CN113204242A - Reconfigurable unmanned vehicle three-section type butt joint control method - Google Patents

Abstract

Translated fromChinese

本发明提供一种可重构无人车三段式对接控制方法，将无人车单元的对接过程分为远端接近阶段、近端捕获阶段和柔性对接阶段三个阶段，能使无人车单元在复杂地面环境下迅速实现自主动态对接。其中在远端接近阶段，采用考虑转向模式切换的远端接近轨迹实时规划算法，以规划出所需时间短的接近轨迹，提高对接效率；在近端捕获阶段，确定由远端接近阶段向柔性对接阶段的切换时机，解决了可重构无人车对接过程中阶段切换确定困难的问题，显著提高了可重构无人车的对接效率。在柔性对接阶段，基于六自由度柔性对接机构设计了包含视觉传感器、激光测距传感器、力传感器等多传感器感知系统的柔性对接过程，提高了柔性对接阶段的准确性和稳定性。

The invention provides a three-stage docking control method for a reconfigurable unmanned vehicle, which divides the docking process of the unmanned vehicle unit into three stages: a remote approach stage, a near-end capture stage and a flexible docking stage, so that the unmanned vehicle can be The unit quickly realizes autonomous dynamic docking in complex ground environment. Among them, in the far-end approaching stage, a real-time planning algorithm of the far-end approaching trajectory considering the steering mode switching is adopted to plan the approaching trajectory with a short time required to improve the docking efficiency; The switching timing of the docking stage solves the problem of difficulty in determining the stage switching during the docking process of the reconfigurable unmanned vehicle, and significantly improves the docking efficiency of the reconfigurable unmanned vehicle. In the flexible docking stage, a flexible docking process including vision sensors, laser ranging sensors, force sensors and other multi-sensor sensing systems is designed based on the six-degree-of-freedom flexible docking mechanism, which improves the accuracy and stability of the flexible docking stage.

Description

Reconfigurable unmanned vehicle three-section type butt joint control method

Technical Field

The invention relates to an unmanned vehicle reconstruction method, in particular to a reconfigurable unmanned vehicle three-section type butt joint control method, and belongs to the technical field of unmanned vehicles.

Background

The unmanned vehicle can independently execute functional tasks such as logistics, transportation, distribution, patrol, public transportation, retail, cleaning, connection, rescue and the like, and is a core element for future intelligent transportation and smart city construction. It is expected that most tasks will be completed by unmanned vehicles instead of human beings in future transportation and travel and human life, and vehicles will be evolved from traditional vehicles into intelligent carriers for performing functional tasks, and have great influence on the development of human society. Compared with the traditional intelligent networked automobile, the unmanned automobile aims at executing functional tasks, does not have a human driving mechanism, subverts the basic design concept of the traditional automobile centered on human, and has innovative and flexible configuration, and revolutionary changes of basic characteristics such as system architecture and the like. Therefore, the fundamental theory and key technology of the unmanned vehicle must realize original breakthrough, is a brand new challenge brought by the era of intelligent vehicles, and is a research hotspot in the international and domestic fields.

With the continuous expansion of the connotation of intelligent transportation and smart cities in the future, the development of unmanned vehicles faces major challenges of complex and variable execution tasks, three-dimensional and multidimensional running environments, continuous expansion of functional requirements, single limitation of carrier configuration and the like. Obviously, the traditional unmanned vehicle with a fixed configuration has difficulty in meeting the challenges and cannot meet the requirements of the intelligent transportation and the smart city for a novel intelligent vehicle in the future. The reconfigurable unmanned vehicle technology thoroughly breaks through the form constraint of the traditional fixed configuration unmanned vehicle, can independently realize complex functions such as function reconfiguration, topology reconfiguration and the like, realizes independent combination, butt joint and disintegration among multiple unmanned vehicle units, comprehensively expands the function task execution boundary of the unmanned vehicle, and is expected to become a subversive innovation technology in the future. How to enable the unmanned vehicle units to be accurately butted under the complex ground environment is a key technology which needs to be solved firstly by the reconfigurable unmanned vehicle.

Disclosure of Invention

In view of the above, the invention provides a reconfigurable unmanned vehicle three-stage docking control method, which divides a docking process of an unmanned vehicle unit into a far-end approaching stage, a near-end capturing stage and a flexible docking stage, and can enable the unmanned vehicle unit to rapidly realize autonomous dynamic docking under a complex ground environment.

The reconfigurable unmanned vehicle three-section type butt joint control method specifically comprises the following steps:

the reconfigurable unmanned vehicle is provided with more than two unmanned vehicle units; the reconstruction of the unmanned vehicle is realized by more than two unmanned vehicle units through butt joint;

each unmanned vehicle unit is provided with a docking mechanism for realizing docking, each docking mechanism comprises a movable end and a fixed end, and during docking, the movable end of the docking mechanism on one unmanned vehicle unit is docked with the fixed end of the docking mechanism on the other unmanned vehicle unit; when the two unmanned vehicle units are butted, the unmanned vehicle unit for providing the movable end of the butting structure is an active butting vehicle, and the unmanned vehicle unit for providing the fixed end of the butting structure is a passive butting vehicle;

the three-section topology reconstruction method divides the butt joint process of two unmanned vehicle units into three stages, which are respectively as follows: a far-end approaching stage, a near-end capturing stage and a butt joint stage;

after receiving a docking instruction, the two unmanned vehicle units enter a far-end approach stage, and in the far-end approach stage, the two unmanned vehicle units are converged to a set target position; when two unmanned vehicle units travel to a set distance at intervals, entering a near-end capturing stage;

a near-end capturing stage, wherein the active docking car judges the switching time from a far-end approaching stage to a docking stage in real time by taking the motion range of the movable end of the docking mechanism as a constraint condition, and enters the docking stage when the motion range of the movable end of the docking mechanism meets the set constraint condition;

and in the docking stage, the active docking vehicle controls the movable end of the docking mechanism to be docked with the fixed end of the docking mechanism of the passive docking vehicle, so that the two unmanned vehicle units complete topology reconstruction.

