3D scanning LiDAR, LiDAR scanning methods and unmanned forklifts

Abstract

This invention discloses a 3D scanning lidar, a lidar scanning method, and an unmanned forklift. The method establishes a mathematical model of the relationship between rotational speed and the spatial angular distribution of a point cloud. It fixes the spindle rotational speed r1 and adjusts the high-speed rotational speed r2 to form a non-integer proportional motion, achieving non-repetitive scanning to generate a uniform point cloud. Based on the model, it calculates the point cloud uniformity index δS, generates a feature map, and selects the optimal rotational speed to support switching between uniform, progressive, or repetitive scanning modes. Furthermore, through pre-calculated gyro effect compensation and bidirectional laser arrangement, it forms a cross-shaped oblique grid, improving boundary recognition accuracy; adjusting the ratio of r1 to r2 controls the grid angle, achieving acute, right, or obtuse angle grids. This invention overcomes the problems of traditional radar resolution being limited by the number of modules, slow coverage speed, and complex control. Through open-loop control and dynamic adjustment, it reduces cost and power consumption, improves point cloud density and task adaptability, and is applicable to fields such as autonomous driving and robotics.

Description

Technical Field

[0001] This invention relates to the field of three-dimensional perception technology, and in particular to a 3D LiDAR scanning method for multiple scanning units, which aims to generate non-repeating point clouds suitable for applications such as autonomous driving, robotics, and mapping. Background Technology

[0002] In existing technologies, 3D LiDAR is widely used in fields such as robot navigation, obstacle avoidance, and target recognition. It outputs data by real-time detection of the three-dimensional coordinates of spatial targets and organizing them into a point cloud. Based on the point cloud organization, it can be divided into repetitive scanning and non-repetitive scanning. Repetitive scanning typically uses a spindle motor to drive multiple transmitting and receiving modules to rotate, forming a standard spatial grid coverage. For example, multiple light sources can complete the longitudinal or lateral field of view coverage, and a rotation mechanism can then fill the grid in other directions. For instance, Chinese patent CN113687387B discloses a LiDAR scanning device and method. This device includes a first laser emitter, a first galvanometer, a second laser emitter, and a second galvanometer. By controlling the first laser emitter to emit a first laser and controlling the rotation of the first galvanometer, the first laser scans the target object laterally. Simultaneously, the second laser emitter is controlled to emit a second laser and the second galvanometer is controlled to rotate, causing the second laser to scan the target object longitudinally. This method emphasizes achieving two-dimensional scanning through the coordinated use of two galvanometers, organizing the point cloud into a repetitive grid, thus improving scanning efficiency. This technology structurally relies on multiple laser emitters and galvanometer assemblies to achieve grid coverage in space. However, its resolution is limited by the number of transmitting / receiving modules; higher resolution requires more modules, leading to increased cost and power consumption. Currently, the highest line count on the market is only 128 lines, which is still lower than visual resolution. The scanning process of this patent relies on precise rotation control of the galvanometer, and its stability in highly dynamic environments depends on a closed-loop feedback mechanism. The algorithm is complex and consumes significant computational resources.

[0003] Non-repetitive scanning employs fewer transmit and receive modules. It controls the scanning mechanism to adjust the phase in two orthogonal directions in a pseudo-random manner, forming a non-repetitive point cloud. As the integration time increases, the coverage area expands, resulting in improved equivalent resolution. For example, Chinese patent CN114782651A discloses an automatic extrinsic parameter calibration method for a non-repetitive scanning 3D LiDAR and thermal camera. This method includes heating or cooling a calibration plate, aligning the LiDAR and camera with the calibration plate and moving them, acquiring multiple sets of point cloud and image data, extracting feature points from the calibration plate, and solving the extrinsic parameters through coordinate pairing and nonlinear optimization. This technology utilizes the non-repetitive scanning characteristic to achieve high-density coverage by accumulating multiple frames of point cloud data and considers the non-repetitive trajectory of the LiDAR during calibration. However, calibration relies on external auxiliary equipment and multiple sets of data acquisition, making it suitable for static calibration scenarios. This method emphasizes filling the preceding gaps with non-repetitive paths in point cloud generation to improve long-term resolution, but the short-term point cloud is sparse, with a repetition rate of less than 200kHz, resulting in a longer time to reach effective resolution. The calibration process of this patent involves multiple optimization calculations and relies on camera assistance, which increases the complexity of the system and results in low calibration efficiency in actual deployment.

[0004] Another type of technology focuses on compensating for scanning deviations to optimize point cloud quality. For example, international patent WO2021226763A1 discloses a lidar system based on non-integer speed ratio and closed-loop synchronous control. Through a master-slave control architecture, it provides real-time feedback of the position information of the driven scanning module, generates characteristic signals describing positional differences (i.e., phase errors), and adjusts the motion parameters of the driven module accordingly to accurately maintain a preset non-integer speed ratio, thereby achieving stable, non-repeating scanning. While this method can optimize the scanning trajectory, it relies on complex feedback loops and algorithms, consuming significant computational resources, increasing system cost and power consumption, and making it difficult to support dynamic adjustment in low-power scenarios. Furthermore, existing scanning point clouds mostly pass through objects in one direction, easily missing detections or being inaccurately measured at edges parallel to the scanning direction, failing to form ideal orthogonal or tilted intersecting grids, resulting in incomplete boundary recognition of small objects. The vertical field of view is generally limited to 0° to 90° or -15° to 30°, making it difficult to cover the top and bottom spaces of robots with height. These limitations restrict the application of LiDAR in smart devices, necessitating a method that simplifies control logic, reduces costs, and improves scanning flexibility.

[0005] Existing repetitive scanning technologies suffer from resolution limitations due to the number of modules, high cost, and high power consumption, hindering their development and failing to meet the high-precision interaction requirements of robots. While non-repetitive scanning can improve resolution through integration, its spatial coverage is slow. Furthermore, existing scanning point clouds often pass through objects in one direction, leading to missed detections or inaccurate measurements at edges parallel to the scanning direction. This prevents the formation of ideal orthogonal or inclined intersecting grids, resulting in incomplete boundary recognition of small objects. Vertical field of view is typically limited to 0° to 90° or -15° to 30°, making it difficult to cover the top and bottom spaces of robots with height. Existing control methods are mostly closed-loop mechanisms. While these can achieve precise trajectory optimization, they rely on complex feedback loops and algorithms, increasing system cost and power consumption, and making it difficult to support dynamic adjustment in low-power scenarios. These shortcomings limit the application of LiDAR in intelligent devices, necessitating a method that simplifies control logic, reduces costs, and improves scanning flexibility. Summary of the Invention

[0006] This invention provides a lidar scanning method that solves the problems of limited resolution due to the number of modules, slow coverage speed, inaccurate boundary recognition, and high cost caused by complex closed-loop control in the prior art.

[0007] To achieve the above objectives, the present invention provides the following technical solution:

[0008] A 3D lidar scanning method is applied to a 3D perception system comprising a first and a second actuator that can move relative to each other. The method generates a non-repeating 3D point cloud based on a non-integer proportional motion relationship between the two actuators. The method includes:

[0009] a) Establish a mathematical model describing the relationship between the rotational speeds of the first and second execution units and the final point cloud spatial angle distribution, wherein the first execution unit is a main axis execution unit with a fixed rotational speed r1, and the second execution unit is a high-speed execution unit with an adjustable rotational speed r2; based on the mathematical model, obtain a distribution array S by performing histogram statistics on the point cloud spatial angle distribution, and calculate the standard deviation δ of the distribution array S. s To quantify the spatial distribution uniformity of the point cloud;

[0010] b) For a series of different second execution unit motor speeds r2, repeat the calculation in step a) to generate the uniformity index δ. s A feature map showing how the motor speed r2 of the second execution unit changes;

[0011] c) Based on the expected point cloud shape, analyze the feature map and select a set of motor speeds that can realize the expected point cloud shape, wherein the expected point cloud shape includes uniform scanning point cloud, line-by-line scanning point cloud or repeated scanning point cloud.

[0012] d) Instruct the first execution unit and the second execution unit to operate at the motor speed corresponding to the optimal operating point, wherein the rotational speed r1 of the first execution unit remains unchanged, the expected point cloud shape is achieved by adjusting the rotational speed r2 of the second execution unit, and the rotational speed is supported to switch between different point cloud shapes during operation.

