0045] In this system, the edge trajectory construction module 200 includes: The feature point discrete unit is used to discretize the standard outer edge contour into multiple feature points. Each feature point corresponds to a spatial direction, which is determined by the horizontal angle and the vertical angle. The first route mapping unit is used to reverse map the spatial direction of each feature point to the deflection angle of the MEMS galvanometer and the incident point position of the laser beam on the curved surface mirror based on the known surface parameters of the annular curved surface mirror. The first trajectory construction unit is used to arrange the galvanometer deflection angles corresponding to all feature points according to the geometric order of the cargo edge, generate a continuous laser emission angle sequence and its corresponding illumination path on the curved reflector, and obtain the cargo edge trajectory.
[0046] The protection trajectory construction module 300 is used to generate a virtual protection line parallel to the edge of the cargo based on the standard outer edge contour, thereby determining the corresponding laser emission angle sequence and the curved mirror irradiation path to form an obstacle protection trajectory.
[0047] In this system, the protection trajectory construction module 300 includes: The protective line construction unit is used to generate a virtual protective line that is parallel to the edge of the cargo and maintains a safe distance from the edge, based on a standard outer edge contour and a fixed safety distance from the outer edge of the cargo. The virtual protective line is composed of a series of spatial points. The second route mapping unit is used to reverse map the direction of each spatial point on the virtual protection line to the deflection angle of the MEMS galvanometer and the position of the incident point of the laser beam on the curved surface mirror according to the surface parameters of the annular curved surface mirror. The second trajectory construction unit is used to sort the galvanometer deflection angles corresponding to all points on the virtual protective line according to the geometric direction of the protective line, generate the corresponding illumination path, and obtain the obstacle protection trajectory.
[0048] The mobile protection module 400 is used to alternately scan the edge trajectory and the protection trajectory during the AGV's operation, and simultaneously perform data analysis to determine whether there is any movement abnormality.
[0049] In this system, the mobile protection module 400 includes: The alternating scanning unit is used to cyclically control the MEMS galvanometer to perform laser scanning of the cargo edge trajectory and obstacle protection trajectory in a point-by-point alternating manner during the driving process; The scanning data storage unit is used to record the cargo surface point cloud data obtained by edge trajectory scanning and the obstacle ranging data obtained by protective trajectory scanning, and to identify and classify the two types of data according to trajectory type for storage; The protection determination unit is used to compare the edge trajectory point cloud data with the standard outer edge contour, compare the protection trajectory ranging data with the preset safety distance threshold, and output the determination result.
[0050] 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 robot perception method based on human-machine collaboration, characterized in that, The method includes: A laser beam driven by a MEMS galvanometer is reflected by a ring-shaped curved mirror to scan the cargo, obtain a three-dimensional point cloud, and establish a standard outer edge contour. Based on the standard outer edge profile, a set of laser emission angle sequences and their corresponding illumination paths on the annular curved surface reflector are determined to form the cargo edge trajectory. Based on the standard outer edge contour, a virtual protective line parallel to the edge of the cargo is generated, and the corresponding laser emission angle sequence and curved mirror irradiation path are determined accordingly to form an obstacle protection trajectory. During the AGV's operation, edge trajectory and protection trajectory scanning are performed alternately, and data analysis is conducted simultaneously to determine whether there are any abnormal movements.
2. The robot perception method based on human-machine collaboration according to claim 1, characterized in that, The step of determining a set of laser emission angle sequences and their corresponding illumination paths on the annular curved surface mirror based on the standard outer edge contour to form the cargo edge trajectory includes: The standard outer edge contour is discretized into multiple feature points, each feature point corresponding to a spatial direction, which is determined by the horizontal angle and the vertical angle. Based on the known surface parameters of the annular curved mirror, the spatial direction of each feature point is reversed and mapped to the deflection angle of the MEMS galvanometer and the incident point position of the laser beam on the curved mirror. Arrange the galvanometer deflection angles corresponding to all feature points according to the geometric order of the cargo edge to generate a continuous laser emission angle sequence and its corresponding illumination path on the curved mirror, thus obtaining the cargo edge trajectory.
3. The robot perception method based on human-machine collaboration according to claim 1, characterized in that, The step of generating a virtual protective line parallel to the edge of the cargo based on the standard outer contour, and determining the corresponding laser emission angle sequence and curved mirror irradiation path accordingly to form an obstacle protection trajectory includes: Based on the standard outer edge contour, a fixed safety distance is preset on the outside of the cargo, and a virtual protective line is generated that is parallel to the edge of the cargo and maintains a safe distance from the edge. The virtual protective line is composed of a series of spatial points. The orientation of each spatial point on the virtual protective line is reverse-mapped to the deflection angle of the MEMS galvanometer and the position of the incident point of the laser beam on the curved surface mirror based on the surface parameters of the annular surface mirror. The mirror deflection angles corresponding to all points on the virtual protective line are sorted according to the geometric direction of the protective line to generate the corresponding illumination path, thus obtaining the obstacle protection trajectory.