As a preferred mode of the present invention, in the far-end approach phase, a far-end approach trajectory real-time planning algorithm considering steering mode switching is adopted to calculate the far-end approach trajectory:

the docking instruction comprises a set target position, and the unmanned vehicle unit receiving the docking instruction firstly obtains a shortest path under the algorithm as an initial approaching track through a track planning algorithm; then optimizing the initial approaching track by selecting a steering mode to obtain a far-end approaching track;

the steering mode is selected according to different working conditions, namely the modes of double-axle steering, crab steering and pivot steering of the unmanned vehicle unit adopting the independent steering technology are selected: wherein the dual axle steering mode is suitable for long distance and long time working conditions; the crab-type steering mode is suitable for the working condition of quick lane change; the pivot steering mode is suitable for the turning working condition in a narrow area.

As a preferred mode of the present invention, in the remote approach stage, in a process that two unmanned vehicle units approach to a set target position, the active docking vehicle calculates a distance between itself and the passive docking vehicle in real time:

if the distance between the two unmanned vehicle units reaches a set distance value before the target position is reached, sequentially entering a near-end capturing stage and a butt joint stage, and moving towards the target position after the butt joint is completed;

if one unmanned vehicle unit reaches the target position first, the unmanned vehicle unit stops at the target position, when the other unmanned vehicle unit runs to the position with the set distance from the unmanned vehicle unit, the unmanned vehicle unit enters a near-end capturing stage and a butt-joint stage in sequence, and butt-joint is completed at the set target position.

As a preferred aspect of the present invention, in the near-end capturing phase, the determination process of the active docking car for switching the timing from the far-end approaching phase to the docking phase is as follows:

the active butt-joint vehicle firstly carries out attitude judgment, and the constraint conditions of the attitude judgment are as follows:

-γ_Dlim≤γ_C≤+γ_Dlim

wherein:γ_Ccapturing a self course angle in a judging coordinate system for the active docking car at the near end;γ_Dlimthe limit direction-seeking angle is an included angle between a butt joint plane of the passive butt joint vehicle and a transverse plane of the active butt joint vehicle; the near-end capturing and distinguishing coordinate system takes the positioning center of the passive docking vehicle as the referenceThe passive butt-joint vehicle is a coordinate system with the longitudinal direction in the x direction and the transverse direction in the y direction;

if the course angle of the active docking vehicle meets the constraint condition, entering position judgment; if not, the active opposite-direction receiving vehicle carries out course adjustment until the course angle of the active opposite-direction receiving vehicle meets the constraint condition;

the constraint conditions for the position judgment are as follows:

-X_lim-X_i1-L_r-X_i2-L_fcos(γ_C)≤X_C≤+X_lim-X_i1-L_r-X_i2-L_fcos(γ_C)

-Y_lim-L_r-X_i2-L_fsin(γ_C)≤Y_C≤+Y_lim-L_r-X_i2-L_fsin(γ_C)

wherein: (X_C，Y_C) Capturing the coordinates of the positioning center of the active docking car in the near-end judging coordinate system;X_lim，Y_limthe longitudinal and transverse limit movement distances of the movable end of the butt joint structure on the active butt joint vehicle are set;X_i1the longitudinal length of the movable end of the docking mechanism on the active docking car is defined;X_i2the longitudinal length of the fixed end of the docking mechanism on the passive docking car is shown;L_f，L_rthe distances from the front end surface and the rear end surface of the body of the passive butt joint vehicle to the positioning center of the body of the passive butt joint vehicle are respectively;

if the position of the active docking car meets the constraint condition of the position judgment, entering a docking stage; and if not, performing attitude adjustment on the active butt joint vehicle until the positioning center of the active butt joint vehicle meets the constraint condition of the position judgment.

In a preferred mode of the invention, in the docking stage, the movable end of the docking mechanism on the active docking vehicle and the fixed end of the docking mechanism on the passive docking vehicle are in flexible docking; namely, the butt joint mechanism is a flexible butt joint structure;

the flexible docking mechanism includes: the device comprises an active capture module, a locking module, a sensing module and a control module; the active capture module adopts a six-degree-of-freedom platform, the fixed end of the six-degree-of-freedom platform is fixedly connected with the unmanned vehicle unit, and the movable end of the six-degree-of-freedom platform is provided with a locking core; the six-degree-of-freedom platform can drive the locking core to move along the transverse direction, the longitudinal direction, the vertical direction, the yaw direction, the rolling direction and the pitching direction so as to adjust the position and the posture of the locking core;

the locking module includes: a locking mechanism and a docking guide block; the butt joint guide block is fixedly connected with the unmanned vehicle unit; the butt joint guide block is provided with a butt joint guide hole matched with the locking core and used for accommodating the locking core; the locking mechanism is used for locking the position of the butted guide block and the locking core after being butted;

the sensing module is used for sensing the position and the posture of the locking core on the active capture module relative to the butt joint guide block on the locking module and sending the position and the posture to the control module; the control module controls the active capture module to adjust the position and the posture of the locking core relative to the butt joint guide block according to the sensing information of the sensing module, so that the locking core is inserted into the butt joint guide hole of the butt joint guide block when two unmanned vehicle units are in butt joint.

As a preferred mode of the present invention, the sensing module comprises a vision sensor mounted at the fixed end of the six-degree-of-freedom platform and more than two laser ranging sensors mounted at the end face of the movable end of the six-degree-of-freedom platform; the vision sensor and the more than two laser ranging sensors are respectively connected with the control module and used for sending detected signals to the control module;

an image recognition positioning plate matched with the vision sensor is arranged on the butt joint guide block, and the vision sensor obtains the position of the butt joint guide block relative to the locking core through recognition of the image recognition positioning plate;

the butt joint guide block is provided with a laser sensor detection board used for being matched with the laser ranging sensors, more than two laser ranging sensors are distributed at intervals along the circumferential direction, the control module obtains an included angle between the axis of the locking core and the axis of the butt joint guide block according to distance information between the detection board of the laser sensor and the distance information detected by the more than two laser ranging sensors respectively, and the control module adjusts the posture of the locking core so that the butt joint guide block is coaxial with the locking core.