[0013] Furthermore, the rotation periods of the first execution unit and the second execution unit are calculated; a mathematical model is established based on the rotational speeds r1 and r2, the output angle is associated with the rotational speeds r1 and r2, and a sequence f(n) representing the angle distribution of all scan lines is generated by recursion; the sequence f(n) is subjected to histogram statistics to obtain the distribution array S, and the standard deviation δ of the distribution array S is calculated. s This is to quantify the spatial distribution uniformity of the point cloud.

[0014] Furthermore, the calculation of the rotation period of the first execution unit and the second execution unit specifically includes: calculating the rotation period of the first execution unit T1 = 60 / r1, where r1 is the spindle motor speed; calculating the scanning period of each prism of the second execution unit T2 = 60 / (nr2), where the number of prisms is the number of prism edges; in the mathematical model, the distribution angle of each prism in space is calculated as θ = mod(360(T2 / T1)*k, 360), where k is an integer index; and generating a sequence of all angle distributions f(n) by recursion based on a given initial phase; performing histogram statistics on the sequence f(n) to obtain a distribution array S, and calculating the standard deviation δ of the distribution array S. s This is to quantify the spatial distribution uniformity of the point cloud.

[0015] Furthermore, based on the rotational speed r2 value and the uniformity index δ s The feature map is generated, wherein the feature map includes multiple curves corresponding to different integration times. The curves are periodic waveforms, used to identify the rotation speed points where all curves coincide with the valley value to achieve uniform densification of the point cloud over time, or to identify the rotation speed points where the short integration curve is at the peak value and the long integration curve is at the valley value to achieve line-by-line scanning of the point cloud.

[0016] Furthermore, by adjusting the rotational speed r2 of the second execution unit, three scanning states are switched: repetitive scanning point cloud state, in which all scan lines overlap; uniform non-repetitive scanning point cloud state, in which the point cloud is evenly distributed and gradually densified; and line-by-line scanning point cloud state, in which the point cloud is filled line by line along a specific direction.

[0017] Furthermore, the method for the gyroscopic effect generated by the rotation of the second execution unit includes: calculating the ratio of the rotational inertia J1 of the driving part of the second execution unit to the rotational inertia J2 of the compensation part; adjusting the rotational speed rb of the compensation part in real time according to the ratio so that it satisfies ra / rb = J2 / J1 with the rotational speed ra of the driving part of the second execution unit; and correcting the rotational speed rb in real time by controlling the driving signal to counteract the gyroscopic effect.

[0018] Furthermore, the method includes controlling the scanning trajectory by adjusting the rotational speed ratio of the first execution unit and the second execution unit to generate a bidirectional intersecting diagonal grid to replace the unidirectional parallel diagonal scan, thereby scanning the object boundary from two directions; gradually refining the intersecting grid as the integration time increases, changing from a coarse grid state to a dense state, wherein the scan lines intersect to fill blank areas; scanning a square object, wherein the intersecting diagonal scan hits the side boundary of the object from two directions, and gradually densifying the point cloud coverage under non-repeating scan conditions to achieve complete capture of the object boundary.

[0019] Furthermore, the method includes controlling the grid intersection angle by adjusting the ratio of the rotation period T1 of the first execution unit to the effective scan period T2 of the second execution unit, wherein an acute-angled grid is generated when the T2 / T1 ratio is small, a right-angled grid is generated when the T2 / T1 ratio is moderate, and an obtuse-angled grid is generated when the T2 / T1 ratio is large.

[0020] Furthermore, step a) further includes: setting the scanning period T1 of the first execution unit and the rotation period T2 of the second execution unit, controlling T1 / T2 to be a non-integer ratio to ensure non-repeating scanning; setting the target non-integer ratio T1 / T2 = N ± 0 + ε, where N is a positive integer, 0 < ε < 1, and 0 is a small irrational number.

[0021] Furthermore, step c) further includes: analyzing the uniformity index δ of the feature map. s The curve determines the rotational speed r2 corresponding to the expected point cloud shape. When all curves coincide at the valley value, a uniform scanning point cloud is generated. When the short integration time curve is at the peak value and the long integration time curve is at the valley value, a line-by-line scanning point cloud is generated. When all curves coincide at the peak value, a repeating scanning point cloud is generated.

[0022] A 3D lidar includes:

[0023] A main scanning unit is used to perform a first scanning motion with first motion parameters;

[0024] An auxiliary scanning unit is used to perform a second scanning motion with second motion parameters;

[0025] A processor is electrically connected to the main scanning unit and the auxiliary scanning unit;

[0026] The processor is configured to: establish a mathematical model describing the relationship between the rotational speeds of the main scanning unit and the auxiliary scanning unit and the spatial angle distribution of the point cloud; control the first rotational speed r1 to remain constant; adjust the second rotational speed r2 to form a non-integer proportional motion relationship; based on the mathematical model, obtain a distribution array S by performing histogram statistics on the spatial angle distribution of the point cloud, and calculate the standard deviation δ of the distribution array S. s For a series of different second rotational speeds r2, the uniformity index δ was repeatedly calculated to obtain the uniformity index. s , generate δ s The feature map changes with r2; based on the expected point cloud shape, the feature map is analyzed to determine the combination of rotation speeds r1 and r2 corresponding to the target operating point, wherein the expected point cloud shape includes uniform scanning point cloud, line-by-line scanning point cloud, or repetitive scanning point cloud; the main scanning unit and the auxiliary scanning unit are instructed to operate according to the rotation speed corresponding to the target operating point, and the real-time adjustment of the second rotation speed r2 is supported to switch the point cloud shape.

[0027] Furthermore, the main scanning unit includes a first motor and a main reflector, and the auxiliary scanning unit includes a second motor and a multi-prism reflector.

[0028] Furthermore, it also includes an inertia compensation unit, which is coaxially arranged with the rotation axis of the auxiliary scanning unit. The processor controls the inertia compensation unit and the auxiliary scanning unit to move with the same angular velocity but in opposite rotation directions.

[0029] Furthermore, the inertia compensation unit includes a third motor and a counterweight.

[0030] Furthermore, the auxiliary scanning unit includes at least two laser transceiver modules, which are arranged to emit laser beams in opposite directions within the same scanning plane to generate a bidirectional intersecting diagonal grid; the processor controls the grid intersection angle to generate acute, right, or obtuse intersecting grids by adjusting the ratio of the first rotation speed r1 to the second rotation speed r2.

[0031] Furthermore, the laser transceiver module includes a laser emitting part and a laser echo receiving part. The laser emitting part and / or the laser echo receiving part are provided with at least one internal reflector for folding the optical path so that the laser emitting optical path or the echo receiving optical path is L-shaped or U-shaped.

[0032] Furthermore, the multi-prism reflecting prism has a hollow structure, the second motor is an external rotor motor, and its rotor is disposed inside the hollow multi-prism reflecting prism and coaxially connected to it; the main reflecting mirror is configured to support a spatial field of view range of -20° to 90° or -45° to 60° by adjusting its position.

[0033] Furthermore, the processor stores multiple preset scanning mode configuration files, each corresponding to a point cloud distribution characteristic. The processor can load different configuration files according to external instructions to switch scanning modes.

[0034] An unmanned forklift includes a vehicle body, a control system, and a 3D LiDAR for environmental perception, wherein the 3D LiDAR is any one of the 3D LiDARs described above, and the control system is configured as follows:

[0035] When performing large-scale path planning or autonomous navigation tasks, control the 3D LiDAR to operate in a globally uniform mode;

[0036] When performing tasks such as approaching shelves or avoiding dynamic obstacles, control the 3D LiDAR to switch to orthogonal obstacle avoidance mode;

[0037] When performing tasks such as cargo identification or fork alignment with pallet slots, control the 3D LiDAR to switch to local encryption mode.

[0038] This invention provides a 3D lidar scanning method, a 3D lidar, and an unmanned forklift, with the following advantages:

[0039] First, this invention establishes a specific non-integer proportional motion relationship between a fixed-rotational-speed spindle actuator (r1) and a dynamically adjustable-rotational-speed high-speed actuator, such as a polygonal prism, with a rotational speed (r2). This achieves efficient, non-repetitive scanning, effectively overcoming the bottleneck of traditional lidar resolution being limited by the number of physical laser modules. By keeping r1 constant and precisely adjusting r2, this method ensures that the scanning trajectory continuously and non-overlaps in the field of view over time. As integration time accumulates, the equivalent line count and spatial density of the point cloud continuously increase. This allows systems using fewer laser transceiver hardware to generate high-resolution, high-density point clouds comparable to or even surpassing those of high-line-count traditional lidar, fundamentally solving the problem that improving resolution in existing technologies requires increasing the number of hardware components, leading to a sharp increase in cost, power consumption, and size.