4. The robot perception method based on human-machine collaboration according to claim 1, characterized in that, The step of alternately scanning edge trajectories and protective trajectories during AGV operation, and simultaneously analyzing data to determine whether there are any movement anomalies, includes: During the journey, the MEMS galvanometer is cyclically controlled to perform laser scanning of the cargo edge trajectory and obstacle protection trajectory in a point-to-point alternation manner. The point cloud data of the cargo surface obtained by edge trajectory scanning and the obstacle ranging data obtained by protective trajectory scanning are recorded separately, and the two types of data are identified and classified according to trajectory type and stored separately. The edge trajectory point cloud data is compared with the standard outer edge contour, and the protection trajectory ranging data is compared with the preset safe distance threshold, and the judgment result is output.
5. The robot perception method based on human-machine collaboration according to claim 1, characterized in that, The distance between the virtual protective line and the edge of the goods is 5-10cm.
6. A robot perception system based on human-machine collaboration, characterized in that, The system includes: The contour detection module is used to scan the cargo by driving a laser beam through a MEMS galvanometer and reflecting it through a ring-shaped curved mirror, thereby acquiring a three-dimensional point cloud and establishing a standard outer edge contour. The edge trajectory construction module is used to determine a set of laser emission angle sequences and their corresponding illumination paths on the annular curved surface mirror based on the standard outer edge contour, thereby forming the cargo edge trajectory; The protection trajectory construction module is used to generate a virtual protection line parallel to the edge of the cargo based on the standard outer edge contour, thereby determining the corresponding laser emission angle sequence and the curved mirror irradiation path to form an obstacle protection trajectory. The movement protection module is used to alternately scan the edge trajectory and the protection trajectory during the AGV's movement, and simultaneously perform data analysis to determine whether there are any movement anomalies.
7. The robot perception system based on human-machine collaboration according to claim 6, characterized in that, The edge trajectory construction module includes: The feature point discrete unit is used to discretize the standard outer edge contour into multiple feature points. Each feature point corresponds to a spatial direction, which is determined by the horizontal angle and the vertical angle. The first route mapping unit is used to reverse map the spatial direction of each feature point to the deflection angle of the MEMS galvanometer and the incident point position of the laser beam on the curved surface mirror based on the known surface parameters of the annular curved surface mirror. The first trajectory construction unit is used to arrange the galvanometer deflection angles corresponding to all feature points according to the geometric order of the cargo edge, generate a continuous laser emission angle sequence and its corresponding illumination path on the curved reflector, and obtain the cargo edge trajectory.
8. The robot perception system based on human-machine collaboration according to claim 6, characterized in that, The protection trajectory construction module includes: The protective line construction unit is used to generate a virtual protective line that is parallel to the edge of the cargo and maintains a safe distance from the edge, based on a standard outer edge contour and a fixed safety distance from the outer edge of the cargo. The virtual protective line is composed of a series of spatial points. The second route mapping unit is used to reverse map the direction of each spatial point on the virtual protection line to the deflection angle of the MEMS galvanometer and the position of the incident point of the laser beam on the curved surface mirror according to the surface parameters of the annular curved surface mirror. The second trajectory construction unit is used to sort the galvanometer deflection angles corresponding to all points on the virtual protective line according to the geometric direction of the protective line, generate the corresponding illumination path, and obtain the obstacle protection trajectory.
9. The robot perception system based on human-machine collaboration according to claim 6, characterized in that, The mobile protection module includes: The alternating scanning unit is used to cyclically control the MEMS galvanometer to perform laser scanning of the cargo edge trajectory and obstacle protection trajectory in a point-by-point alternating manner during the driving process; The scanning data storage unit is used to record the cargo surface point cloud data obtained by edge trajectory scanning and the obstacle ranging data obtained by protective trajectory scanning, and to identify and classify the two types of data according to trajectory type for storage; The protection determination unit is used to compare the edge trajectory point cloud data with the standard outer edge contour, compare the protection trajectory ranging data with the preset safety distance threshold, and output the determination result.
10. The robot perception system based on human-machine collaboration according to claim 6, characterized in that, The distance between the virtual protective line and the edge of the goods is 5-10cm.