As a preferred mode of the present invention, when two unmanned vehicle units are docked, the vision sensor on the active capture module of the active docking vehicle acquires the position information of the image recognition positioning plate on the locking module of the passive docking vehicle and feeds the position information back to the control module, and the control module adjusts the position of the locking core at first step according to the position information transmitted by the vision sensor, so that the relative position of the locking core and the docking guide block meets the set docking position requirement;

after preliminary adjustment, the control module on the active butt joint vehicle adjusts the posture of the locking core to eliminate the calculated included angle between the axis of the locking core and the axis of the butt joint guide block according to the distance information between the detection plate of the laser sensor on the locking module and the detection information detected by the laser ranging sensor, so that the axes of the locking core and the butt joint guide block are overlapped;

then the control module on the active docking vehicle controls the active capturing module to insert the locking core into the docking guide block; and finally, locking the active capture module and the locking module through a locking mechanism.

As a preferable mode of the present invention, the sensing module further includes two or more force sensors; more than two force sensors are arranged on the end face of the movable end of the six-degree-of-freedom platform and are distributed at intervals along the circumferential direction; the force sensor is connected with the control module;

when the two unmanned vehicle units are in butt joint, the force sensor is in contact with the butt joint surface of the butt joint guide block, and the stress of the butt joint surface of the butt joint guide block and the butt joint surface of the locking core is fed back to the control module.

Has the advantages that:

the docking control method divides the docking process of the unmanned vehicle unit into a far-end approaching stage, a near-end capturing stage and a flexible docking stage: in the far-end approaching stage, two unmanned vehicle units to be butted autonomously travel along the planned approaching track until the distance between the two vehicles is set, and a far-end approaching track real-time planning algorithm considering steering mode switching is provided aiming at the stage so as to plan the approaching track with short required time and improve the butting efficiency.

In a near-end capturing stage, a judging method for determining the switching time from a far-end approaching stage to a flexible docking stage in the reconfigurable unmanned vehicle topology reconfiguration process is provided, the judging method solves the problem that the stage switching is difficult to determine in the reconfigurable unmanned vehicle docking process, and the docking efficiency of the reconfigurable unmanned vehicle is remarkably improved.

In the flexible docking stage, a six-degree-of-freedom flexible docking mechanism is adopted, and a flexible docking process comprising a vision sensor, a laser ranging sensor, a force sensor and other multi-sensor sensing systems is designed on the basis of the six-degree-of-freedom flexible docking mechanism, so that the accuracy and the stability of the flexible docking stage are improved.

Drawings

FIG. 1 is a schematic diagram illustrating the adjustment of the attitude of an actively docked vehicle during a near-end capture phase;

FIG. 2 is a schematic structural diagram of an active capture module of the flexible docking mechanism employed in the docking stage;

fig. 3 is a schematic structural diagram of a flexible docking mechanism locking module adopted in a docking stage.

Wherein: the device comprises avision sensor 1, an electrically drivenlinear actuator 2, abase 3, alaser ranging sensor 4, aforce sensor 5, alocking core 6, alocking pin actuator 7, a buttjoint guide block 8, an imagerecognition positioning plate 9, a lasersensor detection plate 10 and a lockingcore connecting plate 11.

Detailed Description

The invention is described in detail below by way of example with reference to the accompanying drawings.

Example 1:

in order to solve the problem of accurate control of the reconfigurable unmanned vehicle topology reconfiguration docking process, the embodiment provides a reconfigurable unmanned vehicle three-section type docking control method, which can meet the complex control requirement of the reconfigurable unmanned vehicle autonomous topology reconfiguration.

The reconfigurable unmanned vehicle is provided with more than two unmanned vehicle units, each unmanned vehicle unit is an unmanned vehicle with two wheels, and the two wheels have independent steering functions. When more than two unmanned vehicle units are required to work together, the more than two unmanned vehicle units are in end-to-end butt joint according to actual use requirements, and reconstruction of the unmanned vehicle is achieved. Each unmanned vehicle unit is provided with a docking mechanism for realizing docking, the docking mechanism comprises a movable end and a fixed end, the movable end of the docking mechanism is arranged at the front end of the unmanned vehicle unit, and the fixed end is arranged at the rear end of the unmanned vehicle unit; when in butt joint, the movable end of the butt joint mechanism on one unmanned vehicle unit is in butt joint with the fixed end of the butt joint mechanism on the other unmanned vehicle unit.

For convenience of description, when the two unmanned vehicle units are butted, the unmanned vehicle unit positioned at the rear for providing the movable end of the butting mechanism is an active butting vehicle, and the unmanned vehicle unit positioned at the front for providing the fixed end of the butting mechanism is a passive butting vehicle. After receiving an external docking instruction, the two unmanned vehicle units are converged to a specified position and are docked, and the three-section topology reconstruction method is used for controlling the docking process and ensuring accurate docking of the two unmanned vehicle units.

The three-section topology reconstruction method divides the butt joint process of two unmanned vehicle units into three stages, which are respectively as follows: the remote end is close to stage, near-end and is caught stage and butt joint stage, through the control to three stages, realizes the accurate control to unmanned vehicle unit butt joint process.

After the two unmanned vehicle units receive the docking instruction, the remote approach stage is started:

the two unmanned vehicle units receive the docking instruction, the docking instruction includes a set target position (and an active docking vehicle and a passive docking vehicle are designated in the docking instruction), and the two unmanned vehicle units receiving the docking instruction are converged with the target position which is quickly approached by track tracking through track planning. And aiming at the stage, a far-end approaching track real-time planning algorithm considering steering mode switching is provided, and two unmanned vehicle units plan a far-end approaching track with shorter required time according to the algorithm. The specific implementation process of the remote approach trajectory real-time planning algorithm comprises the following steps:

the unmanned vehicle unit receiving the docking instruction firstly obtains a shortest path under a conventional path planning algorithm (such as an A star algorithm) as an initial approach path; and then optimizing the initial approach track by selecting a steering mode to obtain a shorter approach time path, and taking the path as a far-end approach track.

The selection of the steering mode is to select the double-axle steering mode, the crab-type steering mode and the pivot steering mode of the unmanned vehicle unit adopting the independent steering technology according to different working conditions (namely, the steering mode is switched): the double-axle steering mode has strong stability and is suitable for working conditions of long distance, long time and the like; the crab steering mode can change the position of the vehicle under the condition of not changing the direction of the vehicle head, and is suitable for working conditions such as rapid lane change and the like; the pivot steering can change the direction of the vehicle head under the condition of not changing the position of the vehicle, and is suitable for working conditions such as head dropping in narrow areas. Selecting different steering modes for different operating conditions may reduce the time required for the distal end approach procedure.