[0040] Secondly, this invention does not simply achieve non-repeating scanning, but rather quantifies and optimizes the uniformity of the point cloud by establishing a mathematical model that correlates the rotational speed relationship (r1 is constant, r2 is adjustable) with the final spatial angular distribution of the point cloud. Specifically, this method uses histogram statistics on the spatial angles of the scanned point cloud to obtain a distribution array, and calculates its standard deviation to quantify uniformity. The feature map generated based on this model clearly shows the pattern of point cloud uniformity index changes with the r2 rotational speed. By selecting the r2 rotational speed at the valley point in the corresponding map as the optimal working point, it ensures that the point cloud maintains a high degree of spatial distribution balance throughout the encryption process. This controlled and uniform encryption method, with fixed r1 and dynamically adjusted r2, avoids the local point cloud clustering or sparsity problems that may occur with random scanning, and can achieve effective, blind-spot-free coverage of the entire sensing area in a shorter time.

[0041] Furthermore, this invention employs an advanced open-loop control strategy. Unlike existing technologies that commonly rely on complex closed-loop feedback mechanisms, the control system of this invention is based on a preset mathematical model, using a fixed r1 and dynamically adjusted r2 for speed control, eliminating the need for expensive external sensors. Pre-compensation for physical disturbances is achieved, for example, by setting an inertia compensation unit (containing a third motor and counterweight) that rotates coaxially and counter-rotates with the high-speed actuator, actively counteracting the gyroscopic effect generated by high-speed rotation, thereby ensuring scanning stability at the physical level. This design eliminates complex real-time feedback loops and algorithms, significantly reducing the system's computational load and potential failure points, and enhancing overall reliability.

[0042] Another significant advantage resulting from this is the high cost-effectiveness of the equipment. By eliminating the precision feedback sensors and complex drive compensation devices required for closed-loop control, and employing a compact design such as a hollow multi-prism reflecting prism with an embedded external rotor motor, this invention significantly reduces the bill of materials and manufacturing costs of the lidar. Simultaneously, the laser transceiver module can achieve L-shaped or U-shaped optical paths by setting internal reflectors to fold the optical path, further reducing the device's size. The simplified system architecture also means lower operating power consumption, allowing for easy integration into application platforms with strict requirements on cost, energy consumption, and space, such as battery-powered unmanned forklifts, demonstrating broad market application prospects.

[0043] Furthermore, this invention endows LiDAR with unprecedented real-time adjustability and task adaptability. The processor can store multiple preset scanning mode configuration files corresponding to different point cloud distribution characteristics, and different configurations can be loaded according to external commands. By adjusting the r2 rotation speed in real time, seamless switching can be achieved between various modes without interrupting system operation, including uniform non-repetitive scanning selecting the valley value of all integral time curves, progressive scanning selecting the peak value at short-time curves, the valley value at long-time curves, and repetitive scanning selecting the peak value of all curves. This optimized perception strategy through dynamic adjustment of r2 enables devices such as unmanned forklifts to flexibly adjust point cloud and grid characteristics according to specific tasks such as large-scale navigation, obstacle avoidance, and cargo recognition, greatly improving operational efficiency and intelligence.

[0044] Finally, this invention significantly improves the ability to recognize object boundaries through a unique hardware layout and control method. By arranging at least two laser transceiver modules with opposite emission directions within the same scanning plane, combined with a control method that keeps r1 constant and r2 adjustable, the system can generate a bidirectional intersecting diagonal grid scanning trajectory. This intersecting grid ensures that any object, especially its contour boundary, can be detected by laser beams from two different directions. More importantly, by adjusting the rotational speed of r2 to change the rotational speed ratio between the first and second execution units, the intersection angle of the grid can be precisely controlled, generating acute, right, or obtuse angle grids to adapt to different detection targets. This method fundamentally solves the problem of missed detections caused by the scan line being parallel to the object edge during unidirectional scanning, enabling the capture of a more complete and accurate three-dimensional shape of the object, and significantly improving the reliability of the robot's autonomous obstacle avoidance and safe interaction. Attached Figure Description

[0045] Appendix Figure 1a : Logic diagram of the 3D LiDAR system framework in this invention;

[0046] Appendix Figure 1b : A schematic diagram of the structure of the 3D lidar in this invention;

[0047] Appendix Figure 2a : A schematic diagram of the scanning unit and inertia compensation unit in this invention;

[0048] Appendix Figure 2b : A schematic diagram of the scanning unit and inertia compensation unit in this invention;

[0049] Appendix Figure 2c : Schematic diagram of the rotational inertia relationship of the inertia compensation unit in this invention;

[0050] Appendix Figure 3 : Schematic diagram of the optical path folding of the laser transceiver module in this invention;

[0051] Appendix Figure 4a , 4b 4c: Point cloud distribution at time points t1, t2, and t3 in this invention;

[0052] Appendix Figure 5a , 5b 5c: In this invention, they respectively correspond to Figure 4a , 4b The projection of 4c onto a two-dimensional plane;

[0053] Appendix Figure 6 This demonstrates the uniformly tilted dense dot pattern exhibited by the 3D lidar scanning method in the generated angle sequence distribution;

[0054] Appendix Figure 7 :for Figure 6 Two-dimensional histogram statistical chart;

[0055] Appendix Figure 8 : This is a cluster of curves showing the change of the point cloud uniformity index δS with the rotational speed of the auxiliary execution unit in this invention;

[0056] Appendix Figure 9 :for Figure 8 A magnified view of a portion of the image;

[0057] Appendix Figure 10 : Flowchart of non-periodic disturbance generation in this invention;

[0058] Appendix Figure 11 : Flowchart of the grid structure adjustment amount generation in this invention;

[0059] Appendix Figure 12 : Schematic diagram of the dynamic adjustment principle of the intersection angle of the point cloud mesh in this invention. Detailed Implementation

[0060] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0061] To make the objectives, technical solutions, and advantages of the present invention clearer and more complete, the following will provide a comprehensive, in-depth, and detailed description of a 3D scanning lidar, a lidar scanning method, and an unmanned forklift using the lidar disclosed in the present invention, in conjunction with the accompanying drawings and specific embodiments.

[0062] 3D scanning laser weight:

[0063] Figure 1aThe system logic architecture diagram of the 3D LiDAR of this invention is shown macroscopically. The diagram mainly includes three core components: a main scanning unit 100, an auxiliary scanning unit 101, an inertia compensation unit 102, and a processor 103, which serves as the system's brain. The main scanning unit and the auxiliary scanning unit work together to scan the laser beam in three-dimensional space through relative motion. The inertia compensation unit is coaxial with the auxiliary scanning unit but rotates in opposite directions, used to counteract the gyroscopic effect generated during motion through pre-calculated static balance, thus ensuring scanning stability. The processor, as the control core, is electrically connected to each unit, responsible for executing the core scanning control algorithm, selecting the rotation speed based on a preset mathematical model and sending commands, and supporting dynamic adjustment to switch scanning modes.

[0064] Figure 1b A schematic diagram of a specific embodiment of the present invention is shown.

[0065] The main scanning unit 100 is one of the core moving components in the 3D LiDAR system of this invention, enabling spatial scanning. Its main function is to perform scanning motion in the first dimension according to preset first motion parameters. In a typical embodiment of this invention, the main scanning unit performs a stable, uniform rotational motion, specifically a 360° horizontal scan. This scanning method ensures broad coverage of the scanned area.

[0066] In this specific embodiment, the function of the main scanning unit 100 is realized by the horizontal turntable 2 and its driving mechanism. The horizontal turntable 2 structurally includes a spindle motor 21 and a horizontal turntable 22. The spindle motor 21 consists of a spindle motor stator 211 and a spindle motor rotor 212. In terms of installation orientation, the spindle motor 21 is placed vertically, and its spindle motor stator 211 is firmly fixed to the radar base 1. The spindle motor rotor 212 is connected to the horizontal turntable 22. Based on this configuration, the spindle motor 21, through the rotation of its rotor 212, can drive the horizontal turntable 22 to rotate stably in the horizontal direction, thereby completing the scanning task defined by the main scanning unit.