And after the unmanned vehicle unit obtains the far-end approaching track considering the switching of the steering mode, tracking the track according to the obtained far-end approaching track, and further quickly approaching the set target position.

In the approaching process, the active docking vehicle obtains the position coordinates of the passive docking vehicle in real time according to workshop communication, so as to obtain the distance between the active docking vehicle and the passive docking vehicle, if the distance between two unmanned vehicle units before reaching the target position reaches a set distance value, the active docking vehicle enters a near-end capturing stage and a docking stage in sequence, and moves to the target position after docking is completed; if one unmanned vehicle unit reaches the target position first, the unmanned vehicle unit stops at the target position, when the other unmanned vehicle unit runs to the position with the set distance from the unmanned vehicle unit, the unmanned vehicle unit enters a near-end capturing stage and a butt-joint stage in sequence, and butt-joint is completed at the set target position.

And the active docking vehicle in the near-end capturing stage adopts a near-end capturing and judging method to determine the switching time from the far-end approaching stage to the docking stage in the topology reconstruction process. The near-end capturing discrimination method comprises the following steps: and when the two judgment conditions are simultaneously met, the two unmanned vehicle units to be butted can complete the switching from the far-end approaching stage to the butting stage, namely the two unmanned vehicle units to be butted reach the butting time, and the movable end and the fixed end of the butting mechanism can be butted.

And a near-end capturing stage, wherein the butt joint opportunity of the butt joint vehicle is actively judged in real time, and the judgment is to obtain the switching opportunity from the far-end approaching stage to the butt joint stage by taking the motion range of the movable end of the butt joint mechanism as a constraint condition in a near-end capturing and judging coordinate system. The near-end capturing and distinguishing coordinate system is established by taking a positioning center of the passive docking car (the positioning center is a set position, generally a centroid position of the passive docking car) as an origin of coordinates, wherein the x direction of the near-end capturing and distinguishing coordinate system is the longitudinal direction of the passive docking car, and the y direction of the near-end capturing and distinguishing coordinate system is the transverse direction of the passive docking car. The external docking instruction received by the active docking vehicle comprises absolute coordinates of a passive docking vehicle positioning center needing to be docked with the active docking vehicle in a world reference system.

During the approach process of the active docking vehicle to the passive docking vehicle, firstly, the attitude judgment is carried out, and the active docking vehicle acquires the self course angle in the near-end capturing and judging coordinate system through sensors such as an inertial sensor (IMU) and a GPS (global positioning system)γ_CDefining the included angle between the butt-joint plane of the passive butt-joint vehicle and the transverse plane of the active butt-joint vehicleγ_DFor the direction-finding angle, as shown in FIG. 1, in the near-end capturing discrimination coordinate system, the heading angle of the vehicle itself is actively dockedγ_CAngle with direction findingγ_DEqual, the movable end of the docking mechanism is required to be in the docking range and have a direction-finding angleγ_DThe set attitude constraint condition is required to be met, and the heading angle of the active docking vehicle isγ_CAngle with direction findingγ_DEquality, i.e. need to judgeγ_CWhether the following set posture constraint conditions are satisfied:

-γ_Dlim≤γ_C≤+γ_Dlim

wherein:γ_Dlimis the limit steering angle.

If the vehicle is actively docked, the self course angle of the vehicleγ_CIf the attitude constraint condition is met, entering position judgment; if not, the heading of the butt joint vehicle is actively adjusted until the heading angle of the butt joint vehicle meets the attitude constraint condition.

After the active docking vehicle completes the attitude judgment, based on the course angle meeting the attitude constraint conditionγ_CAnd (4) judging the position:

the active docking vehicle obtains the absolute coordinates of the self vehicle and the passive docking vehicle under a world reference system through GPS and vehicle-to-vehicle communication, and determines the positioning center coordinate of the active docking vehicle in a near-end capturing and distinguishing coordinate system taking the positioning center of the passive docking vehicle as the origin of coordinates through the relative position relationship of the two vehicles (the step (a)X_C，Y_C) (typically the active docking vehicle centroid coordinates). To activate the docking mechanismThe positioning center coordinate of the active butt-joint vehicle needs to be judged when the moving end is in the butt-joint range (X_C，Y_C) Whether the following position constraint conditions are satisfied:

-X_lim-X_i1-L_r-X_i2-L_fcos(γ_C)≤X_C≤+X_lim-X_i1-L_r-X_i2-L_fcos(γ_C)

-Y_lim-L_r-X_i2-L_fsin(γ_C)≤Y_C≤+Y_lim-L_r-X_i2-L_fsin(γ_C)

wherein:X_lim，Y_limrespectively representing the longitudinal and transverse limit movement distances of the movable end of the butt joint structure on the active butt joint vehicle;X_i1indicating dockingThe longitudinal length of the movable end of the mechanism,X_i2the longitudinal length of the fixed end of the docking mechanism is shown;L_f，L_rrespectively, the distances from the front end face and the rear end face of the passive docking station to the positioning center thereof (the distances do not include the length of the docking mechanism), and in fig. 1, (b) isX_D1,Y_D1) Showing the coordinates of the movable end of the docking mechanism and the center of the connecting end face of the vehicle body of the active docking mechanism in a near-end capturing and judging coordinate system (a)X_D2,Y_D2) And the coordinates of the fixed end of the docking mechanism and the center of the connecting end face of the vehicle body of the passive docking vehicle in the near-end capturing and distinguishing coordinate system are represented.

The position constraint conditions form a position envelope area of the active docking vehicle, if the positioning center of the active docking vehicle is within an envelope range, the docking condition is met, and a docking stage can be entered; if the position of the vehicle is not in the envelope curve, the vehicle posture of the vehicle is actively adjusted until the positioning center coordinate of the vehicle meets the position constraint condition.

And when the active docking vehicle meets the docking condition, entering a docking stage, and controlling the active docking vehicle to control the movable end of the docking mechanism to be in high-precision docking with the fixed end of the docking mechanism of the passive docking vehicle so as to enable the two unmanned vehicle units to complete topology reconstruction.

Example 2:

on the basis of theembodiment 1, in the docking stage, the movable end of the docking mechanism on the active docking vehicle and the fixed end of the docking mechanism on the passive docking vehicle are subjected to high-precision flexible docking.

On hardware, a six-degree-of-freedom flexible docking mechanism shown in fig. 2 and 3 is adopted; in software, the flexible docking process is controlled based on a multi-sensor sensing system such as a visual sensor, a laser ranging sensor and a force sensor, so that the accuracy and the stability of the flexible docking stage are improved.

Specifically, the method comprises the following steps: flexible docking mechanism includes: the device comprises an active capture module, a locking module, a sensing module and a control module; the active capturing module is a movable end of the docking mechanism, and the locking module is a fixed end of the docking mechanism.

As shown in fig. 2, the active capture module includes: an electrically drivenlinear actuator 2, abase 3 and alock core 6; the active capture module adopts a six-degree-of-freedom platform, thebase 3 is used as a fixed end of the six-degree-of-freedom platform, and thebase 3 is fixedly connected with a vehicle body of the unmanned vehicle unit; thelocking core 6 is fixed in the middle of the lockingcore connecting plate 11, and three groups of pin holes are uniformly distributed on the outer circumferential surface of thelocking core 6 at intervals along the circumferential direction.

Every two six electric-drivenlinear actuators 2 form a group, three groups of electric-drivenlinear actuators 2 are uniformly distributed on thebase 3 at intervals along the circumferential direction, and the other ends of the two electric-drivenlinear actuators 2 in each group are respectively hinged with the lockingcore connecting plate 11; namely, the fixed end of the electric drivelinear actuator 2 is hinged with thebase 3, and the actuating end is hinged with the lockingcore connecting plate 11. And a lockingcore connecting plate 11 connected with alocking core 6 is used as a movable end of the six-degree-of-freedom platform. By controlling the extension and retraction of the six electric-drivenlinear actuators 2, the postures of the active capture module in the transverse, longitudinal, vertical, yaw, roll and pitch directions can be adjusted.

When the flexible docking mechanism is docked, the control module controls the six electric-drivenlinear actuators 2 to move according to the expected position, so that the motion of the movable end of the platform in six freedom directions (transverse, longitudinal, vertical, yaw, roll and pitch) in a Cartesian coordinate system is realized, and finally thelocking core 6 on the movable end of the platform is dynamically controlled to be aligned with thedocking guide block 8 on the locking module in a high-precision mode, so that the docking action is completed.

The sensing module is arranged on the active capture module and comprises avision sensor 1 arranged on abase 3, threelaser ranging sensors 4 and threeforce sensors 5 arranged on a connecting plate of alocking core 6; wherein thevision sensor 1 is positioned right above thebase 3, and the image acquisition direction of thevision sensor 1 faces to the movable end of the six-degree-of-freedom platform; the threelaser ranging sensors 4 are uniformly distributed at intervals along the circumferential direction of the lockingcore connecting plate 11; the threeforce sensors 5 are arranged on the end face of the end of the lockingcore connecting plate 11 where thelocking core 6 is located and are uniformly distributed at intervals along the circumferential direction; preferably, the threeforce sensors 5 and the three laserdistance measuring sensors 4 are offset with respect to one another. And each sensor in the sensing module is respectively connected with the control module and used for sending the detected signal to the control module.

As shown in fig. 3, the locking module includes: the device comprises a locking mechanism, an imagerecognition positioning plate 9, a buttjoint guide block 8 and a lasersensor detection plate 10; wherein the buttjoint guide block 8 is fixedly connected with the vehicle body of the unmanned vehicle unit through a bracket; thedocking guide block 8 is centrally provided with a docking guide hole for cooperating with the lockingcore 6 for accommodating thelocking core 6. The lasersensor detection plate 10 is arranged on the outer circumference of the middle part of the buttjoint guide block 8, and divides the buttjoint guide block 8 into two parts along the axial direction, wherein one part is used for butt joint with the active capture module, and the other part is used for installing a locking mechanism.

The locking mechanism is used for realizing the position locking after the butt joint of the buttjoint guide block 8 and thelocking core 6, and adopts a locking pin and alocking pin actuator 7. Specifically, three lockingpin actuators 7 are uniformly distributed on the outer circumference of the buttjoint guide block 8 at intervals along the circumferential direction, the actuating end of each lockingpin actuator 7 is provided with locking pins which are in one-to-one correspondence with pin holes on thelocking core 6, and in order to ensure that the locking pins can be smoothly inserted into the corresponding pin holes, a spring is arranged inside each locking pin; initially, thelocking pin actuator 7 pulls the locking pin to compress the spring, so that the spring is in a compressed state and the locking pin is not pushed out; after thelocking core 6 enters the butt joint guide hole in the buttjoint guide block 8, thelocking pin actuator 7 releases force, the lockingcore 6 is rotated through the six-degree-of-freedom platform, when thelocking core 6 rotates to the pin hole and corresponds to the locking pin in position, the locking pin automatically extends out under the action of the restoring force of the spring and enters the pin hole, and therefore locking between the buttjoint guide block 8 and thelocking core 6 is achieved. An imagerecognition positioning plate 9 is connected to one of thelocking pin actuators 7; preferably, the imagerecognition positioning plate 9 is located at a position right above thedocking guide block 8, and corresponds to the position of the vision sensor on thebase 3.

The imagerecognition positioning plate 9 is used for being matched with thevision sensor 1, and thevision sensor 1 can obtain the relative position information of the imagerecognition positioning plate 9 on the unmanned vehicle unit where the locking module is located based on a position area recognition algorithm and an edge line recognition algorithm and sends the relative position information to the control module; the control module adjusts the position of thelocking core 6 on the six-degree-of-freedom platform according to the position, so that the lockingcore 6 and the buttjoint guide block 8 reach an expected relative position, and the accurate butt joint requirement is met.

The lasersensor detection plate 10 is used for being matched with the threelaser ranging sensors 4; when the active capture module is in butt joint with the locking module, the control module establishes a two-plane parallel mathematical model according to distance information between the threelaser ranging sensors 4 and the lasersensor detection plate 10 on the locking module, calculates an included angle between the axis of thelocking core 6 and the axis of theguide block 5, then controls thelocking core 6 on the six-freedom-degree platform to move to eliminate the included angle, enables the buttjoint guide block 8 and thelocking core 6 to be coaxial, and ensures that the lockingcore 6 can be accurately inserted into the buttjoint guide block 8 during butt joint.