[0067] The auxiliary scanning unit 101 is another core moving component, responsible for performing a second-dimensional scanning motion that is orthogonal to or at a certain angle to the motion of the main scanning unit. This unit works in conjunction with the main scanning unit according to preset second motion parameters to jointly complete a comprehensive scan of the three-dimensional space. The core control method of this invention ensures that the system can cope with different working environment requirements by precisely adjusting the motion parameters of the auxiliary scanning unit, especially during dynamic adjustment and multi-mode switching.

[0068] In this specific embodiment, the function of the auxiliary scanning unit is implemented by the rotating mirror module 5. The rotating mirror module 5 structurally includes a high-speed motor bracket 51, a high-speed motor 52, and a reflecting prism 53. The high-speed motor 52 consists of a high-speed motor stator 521 and a high-speed motor rotor 522. In terms of installation orientation, the high-speed motor 52 is placed horizontally, and its high-speed motor stator 521 is firmly fixed to the horizontal turntable 22 via the high-speed motor bracket 51. In this preferred embodiment, the high-speed motor rotor 522 is an external rotor structure and is coaxially connected to the reflecting prism 53. Its function is to drive the reflecting prism 53 to rotate at high speed horizontally above the laser transceiver module 4, thereby completing the scanning task defined by the auxiliary scanning unit.

[0069] In addition to the two major motion units mentioned above, the 3D LiDAR also includes key optical components for realizing optical path transmission and reception. These components are carried on the horizontal turntable 22 of the main scanning unit and rotate accordingly.

[0070] Specifically, a laser transceiver module 4 and a multi-faceted reflector 3 are directly fixed to the horizontal turntable 22. In this preferred embodiment, they are arranged axially symmetrically on the horizontal turntable 22 to ensure dynamic balance during rotation. The laser transceiver module 4 is located in the central region of the horizontal turntable 22, while the two reflectors 3 are located on the two laser transceiver sides of the laser transceiver module 4, respectively. The function of the reflectors 3 is to reflect the laser beam emitted from the laser transceiver module 4 upwards at a preset angle to the reflecting surface of the reflecting prism 53 of the auxiliary scanning unit, and finally, through the further reflection of the reflecting prism 53, the scanning laser is directed into the external environment.

[0071] The processor 103 is the core control unit of the lidar system of this invention, responsible for the management and scheduling of the entire system. It establishes electrical connections with key units such as the main scanning unit and auxiliary scanning unit through an internal bus or interface circuit to ensure the coordinated operation of each unit. The processor's main responsibilities include:

[0072] The core scanning control algorithm disclosed in this invention is executed;

[0073] Based on a preset mathematical model, a non-integer motion speed ratio between the main scanning unit and the auxiliary scanning unit is set and maintained.

[0074] Load the pre-stored scan mode configuration file according to external instructions, and select the corresponding speed combination;

[0075] Send a fixed rotation speed command to control the operation of the main scanning unit and the auxiliary scanning unit;

[0076] It supports dynamic speed adjustment based on external commands to achieve switching between different scanning modes;

[0077] It interacts with upper-layer application systems to exchange data and commands.

[0078] Figure 2a , 2b Figure 2c details the core component of the multi-motor composite rotational angular momentum cancellation structure mounted on the horizontal turntable 22 in this embodiment of the invention. This structure aims to actively eliminate the gyroscopic effect generated by the composite rotation of the multi-motor system through pre-calculated static equilibrium, thereby ensuring the stability of the 3D lidar operation and the accuracy of measurements.

[0079] The structure mainly includes a rotating mirror module 5 and a set of counter-rotating modules, both of which are fixed to the horizontal turntable 22 by a motor bracket 51. The rotating mirror module 5 includes a high-speed motor 52 and a reflecting prism 53 for vertical scanning. However, this rotation, combined with the horizontal rotation of the spindle motor 21, produces a compound gyroscopic effect, which in turn causes equipment vibration.

[0080] To eliminate this vibration, one of the features of this solution is the inclusion of the reverse rotation module. This module includes a third motor 6 and a counterweight load 7. Its key structural feature is that the third motor 6 is coaxially arranged with the high-speed motor 52. The operating principle of the compensation structure is that the third motor 6 drives the counterweight load 7 to rotate in the opposite direction, thereby generating an angular momentum that compensates for the rotation of the high-speed motor 52.

[0081] Composite rotational angular momentum equilibrium equation:

[0082] ra·J1=rb·J2

[0083] in:

[0084] ra: is the configured speed of the high-speed motor 52;

[0085] J1: The total moment of inertia of the high-speed motor 52 and the reflective prism 53 fixed thereon;

[0086] rb: the rotational speed of the third motor 6;

[0087] J2: The total moment of inertia of the third motor 6 and the counterweight load 7 fixed on it.

[0088] During the design phase, the processor 103 sets the rotational speed of the third motor 6 to rb = (J1 / J2)·ra based on the pre-calculated rotational inertia ratio J1 / J2, and controls the operation of the counter-rotating module through a fixed drive signal. When this balance equation holds, the angular momentum generated by the two counter-rotating modules is equal in magnitude and opposite in direction, thus theoretically canceling each other out and ensuring the smooth operation of the lidar. Thanks to the stability of the high-precision motor and the pre-calibrated mechanical design, this invention can maintain long-term stability without real-time vibration monitoring or dynamic speed adjustment.

[0089] Figure 3 The schematic diagram visually illustrates the working process of the 3D LiDAR achieving vertical scanning in this embodiment of the invention. The diagram clearly reveals how the rotation of the rotating mirror module deflects the laser beam emitted from the laser transceiver module 4 to different elevation angles, thereby achieving scanning coverage of the vertical dimension of space.

[0090] The vertical scanning principle is achieved by the high-speed motor 522 in the rotating mirror module driving the reflecting prism 53 to rotate at high speed, causing a continuous and periodic change in the angle between the incident laser and different reflecting surfaces on the reflecting prism 53. According to the law of reflection of light, this change in angle directly leads to a continuous change in the exit angle of the laser beam, after being reflected by the reflecting prism 53 and finally entering the external environment, in the vertical plane, thereby achieving vertical scanning.

[0091] Figure 3 The specific radar scanning optical path process in this preferred embodiment is also illustrated in detail. The laser transceiver module 4 is mounted on the horizontal turntable 22 and emits a laser beam from its laser transceiver side. The laser beam is first reflected by the reflector 3 and then shines upward at a preset tilt angle onto a reflective surface of the reflective prism 53. The reflective prism 53 is supported by a high-speed motor bracket 51. Then, the reflective surface of the reflective prism 53 reflects the laser beam again, causing it to exit from the optics cover 6 of the lidar into the external environment.

[0092] A significant feature of this solution is that the laser transceiver module 4 has two laser transceiver sides, enabling it to synchronously emit and process two scanning lasers on different optical paths during operation. For example... Figure 3As shown, the two laser beams are reflected obliquely upwards by their respective corresponding reflectors 3, and then reflected again on two different cylindrical surfaces of the rotating reflecting prism 53 before finally exiting the radar. This embodiment features a precisely designed optical path for these two scanning laser beams, ensuring that after striking different cylindrical surfaces of the reflecting prism, the elevation angle of one laser beam entering the environment increases as the prism rotates, while the elevation angle of the other laser beam entering the environment decreases accordingly. In this way, the present invention achieves simultaneous scanning at both the top and bottom in the vertical direction, significantly improving the vertical scanning efficiency in the 3D scanning process.

[0093] 3D scanning LiDAR scanning method:

[0094] Based on the aforementioned structure of a 3D LiDAR in this embodiment, the present invention further discloses an innovative 3D LiDAR scanning method that complements it. The hardware structure provides a robust physical platform for implementing this method, while the method fully leverages the hardware's potential through open-loop control to generate high-quality 3D point cloud data.