In addition, during butt joint, theforce sensors 5 are in contact with the plane where the buttjoint guide block 8 on the locking module is located, the stress between the plane where the buttjoint guide block 8 is located and the plane where thelocking core 6 is located is fed back to the control module, and whether the two planes are parallel or not is judged according to the stress (if the two planes are parallel, the stress at the positions where the threeforce sensors 5 are located are the same). Meanwhile, a threshold value of the stress detected by theforce sensors 5 is preset in the control module, and the threshold value indicates that the lockingcore 6 and the buttjoint guide block 8 are in butt joint in place, namely when the stress fed back by the threeforce sensors 5 reaches the preset threshold value, the lockingcore 6 is inserted to reach a specified position. In addition,force sensor 5 still is used for detecting the stress sudden change that leads to because the unequally disturbance of ground when the butt joint, and when the sudden change appears in stress, control module in time controls thelocking core 6 of six degrees of freedom platforms and adjusts, and the rocking that the unequally disturbance of ground arouses when avoiding the butt joint leads to the mechanism to damage.

The multi-sensor sensing module based on the vision sensor, thelaser ranging sensor 4 and theforce sensor 5 can ensure the accuracy and stability of the butt joint process. When the butt joint is started, thevision sensor 1 acquires the position information of the imagerecognition positioning plate 9 and feeds the position information back to the control module, and the relative position of the flexible butt joint mechanisms of the two unmanned vehicle units to be butt jointed is adjusted to initially meet the requirement required by the flexible butt joint; then, thelaser ranging sensor 4 acquires distance information of the active capture module and the locking module of the two unmanned vehicle units to be butted, so that the movable end (the active capture module) and the fixed end (the locking module) of the flexible butting mechanism of the two unmanned vehicle units are kept parallel; when the movable end and the fixed end are aligned, the lockingcore 6 on the active capture module is slowly inserted into the buttjoint guide block 8 of the locking module, and in the process, theforce sensor 5 acquires stress information between the active capture module and the locking module during butt joint, so that deviation and collision caused by road jolt can be avoided during flexible butt joint.

In summary, the above description is only a preferred embodiment of the present invention, and is not intended to limit the scope of the present invention. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims (8)

Translated fromChinese

1.可重构无人车三段式对接控制方法，其特征在于，所述可重构无人车具有两个以上无人车单元；两个以上所述无人车单元通过对接实现无人车的重构；1. A three-stage docking control method for a reconfigurable unmanned vehicle, characterized in that the reconfigurable unmanned vehicle has more than two unmanned vehicle units; two or more of the unmanned vehicle units are connected to realize unmanned reconstruction of the car;每个无人车单元上均设置有用于实现对接的对接机构，所述对接机构包括活动端和固定端，对接时，其中一个无人车单元上对接机构的活动端与另一个无人车单元上对接机构的固定端对接；令两个无人车单元对接时，用于提供对接结构活动端的无人车单元为主动对接车，用于提供对接结构固定端的无人车单元为被动对接车；Each unmanned vehicle unit is provided with a docking mechanism for docking. The docking mechanism includes a movable end and a fixed end. When docking, the movable end of the docking mechanism on one unmanned vehicle unit is connected to the other unmanned vehicle unit. The fixed end of the upper docking mechanism is docked; when two unmanned vehicle units are docked, the unmanned vehicle unit used to provide the active end of the docking structure is an active docking vehicle, and the unmanned vehicle unit used to provide the fixed end of the docking structure is a passive docking vehicle;所述三段式对接控制方法将两个无人车单元的对接过程分为三个阶段，分别为：远端接近阶段、近端捕捉阶段及对接阶段；The three-stage docking control method divides the docking process of the two unmanned vehicle units into three stages, which are: a distal approach stage, a proximal capture stage, and a docking stage;当两辆无人车单元收到对接指令后，进入远端接近阶段；在所述远端接近阶段，两辆无人车单元向设定的目标位置交汇；当两辆无人车单元行进至间隔设定距离时，进入近端捕捉阶段；When the two unmanned vehicle units receive the docking command, they enter the remote approach stage; in the remote approach phase, the two unmanned vehicle units converge to the set target position; when the two unmanned vehicle units travel to the When the interval is set at a distance, it enters the near-end capture stage;近端捕捉阶段，所述主动对接车以对接机构活动端的运动范围为约束条件，实时进行由远端接近阶段向对接阶段切换时机的判断，当所述对接机构活动端的运动范围满足设定的约束条件时，进入对接阶段；In the near-end capture stage, the active docking vehicle takes the movement range of the active end of the docking mechanism as a constraint, and judges the timing of switching from the distal approach stage to the docking stage in real time. When the movement range of the active end of the docking mechanism meets the set constraints When conditions are met, enter the docking stage;对接阶段，所述主动对接车控制其对接机构活动端与被动对接车对接机构固定端进行对接，使两个无人车单元完成拓扑重构。In the docking stage, the active docking vehicle controls the active end of the docking mechanism to dock with the fixed end of the docking mechanism of the passive docking vehicle, so that the two unmanned vehicle units complete topology reconstruction.2.如权利要求1所述的可重构无人车三段式对接控制方法，其特征在于，在所述远端接近阶段，采用考虑转向模式切换的远端接近轨迹实时规划算法计算远端接近轨迹：2. The three-stage docking control method for reconfigurable unmanned vehicles as claimed in claim 1, characterized in that, in the remote approach stage, a remote approach trajectory real-time planning algorithm considering steering mode switching is used to calculate the remote Approach track:所述对接指令中包含了设定的目标位置，收到对接指令的无人车单元首先通过轨迹规划算法，得到一条该算法下的最短路径作为初始接近轨迹；然后再通过转向模式的选用对所述初始接近轨迹进行优化，得到远端接近轨迹；The docking instruction contains the set target position, and the unmanned vehicle unit that receives the docking instruction first obtains a shortest path under the algorithm as the initial approach trajectory through the trajectory planning algorithm; The initial approach trajectory is optimized to obtain the remote approach trajectory;所述转向模式的选用是根据不同工况对采用独立转向技术的无人车单元的双桥转向、蟹型转向及原地转向模式进行选用：其中所述双桥转向模式适用于长距离和长时间工况；所述蟹型转向模式适用于快速换道工况；所述原地转向模式适用于狭窄区域掉头工况。The selection of the steering mode is to select the dual-axle steering, crab steering and in-situ steering modes of the unmanned vehicle unit using the independent steering technology according to different working conditions: wherein the dual-axle steering mode is suitable for long-distance and long-distance steering. time conditions; the crab-shaped steering mode is suitable for fast lane changing conditions; the in-situ steering mode is suitable for U-turn conditions in narrow areas.3.如权利要求1所述的可重构无人车三段式对接控制方法，其特征在于，在所述远端接近阶段，两辆无人车单元向设定的目标位置接近过程中，所述主动对接车实时计算自身与被动对接车之间的距离：3 . The three-stage docking control method for reconfigurable unmanned vehicles according to claim 1 , wherein, in the remote approach stage, during the approach of two unmanned vehicle units to the set target position, 3 . The active docking vehicle calculates the distance between itself and the passive docking vehicle in real time:若在到达目标位置前两辆无人车单元之间的距离已达到设定距离值，则依次进入近端捕捉阶段和对接阶段，在完成对接后一起向目标位置行进；If the distance between the two unmanned vehicle units has reached the set distance value before reaching the target position, then enter the near-end capture stage and the docking stage in turn, and travel to the target position together after the docking is completed;若其中一辆无人车单元先达到目标位置，则停止在目标位置，当另一辆无人车单元行驶至与该无人车单元设定距离位置处时，依次进入近端捕捉阶段和对接阶段，在设定的目标位置处完成对接。If one of the unmanned vehicle units reaches the target position first, it will stop at the target position. When the other unmanned vehicle unit travels to the set distance from the unmanned vehicle unit, it will enter the near-end capture stage and docking in turn. stage, the docking is completed at the set target location.4.如权利要求1所述的可重构无人车三段式对接控制方法，其特征在于，所述近端捕捉阶段，所述主动对接车进行由远端接近阶段向对接阶段切换时机的判断过程为：4 . The three-stage docking control method for reconfigurable unmanned vehicles according to claim 1 , wherein, in the near-end capture stage, the active docking vehicle performs a switching timing from the far-end approach stage to the docking stage. 5 . The judgment process is:所述主动对接车首先进行姿态判断，所述姿态判断的约束条件为：The active docking vehicle first performs attitude judgment, and the constraints of the attitude judgment are:-γ_Dlim≤γ_C≤+γ_Dlim-γ_Dlim ≤γ_C ≤+γ_Dlim其中：γ_C为所述主动对接车在近端捕获判别坐标系中的自身航向角；γ_Dlim为极限寻向角，寻向角是指被动对接车对接平面与主动对接车横向平面之间的夹角；所述近端捕获判别坐标系指以被动对接车定位中心为坐标原点，被动对接车的纵向为x向为，横向为y向的坐标系；Where:γ_C is the own heading angle of the active docking vehicle in the near-end capture and discriminant coordinate system;γ_Dlim is the limit search angle, which refers to the distance between the docking plane of the passive docking vehicle and the lateral plane of the active docking vehicle. Included angle; the near-end capture discrimination coordinate system refers to a coordinate system with the positioning center of the passive docking vehicle as the coordinate origin, the longitudinal direction of the passive docking vehicle is the x-direction, and the lateral direction is the y-direction;若所述主动对接车自身航向角满足上述约束条件，则进入位置判断；若不满足，则所述主动对接车进行航向调整，直至自身航向角满足上述约束条件；If the heading angle of the active docking vehicle satisfies the above constraint conditions, enter the position judgment; if not, the actively docking vehicle will adjust the heading until its own heading angle satisfies the above constraint conditions;所述位置判断的约束条件为：The constraints of the location judgment are:-X-X_limlim-X-X_i1i1-L-L_rr-X-X_i2i2-L-L_ffcos(γcos(γ_CC)≤X)≤X_CC≤+X≤+X_limlim-X-X_i1i1-L-L_rr-X-X_i2i2-L-L_ffcos(γcos(γ_CC))-Y-Y_limlim-L-L_rr-X-X_i2i2-L-L_ffsin(γsin(γ_CC)≤Y)≤Y_CC≤+Y≤+Y_limlim-L-L_rr-X-X_i2i2-L-L_ffsin(γsin(γ_CC))其中：（X_C，Y_C）为在所述近端捕获判别坐标系中所述主动对接车的定位中心坐标；X_lim，Y_lim为所述主动对接车上对接结构活动端纵向和横向极限运动距离；X_i1为所述主动对接车上对接机构活动端的纵向长度；X_i2为所述被动对接车上对接机构固定端的纵向长度；L_f，L_r分别为所述被动对接车车身前端面和后端面距其定位中心的距离；Wherein: (X_C ,Y_C ) are the coordinates of the positioning center of the active docking vehicle in the near-end capture and discrimination coordinate system;X_lim ,Y_lim are the longitudinal and lateral limits of the active end of the docking structure on the active docking vehicle Movement distance;X_i1 is the longitudinal length of the active end of the docking mechanism on the active docking vehicle;X_i2 is the longitudinal length of the fixed end of the docking mechanism on the passive docking vehicle;L_f ,L_r are the front end surfaces of the passive docking vehicle body respectively and the distance from the rear end face to its positioning center;若所述主动对接车位置满足上述位置判断的约束条件，则进入对接阶段；若不满足，则所述主动对接车进行姿态调整，直至所述主动对接车的定位中心满足上述位置判断的约束条件。