[0095] The 3D scanning lidar scanning method in this embodiment is applied to a three-dimensional perception system including a first execution unit (spindle motor 21) and a second execution unit (high-speed motor 52) capable of relative motion. The core of this method lies in adjusting the rotational speeds of the spindle motor 21 and the high-speed motor 52 using a preset mathematical model, ensuring their speed ratio is a non-integer multiple. This achieves non-repetitive 3D scanning, resulting in a pseudo-random point cloud. The density of this point cloud continuously increases with scanning time, avoiding the drawbacks of traditional periodic scanning, such as point cloud overlap or persistent blank areas. The method includes establishing a mathematical model of the rotational speed relationship and the spatial angular distribution of the point cloud; fixing the rotational speed of the first execution unit; adjusting the rotational speed of the second execution unit to form a non-integer proportional motion; and performing histogram statistics on the point cloud angular distribution based on the model to obtain a distribution array S, and calculating its standard deviation δ. s Quantization uniformity; δ is repeatedly calculated and generated for different second execution unit rotation speeds. s Feature map that changes with rotational speed; select the optimal rotational speed based on the analysis map of the expected point cloud shape; the instruction execution unit runs at the optimal rotational speed, and supports real-time adjustment and shape switching at the second rotational speed.

[0096] Figure 4a , 4b Figure 4c visually illustrates the three-dimensional diagram showing the evolution of the non-repeating scan trajectory generated by the scanning method of the present invention over time. Figure 5a , 5b 5c corresponds to respectively Figure 4a , 4b4c is projected onto a two-dimensional plane. This aims to illustrate how the present invention, through a specific open-loop control strategy, avoids the drawbacks of traditional periodic scanning, thereby obtaining a three-dimensional point cloud whose point cloud density continuously and uniformly increases over time.

[0097] like Figure 4a As shown, this is a 3D plot of the point cloud at the minimum single-loop coverage time of the scan (t1: 33ms). Figure 5a for Figure 4a In a planar projection, the image presents a series of parallel, slightly slanted lines, forming a striped pattern with uniform spacing but noticeable gaps. The 3D image captures the volumetric curve scan path (suitable for comprehensive spatial encryption visualization such as that used by unmanned forklifts for object boundary detection), while the 2D projection simplifies it to a linear representation, emphasizing angular uniformity and the potential to increase density through integration time. The slant effect is particularly pronounced in the central region (x≈150–200), where the lines converge into a dense vertical band, reflecting the focal point of scan concentration, indicating low point cloud coverage, and the pseudo-random trajectory features generated by non-integer velocity ratios ensure subsequent gap filling.

[0098] At time t1, the point cloud generated by the 3D LiDAR exhibits a sparse distribution, resembling a low-density grid. This stage reflects the initial state of non-repetitive scanning. The main scanning unit 100 drives the horizontal turntable 22 at a fixed rotation speed to perform a 360° horizontal scan, while the auxiliary scanning unit 101 drives the reflecting prism 53 at a high speed to achieve rapid vertical scanning. Because the rotation speed ratio is set to a non-integer value, the angular distribution of the scan lines in space exhibits pseudo-random characteristics, avoiding the periodic overlap of traditional repetitive scanning.

[0099] The sparsity of the point cloud is determined by non-integer proportional motion, resulting in larger gaps between scan lines and lower coverage. The mathematical model generates an initial scan trajectory sequence by calculating the spatial distribution angles of each prism, and histogram statistics yield the distribution array S, with its standard deviation δ... s A higher value indicates lower point cloud uniformity.

[0100] This stage is suitable for rapidly covering large areas, such as the initial environmental perception of unmanned forklifts when performing large-scale path planning. The point cloud is sparse but sufficient to provide preliminary spatial contour information.

[0101] like Figure 4b As shown, this is a 3D plot of the point cloud at the midpoint of time (t2: 500ms). Figure 5b for Figure 4bIn the planar projection, the image shows interlaced, wavy lines with significantly reduced gaps, increased point cloud density, and improved coverage. The 3D image demonstrates further refinement of the spiral scanning path; the dynamic rotation speed adjustment of the auxiliary scanning unit optimizes spatial coverage. The wavy pattern in the 2D projection reflects the initial cross-effect of the bidirectional laser arrangement, enhancing boundary recognition capabilities. The standard deviation δS in this stage is smaller than t1, and uniformity is improved, making it suitable for medium-resolution scenarios, such as orthogonal obstacle avoidance modes when an unmanned forklift approaches a shelf or avoids dynamic obstacles.

[0102] The increase in point cloud density stems from the cumulative effect of non-integer proportional motion. The mathematical model recursively generates a new angular distribution sequence, and the histogram-based distribution array S shows a denser distribution with a standard deviation δ. s The decrease compared to time t1 indicates improved uniformity. The characteristic spectrum (the curve of δs versus rotational speed) can be used to select the optimal rotational speed point to optimize the filling effect.

[0103] This stage is suitable for scenarios requiring medium resolution, such as when an unmanned forklift approaches a shelf or avoids dynamic obstacles. Orthogonal obstacle avoidance mode (right-angle intersecting grid) enhances boundary recognition accuracy. Filling gaps in the point cloud improves the ability to capture object contours.

[0104] like Figure 4c As shown, this is a 3D plot of the point cloud at the final time (t3: 1500ms). Figure 5c for Figure 4c In a planar projection, the image presents a highly uniform, dense grid with no obvious gaps, achieving near-complete coverage. The 3D image shows complete spherical or ellipsoidal field coverage, with spiral lines forming complex intersecting patterns, while the grid structure in the 2D projection highlights the fine boundary capture capability of the bidirectional intersecting diagonal grid. The uniformity index δS in this stage approaches its minimum, reflecting the optimization of the high-density point cloud, making it particularly suitable for high-precision tasks, such as the localized densification mode used by unmanned forklifts for cargo recognition or fork alignment with pallet slots.

[0105] The uniformity and density of the point cloud are optimized by a mathematical model, with the standard deviation δ of the distribution array S. s The minimum value reflects the high uniformity of spatial distribution. The bidirectional intersecting diagonal grid (which controls the acute, right, or obtuse angles by adjusting the rotation speed ratio) further enhances the boundary recognition capability, making it particularly suitable for capturing the side boundaries of square objects.

[0106] This stage is suitable for high-precision tasks, such as the localized encryption mode of unmanned forklifts when identifying goods or aligning forks with pallet slots. High-density point clouds ensure complete capture of the three-dimensional shape of objects, improving operational efficiency and safety, especially in complex environments (such as dense shelving).

[0107] The processor 103, based on a preset mathematical model, sets and controls the speed ratio of the spindle motor 21 to the high-speed motor 52 to be a non-integer multiple. This is to ensure that after the spindle motor 21 drives the entire scanning head to rotate one revolution, the next scan line generated by the high-speed motor 52 driving the reflecting prism 53 will not completely overlap with the previous trajectory, but will precisely fill the previous gaps, thus achieving a non-repeating scanning effect where the point cloud density increases over time. This non-integer ratio is set as follows: the scanning cycle of the first execution unit and the rotation cycle of the second execution unit are preset to a non-integer ratio to ensure non-repeating scanning; a pre-calculated speed fluctuation correction value is used, and an open-loop control system based on a pre-calculated mathematical model is constructed according to the target non-integer ratio calculation method. This system uses a precise mathematical model as its target trajectory:

[0108]

[0109] In this model:

[0110] T1: This is the scanning cycle of the spindle motor 21, which is fixed and usually set to T1 = 60 / r1, where r1 is the spindle speed (unit: rpm). For example, when fixed at 3000 rpm, T1 = 0.02 seconds.

[0111] T2: This is the scanning cycle of the high-speed motor 52, which can be dynamically changed by adjusting the rotational speed. The calculation is T2 = 60 / (nr2), where n is the number of edges of the reflecting prism 53 (e.g., 4 or 6), and r2 is the high-speed rotational speed (range 5000-10000 rpm). For example, when n = 4 and r2 = 6123 rpm, T2 = 60 / (4 × 6123) = 0.0025 seconds.

[0112] N: As a positive integer, it represents the basic integer multiple relationship between two cycles. For example, N=4 means that the basic ratio is close to 4 times.

[0113] O: is a preset offset coefficient used to optimize the distribution of scan lines. It controls the position of the subsequent scan trajectory relative to the gap of the previous scan trajectory to improve the uniformity of coverage. It is usually set to a rational number between 0.1 and 0.5, for example, O = 0.314.

[0114] ε: A small adjustment to ensure long-term non-repeatability, used to counteract the instability of T1 and T2. It is pre-generated based on an irrational constant, such as ε = π / 100 ≈ 0.0314.

[0115] Figure 4a , 4bFigure 4c demonstrates the dynamic filling characteristic of the scan point cloud distribution at three consecutive time points t1, t2, and t3. At the initial time t1, the scan lines are relatively sparse. Since the speed ratio is preset to a non-integer value, at the subsequent time t2, new scan lines are guided into the gaps between previous scan lines. As time progresses to t3, the scan lines continuously increase the blank areas in the field of view, resulting in a significant and uniform increase in the coverage and density of the point cloud. Ultimately, this method generates a dense 3D point cloud that exhibits a pseudo-random distribution macroscopically and is microscopically controlled by a precise mathematical model. The point cloud generation process is as follows: pre-calculate the rotation cycles of the first and second execution units; correlate the output angles based on the model and generate a scan line angle sequence recursively; obtain S from the sequence histogram and calculate δ. s Quantitative uniformity; spatial distribution angle is

[0116]

[0117] The point cloud generation process is as follows: First, we pre-calculate the rotation period T1 of the first execution unit (spindle motor 21) (e.g., when r1 = 3000 rpm, T1 = 0.02 seconds) and the rotation period T2 of the second execution unit (high-speed motor 52) under a certain speed ratio (n is the number of prism edges, e.g., when n = 4 and r2 = 6000 rpm, T2 ≈ 0.0025 seconds); based on the model, we associate the output angle with the speed ratio, and generate the scan line angle sequence through recursion, for example, θk = θ0 + k·(360° / n)·(T1 / T2) mod 360°, where θ0 = 0° initial phase, k is the scan index (e.g., ... The first 20 k generate an angle sequence: f(n) [0.0, 73.644, 147.288, 220.932, 294.576, 8.22, 81.864, 155.508, 229.152, 302.796, 16.44, 90.084, 163.728, 237.372, 311.016, 24.66, 98.304, 171.948, 245.592, 319.236]); histogram statistics are performed on this sequence (e.g., dividing the 360° of space into 1° grids) to obtain the distribution array S, and its standard deviation δ is calculated. s To quantify uniformity.

[0118] like Figure 6 The figure shows the numerical distribution of the first 6140 angle sequences within 3000 ms of the aforementioned distribution sequence θk. The horizontal axis represents the ordinal number generated by the above method, and the vertical axis represents the corresponding scan line angle (0°-360°).

[0119] Histogram statistics were obtained from the array set. Figure 7This image is a two-dimensional histogram statistical chart, showing the point cloud angular distribution of a 3D LiDAR at a fixed rotation speed and within a 3000ms integration time. The vertical bars and the black dots at the top reflect density fluctuations. The distribution array S = [S1, S2, ..., sn] is generated by statistically analyzing the first 6140 angle sequences. The standard deviation δs quantifies uniformity: a small δS indicates uniform distribution and strong non-repetitiveness, while a large δS indicates uneven distribution and increased repetitiveness.

[0120] It represents the number of points containing the above θk sequence within a 360-degree unit divided into 1-degree intervals. The resulting array is denoted as S = [S1, S2, S3, ..., Sn]. The standard deviation of the array S is:

[0121]

[0122] Thus, we have obtained the point cloud distribution uniformity under a certain fixed rotational speed (r1, r2 determined) and a certain fixed time interval (integration time determined, 3000 ms). δ s The smaller the value, the better the uniformity and the greater the non-repetitiveness of the point cloud; conversely, the larger the value, the greater the repetitiveness.

[0123] Determine the rotational speed r1, change the rotational speed r2, and then plot the point cloud uniformity value (δ) based on different integration times. s The family of curves that vary with r2, such as Figure 8 As shown, selecting a working point where all curves in the graph are at their troughs will yield a point cloud with uniform density over time. Selecting a peak will produce a repetitive point cloud, while selecting short-term troughs and long-term peaks will produce a fast, line-by-line scan point cloud.

[0124] After magnification of the local area, as shown Figure 9 As shown, the curve details are clearer, and the boundary between peaks and valleys is more obvious. Short integration time curves (e.g., 100ms) show a sharp rise at the peak, reflecting the significant impact of speed fine-tuning on the early distribution, while long integration time curves (e.g., 3000ms) tend to flatten at the valley, indicating the stabilizing effect of time accumulation on the uniformity of the point cloud. The magnified view highlights key points within specific speed ranges. For example, the valley overlap region is suitable for the global uniform mode, the peak region supports the repetitive scan mode, and the transition region of short valleys and long peaks optimizes the line-by-line filling effect, further verifying the flexibility of the patented method in achieving multi-mode adaptation through dynamic adjustment of r2. Selecting a peak yields a repetitive point cloud, while selecting short-duration valleys and long-duration peaks yields a fast line-by-line scan point cloud.

[0125] To ensure scanning stability in highly dynamic operating environments, this embodiment includes a pre-calculated gyroscopic effect compensation step. Gyroscopic effects in a compound axis rotation system can cause equipment vibration, especially at high speeds (e.g., r2 > 8000 rpm), leading to point cloud distortion. Therefore, the processor 103, based on a pre-calculated moment of inertia ratio J1 / J2, sets the compensation rotation speed to satisfy ra / rb = J2 / J1, and controls the third motor 6 to counteract the effect via a fixed drive signal.

[0126] The combined rotation of the main shaft and high-speed shaft generates a pre-advance effect. This embodiment employs reverse rotation compensation: J1 is the inertia of the high-speed motor + prism, and J2 is the inertia of the compensation unit (third motor 6 + counterweight 7). The ratio J1 / J2 is pre-calculated (e.g., 1.2), then rb = 1.2ra. Open-loop control is achieved through a fixed signal, eliminating the need for real-time sensor fusion.

[0127] The advancement of this scanning method lies in its high flexibility and task adaptability, achieved through two pre-generated adjustment parameters: aperiodic perturbation and mesh structure adjustment. These adjustments are based on pre-computed mathematical models, ensuring that the system can dynamically optimize point cloud generation under different scenarios (such as unmanned forklift navigation, obstacle avoidance, and cargo recognition), improving coverage, uniformity, and task adaptability. The implementation mechanisms of these two adjustment parameters are described in detail below:

[0128] First, non-periodic perturbation: In order to break the potential long-term periodicity, the processor 103 pre-generates a set of non-periodic perturbation sequences through an algorithm before running, and stores them in the configuration file. Figure 10 The generation process is demonstrated: parameter initialization, setting the maximum perturbation amplitude (e.g., 0.01) and irrational constant (e.g., π), pre-calculating the perturbation value, and calculating frac for each scan cycle i. i =modf(i·π / 100)[0], where,

[0129] frac i The fractional part of the i-th scan period is used as the intermediate value for generating the perturbation sequence.

[0130] i: Scan cycle index, which is a positive integer (e.g., i = 1, 2, 3, ...), representing continuous scan iterations.

[0131] π / 100: A constant based on the irrational number π (approximately 3.14159), divided by 100 to obtain a smaller value.

[0132] i·π / 100: The product of the periodic index i and the constant π / 100, generating a linearly growing sequence with non-repeating fractional parts.

[0133] [0]: Extract the fractional part from the modf output and ignore the integer part.

[0134] The sequence is stored in processor 103 and loaded into the speed ratio model sequentially during runtime. The speed of high-speed motor 52 is adjusted using fixed instructions. A specific formula example is as follows:

[0135] For a period i = 1, frac1 = modf(0.0314) = 0.0314, disturb1 = 0.01·(2·0.0314-1) ≈ -0.00937

[0136] For a period i = 2, frac2 = modf(0.0628) = 0.0628, disturb2 ≈ -0.00874.

[0137] For a period i = 3, frac3 = modf(0.0942) = 0.0942, disturb3 ≈ -0.00812.

[0138] ...

[0139] In this embodiment, modf is a mathematical function.

[0140] This function is used to decompose a floating-point number into an integer part and a fractional part. It returns two values: the fractional part and the integer part.

[0141] Input: 0.0314 (i.e., π / 100 ≈ 0.0314)

[0142] Output: modf(0.0314) = (0.0314, 0)

[0143] Decimal part: 0.0314 (because the input value is less than 1, there is no integer part).

[0144] Integer part: 0.

[0145] frac1=modf(0.0314)[0]=0.0314

[0146] This means taking the fractional part of the input value 0.0314, which is 0.0314 itself. This step is to ensure that the perturbation value is based on the aperiodic nature of irrational numbers.

[0147] The generated disturbance sequence (e.g., [0.00937, 0.00874, 0.00812, ...]) is used to fine-tune the speed ratio. For example, the target ratio T1 / T2 = 4.314, and the actual ratio fluctuates to 4.32, resulting in a deviation Δ = 4.32 - 4.314 = 0.006. Combining this with the disturbance value disturb1 = -0.00937, the adjustment amount G = 0.5·Δ + disturb1 = 0.5·0.006 - 0.00937 = 0.003 - 0.00937 = -0.00637. Assuming the initial speed r2 = 6123 rpm, the proportional adjustment is converted into a speed change:

[0148] Assuming an initial speed r2 = 6123 rpm, proportional adjustment is converted into speed change:

[0149] Δr2=-r2·GT1 / T2=≈6123·0.001477≈0.904rpm

[0150] r2′=6123+0.904≈6123.904rpm

[0151] Simulation results show that the sample perturbation sequence is [0.00937, 0.00874, 0.00812, ...], which ensures that the proportional fluctuation is small but non-periodic, avoiding potential cycles such as those found in non-repeating patterns.

[0152] Secondly, the mesh structure adjustment amount: when the task requirements change, the processor 103 loads the pre-stored rotation speed ratio configuration file according to the external instructions and adjusts the ratio coefficient to change the scan line intersection angle α. Figure 11 The adjustment process is demonstrated: the grid structure adjustment amount controls the scan line intersection angle by dynamically adjusting the cycle ratio coefficient O to adapt to different task requirements (such as global navigation, obstacle avoidance, or cargo recognition for unmanned forklifts). Here, T1 is the rotation cycle of the first execution unit (spindle motor), T2 is the scanning cycle of each prism face of the second execution unit (high-speed motor), and the ratio T1 / T2 = N + O, where N is a positive integer and O is a small irrational number. The processor 103 stores multiple preset configuration files, each corresponding to a different grid type: acute-angled grid (high-density local refinement), right-angled grid (balanced coverage), or obtuse-angled grid (large-area coarse scanning). When an external command is triggered, the processor loads the configuration file, adjusts the high-speed motor speed r2, and changes the cycle ratio T1 / T2 to achieve the target intersection angle. The intersection angle α is determined by the laser beam optical path geometry, using the formula:

[0153] Where n is the period ratio and n is the number of prism edges. Bidirectional laser arrangement ( Figure 3A cross mesh is generated using two laser beams in opposite directions, ensuring that the object boundary is scanned from two directions and overcoming the missed detection problem of unidirectional parallel scanning. The cross angle varies with scale as follows:

[0154] Small-scale (e.g., T1 / T2 < 4) generation of acute-angled grids is suitable for high-density local scanning. The increased deflection frequency of the laser beam in the vertical direction results in denser intersections of scan lines in space, with a smaller angle α (acute angle). This is consistent with bidirectional laser arrangement (…). Figure 3 Relatedly, two laser beams in opposite directions form a sharper intersecting trajectory, which is suitable for local encryption.

[0155] A moderate ratio (e.g., T1 / T2≈4) generates a right-angled grid for balanced coverage. The laser beam's scanning frequency is balanced in both the vertical and horizontal directions, using a bidirectional laser beam ( Figure 3 This forms orthogonal trajectories. Compared to unidirectional parallel scanning, this allows for balanced coverage of object boundaries from both directions, reducing missed detections (such as shelf edges).

[0156] Large-scale (e.g., T1 / T2>4) generation of obtuse-angled grids is suitable for rapid, large-area scanning. The laser beam scanning frequency decreases in the vertical direction, resulting in a wider intersection angle in space. Bidirectional laser arrangement ( Figure 3 This allows the scan lines to intersect at a large angle, forming a coarse grid, which is suitable for quickly covering a large field of view (such as unmanned forklift navigation).

[0157] like Figure 12 The adjustment process in this embodiment is as follows:

[0158] Receive external commands and determine the target intersection angle. For example, an acute angle of 45° is used for local encryption (such as tray socket alignment), a right angle of 90° is used for obstacle avoidance, and an obtuse angle of 120° is used for large-scale navigation.

[0159] Locate the configuration file and select the corresponding cycle ratio.

[0160] For example, an acute-angled grid corresponds to T1 / T2 = 2.5, a right-angled grid corresponds to T1 / T2 ≈ 4, and an obtuse-angled grid corresponds to T1 / T2 = 5.5.

[0161] Calculate the new rotational speed r2. The period formula is:

[0162]

[0163] Example: If r1 = 3000 rpm, n = 4, and the target angle T1 / T2 = 2.5 (acute angle)

[0164] but:

[0165] The processor 103 instruction enables the high-speed motor 52 to operate at a new speed r2, supporting real-time switching between different grid types.

[0166] Based on a combination of the above methods, including non-integer proportional motion models, histogram statistical uniformity quantization, gyroscope effect compensation, non-periodic disturbance sequence generation, and bidirectional cross-grid control, the processor 103 can efficiently control the 3D LiDAR to achieve various advanced scanning modes. These modes are seamlessly switched through pre-computed mathematical models and open-loop control mechanisms, optimizing point cloud generation to adapt to dynamic environmental requirements. Figure 8 The demonstration showcases the scanning mode switching process, highlighting how the processor loads configuration files based on external instructions (such as task signals sent by the automated forklift control system), enabling a flexible transition from global coverage to local encryption. This mode switching not only improves the system's task adaptability but also reduces computational load and power consumption, supporting low-power devices such as battery-powered automated forklifts.

[0167] The processor 103, based on a preset mathematical model and feature map (a periodic waveform curve of δS varying with rotational speed), controls the high-speed motor 52 to rotate at speed r2, generating three main point cloud shapes: uniform scanning, line-by-line scanning, and repetitive scanning. Each mode corresponds to different integral time curve analysis.

[0168] Global Uniform Mode: The processor loads a uniform scan configuration file, selects all rotational speed points in the feature map where the integral time curves coincide with a valley value, and generates a point cloud that gradually densifies over time. The point cloud density increases uniformly, avoiding local clustering or sparseness, making it suitable for large-scale environmental perception. For example, when an unmanned forklift performs path planning, the uniform mode provides full field-of-view coverage, reducing δS to below 0.3.

[0169] Local encryption mode: Loads a local weighted perturbation configuration file to increase the point cloud density of the target region through non-periodic perturbation sequences (such as disturbances). i =0.01·(2·frac i -1)) Weighted rotation speed adjustment focuses on a specific area (such as a cargo pallet). This increases the point cloud density in the target area by 20%, making it suitable for cargo recognition or alignment tasks.

[0170] Line-by-line scanning mode: Select the rotational speed point (e.g., r2≈6200rpm) of the short integral curve peak (sparse filling) and the long integral curve valley (gradual uniformity), and fill the point cloud line by line along a specific direction.

[0171] Repeated scan mode: Select the rotational speed point where the curve coincides with the peak (e.g., r2≈6500rpm), the scan lines coincide, and a stable grid is generated, which is suitable for static calibration.

[0172] The configuration file stores specific control parameters, including speed values, integral time curves, and disturbance sequences. It can be quickly loaded according to external commands (such as task signals), reducing task response time to the millisecond level and significantly improving system efficiency.

[0173] In the application scenario of this invention on unmanned forklifts, a 3D LiDAR is installed on the vehicle body for environmental perception.

[0174] Global uniform mode: used for large-scale path planning and autonomous navigation, selects valley rotation speed to generate uniform point cloud;

[0175] Orthogonal obstacle avoidance mode: Used to approach shelves or avoid dynamic obstacles, switching to a right-angled grid profile;

[0176] Local encryption mode: Used for cargo identification or fork alignment with pallet jacks, loading a sharp-angled mesh profile to improve local resolution.

Claims

1. A 3D laser radar scanning method applied to a three-dimensional perception system comprising a first execution unit and a second execution unit capable of relative motion, the method is based on a non-integer proportional motion relationship between the two execution units to generate a non-repeated scanning three-dimensional point cloud, characterized in that, The method comprises: a) Calculate the rotation periods of the first execution unit and the second execution unit; establish a mathematical model describing the relationship between the rotational speeds of the first execution unit and the second execution unit and the final point cloud spatial angle distribution, wherein the first execution unit is the main axis execution unit, and its rotational speed... The second execution unit is a high-speed execution unit, and its rotational speed remains constant. Adjustable; based on the stated rotational speed and Establish a mathematical model to correlate the output angle with the rotational speed. and Correlation, and by recursion, a sequence representing the angular distribution of all scan lines is generated. ; for the sequence Histogram statistics are performed to obtain the distribution array S, and the standard deviation δ of the distribution array S is calculated. S To quantify the spatial distribution uniformity of the point cloud; b) for a range of different said second actuator motor speeds repeating the calculation of step a) thereby generating a profile showing the uniformity index δ S as a function of said second actuator motor speed ​ c) analyzing the feature map according to the expected point cloud shape, and selecting a corresponding set of motor speeds that can achieve the expected point cloud shape, wherein the expected point cloud shape includes a uniform scanning point cloud, a row-by-row scanning point cloud, or a repeated scanning point cloud; d) instructing the first and second execution units to operate at motor speeds corresponding to the optimal operating point, wherein the rotational speed of the first execution unit remains unchanged, and the rotational speed of the second execution unit is adjusted to achieve the intended point cloud shape, and to support real-time adjustment of the rotational speeds to switch between different point cloud shapes during operation.

2. The method of claim 1, wherein, The calculation of the rotation period of the first execution unit and the second execution unit specifically includes: calculating the rotation period of the first execution unit. ,in Calculate the spindle motor speed; calculate the scanning cycle of each prism of the second execution unit. Where n is the number of prism edges; in the mathematical model, the spatial distribution angle of each prism is calculated as... Where k is an integer index, all angular distribution sequences are recursively generated given an initial phase. ; for the sequence Histogram statistics are performed to obtain the distribution array S, and the standard deviation δ of the distribution array S is calculated. S This is to quantify the spatial distribution uniformity of the point cloud.

3. The method of claim 1, wherein, According to the rotation speed The value of the uniformity index δ S Generating the feature map, wherein the feature map comprises a plurality of curves corresponding to different integration times, the curves being periodic waveforms, for identifying a rotation speed point at which all curves coincide at a valley value to achieve a point cloud with uniform encryption over time, or a rotation speed point at which a short integration curve is at a peak value and a long integration curve is at a valley value to achieve a point cloud with line-by-line scanning.

4. The method of claim 1, wherein, by adjusting the rotational speed of the second execution unit Switching between three scanning states: a repeated scanning point cloud state, in which all scanning lines coincide; a uniform non-repeated scanning point cloud state, in which the point cloud is uniformly distributed and gradually encrypted; a row-by-row scanning point cloud state, in which the point cloud is filled row by row in a particular direction.

5. The method of claim 4, wherein, The method for the gyroscopic effect generated by the rotation of the second execution unit includes: calculating the ratio of the rotational inertia J1 of the driving part of the second execution unit to the rotational inertia J2 of the compensation part; and adjusting the rotational speed of the compensation part in real time according to the ratio. So that its rotational speed is similar to that of the second execution unit drive section. satisfy The rotational speed is corrected in real time by controlling the drive signal. To counteract the gyro effect.

6. The method of claim 1, wherein, The method comprises controlling the scanning trajectory by adjusting the speed ratio of the first execution unit and the second execution unit to generate a bidirectional cross-hatch grid, replacing the unidirectional parallel hatch scan, so as to scan the object boundary from two directions; gradually refining the cross-hatch grid as the integration time increases, from a coarse grid state to a dense state, wherein the scanning lines cross to fill the blank area; scanning a square object, wherein the cross-hatch scan hits the object side boundary from two directions, and gradually encrypts the point cloud coverage under the condition of non-repeated scanning, so as to achieve complete capture of the object boundary.

7. The method of claim 6, wherein, The method includes adjusting the rotation period of the first execution unit. Effective scan cycle of the second execution unit The proportional control grid intersection angle, where when When the scale is small, it generates acute-angled intersecting meshes. When the proportions are appropriate, a right-angled intersecting grid is generated. When the scale is large, obtuse-angled intersecting meshes are generated.

8. The method of claim 1, wherein, Step c) further includes: analyzing the uniformity index δ of the feature map. S The curve determines the rotational speed corresponding to the expected point cloud shape. When selecting the rotational speed at which all curves coincide with the valley value, When generating a uniform scanning point cloud, the rotation speed is selected such that the short integration time curve is at its peak and the long integration time curve is at its trough. The point cloud is generated line by line, and the rotational speed at which all curves coincide with the peak value is selected. Repeatedly scanned point clouds are generated.

9. A 3D laser radar applying the 3D laser radar scanning method of claim 1, comprising: a main scanning unit for performing first scanning motion with first motion parameters; an auxiliary scanning unit for performing second scanning motion with second motion parameters; a processor electrically connected with the main scanning unit and the auxiliary scanning unit; The processor is characterized in that it is configured to: establish a mathematical model describing the relationship between the rotational speeds of the main scanning unit and the auxiliary scanning unit and the spatial angular distribution of the point cloud, and control the first rotational speed.

1. Keep the speed constant and adjust the second speed. To form a non-integer proportional motion relationship; based on the mathematical model, a distribution array S is obtained by performing histogram statistics on the spatial angle distribution of the point cloud, and the standard deviation δ of the distribution array S is calculated. S For a series of different second speeds The uniformity index δ was obtained by repeated calculations. S , generate δ S Follow The changing feature map; based on the expected point cloud shape, the feature map is analyzed to determine the rotational speed corresponding to the target operating point. 1 and The combination, wherein the expected point cloud shape includes a uniformly scanned point cloud, a line-by-line scanned point cloud, or a repetitive scanned point cloud; instructs the main scanning unit and the auxiliary scanning unit to operate at the rotational speed corresponding to the target operating point, and supports the second rotational speed. Real-time adjustments are made to switch the point cloud shape.

10. The 3D lidar of claim 9, wherein, the main scanning unit comprises a first motor and a main mirror, and the auxiliary scanning unit comprises a second motor and a multi-prism reflecting prism.

11. The 3D lidar of claim 9, wherein, Further comprising an inertia compensation unit coaxially arranged with the rotation axis of the auxiliary scanning unit, and the processor controls the inertia compensation unit and the auxiliary scanning unit to move at the same angular velocity and in opposite rotation directions.

12. The 3D lidar of claim 11, wherein, The inertia compensation unit comprises a third motor and a counterweight.

13. The 3D lidar of claim 9, wherein, The auxiliary scanning unit comprises at least two laser transceiver modules arranged in a manner that the laser beams emitted in the same scanning plane are in opposite directions, generating a two-way cross oblique line grid; the processor controls the grid intersection angle to generate an acute angle, a right angle or an obtuse angle intersection grid by adjusting the ratio of the first rotation speed to the second rotation speed .

14. The 3D lidar of claim 13, wherein, The laser transceiver module comprises a laser emitting part and a laser echo receiving part, and at least one internal mirror is arranged inside the laser emitting part and / or the laser echo receiving part to fold the optical path, so that the laser emitting path or the echo receiving path is in L shape or U shape.

15. The 3D lidar of claim 10, wherein, The multi-prism reflecting prism is a hollow structure, and the second motor is an external rotor motor, the rotor of which is arranged inside the hollow multi-prism reflecting prism and coaxially connected therewith; the main mirror is configured to support a spatial field of view angle range of -20° to 90° or -45° to 60° through position adjustment.

16. The 3D lidar of claim 9, wherein, The processor stores a plurality of preset scanning mode configuration files, each configuration file corresponding to a point cloud distribution characteristic, and the processor can load different configuration files to switch the scanning mode according to external instructions.

17. An unmanned fork truck comprising a truck body, a control system, and a 3D lidar for environmental perception, characterized in that, The 3D laser radar is the 3D laser radar of any one of claims 9-16, and the control system is configured to: when performing a large-range path planning or autonomous navigation task, control the 3D laser radar to operate in a global uniform mode; when performing a task of approaching a shelf or avoiding a dynamic obstacle, control the 3D laser radar to switch to an orthogonal obstacle avoidance mode; when performing a task of identifying goods or aligning a fork with a tray insertion hole, control the 3D laser radar to switch to a local encryption mode.

Images