If the position of the actively docked vehicle satisfies the above constraint conditions for position determination, the docking stage is entered; if not, the actively docked vehicle performs attitude adjustment until the positioning center of the actively docked vehicle satisfies the above constraint conditions for position determination .5.如权利要求1所述的可重构无人车三段式对接控制方法，其特征在于，所述对接阶段，主动对接车上对接机构活动端与被动对接车上对接机构固定端进行柔性对接；即所述对接机构为柔性对接机构；5 . The three-stage docking control method for reconfigurable unmanned vehicles according to claim 1 , wherein, in the docking stage, the active end of the docking mechanism on the active docking vehicle and the fixed end of the docking mechanism on the passive docking vehicle are flexible. 6 . Docking; that is, the docking mechanism is a flexible docking mechanism;所述柔性对接机构包括：主动捕捉模块、锁定模块、传感模块及控制模块；所述主动捕捉模块采用六自由度平台，所述六自由度平台的固定端与无人车单元固接，活动端设置有锁定芯；所述六自由度平台能够带动所述锁定芯沿横向、纵向、垂向、横摆、滚转及俯仰方向运动，以调整所述锁定芯的位置和姿态；The flexible docking mechanism includes: an active capture module, a locking module, a sensing module and a control module; the active capture module adopts a six-degree-of-freedom platform, and the fixed end of the six-degree-of-freedom platform is fixedly connected to the unmanned vehicle unit, and is movable. The end is provided with a locking core; the six-degree-of-freedom platform can drive the locking core to move in the lateral, longitudinal, vertical, yaw, roll and pitch directions to adjust the position and attitude of the locking core;所述锁定模块包括：锁定机构和对接导向块；所述对接导向块与无人车单元固接；所述对接导向块上设置有用于和所述锁定芯配合的对接导向孔，用于容纳所述锁定芯；所述锁定机构用于所述对接导向块和锁定芯对接后的位置锁定；The locking module includes: a locking mechanism and a docking guide block; the docking guide block is fixedly connected to the unmanned vehicle unit; the docking guide block is provided with a docking guide hole for cooperating with the locking core, for accommodating all the the locking core; the locking mechanism is used to lock the position of the docking guide block and the locking core after docking;所述传感模块用于感知所述主动捕捉模块上锁定芯相对锁定模块上对接导向块的位置和姿态，并发送给所述控制模块；所述控制模块依据所述传感模块的感知信息控制所述主动捕捉模块调整所述锁定芯相对所述对接导向块的位置和姿态，使两辆无人车单元对接时，所述锁定芯插入所述对接导向块的对接导向孔中。The sensing module is used to sense the position and posture of the locking core on the active capture module relative to the docking guide block on the locking module, and send it to the control module; the control module controls the sensor according to the sensing information of the sensing module. The active capture module adjusts the position and posture of the locking core relative to the docking guide block, so that when two unmanned vehicle units are docked, the locking core is inserted into the docking guide hole of the docking guide block.6.如权利要求5所述的可重构无人车三段式对接控制方法，其特征在于，所述传感模块包括安装在六自由度平台固定端的视觉传感器以及安装在六自由度平台活动端端面的两个以上激光测距传感器；所述视觉传感器和两个以上所述激光测距传感器分别与所述控制模块相连，用于将检测的信号发送给控制模块；6 . The three-stage docking control method for a reconfigurable unmanned vehicle according to claim 5 , wherein the sensing module comprises a vision sensor installed on the fixed end of the six-degree-of-freedom platform and a movable sensor installed on the six-degree-of-freedom platform. 7 . two or more laser ranging sensors on the end face; the visual sensor and the two or more laser ranging sensors are respectively connected to the control module for sending detected signals to the control module;所述对接导向块上设置有用于和所述视觉传感器配合的图像识别定位板,所述视觉传感器通过对所述图像识别定位板的识别，获得所述对接导向块相对所述锁定芯的位置；The docking guide block is provided with an image recognition positioning plate for cooperating with the vision sensor, and the visual sensor obtains the position of the docking guide block relative to the locking core by recognizing the image recognition positioning plate;所述对接导向块设置有用于和激光测距传感器配合的激光传感器检测板，两个以上所述激光测距传感器沿周向间隔分布，所述控制模块根据两个以上激光测距传感器分别检测到的与所述激光传感器检测板之间的距离信息，获得所述锁定芯轴线与所述对接导向块轴线之间的夹角，所述控制模块以此调整所述锁定芯的姿态使所述对接导向块与所述锁定芯同轴。The docking guide block is provided with a laser sensor detection board for cooperating with the laser ranging sensor, two or more of the laser ranging sensors are distributed at intervals along the circumferential direction, and the control module detects according to the two or more laser ranging sensors respectively. The distance information between the detection board and the laser sensor detection board is obtained, and the angle between the axis of the locking core and the axis of the docking guide block is obtained, and the control module adjusts the posture of the locking core to make the docking The guide block is coaxial with the locking core.7.如权利要求6所述的可重构无人车三段式对接控制方法，其特征在于，所述对接阶段，两个无人车单元对接时，所述主动对接车主动捕捉模块上的视觉传感器获取被动对接车锁定模块上的图像识别定位板的位置信息反馈给控制模块，所述控制模块根据视觉传感器传递的位置信息初步调整锁定芯的位置，使锁定芯和所述对接导向块的相对位置达到设定的对接位置要求；7. The three-stage docking control method for reconfigurable unmanned vehicles according to claim 6, wherein, in the docking stage, when two unmanned vehicle units are docked, the active docking vehicle actively captures the The visual sensor obtains the position information of the image recognition positioning plate on the locking module of the passive docking vehicle and feeds it back to the control module, and the control module preliminarily adjusts the position of the locking core according to the position information transmitted by the visual sensor, so that the locking core and the docking guide block are in the same position. The relative position meets the set docking position requirements;初步调整后，所述主动对接车上的控制模块根据所述激光测距传感器检测到的与锁定模块上激光传感器检测板之间的距离信息，然后调整所述锁定芯的姿态消除解算出的锁定芯轴线与对接导向块轴线的夹角，使其轴线重合；After the preliminary adjustment, the control module on the active docking vehicle adjusts the posture of the locking core to eliminate the calculated lock according to the distance information detected by the laser ranging sensor and the detection board of the laser sensor on the locking module. The included angle between the axis of the core and the axis of the docking guide block makes its axis coincide;然后所述主动对接车上的控制模块控制主动捕捉模块将锁定芯插入到所述对接导向块中；最后通过锁定机构完成主动捕捉模块和锁定模块的锁定。Then the control module on the active docking vehicle controls the active capture module to insert the locking core into the docking guide block; finally, the locking mechanism is used to complete the locking of the active capture module and the locking module.8.如权利要求6所述的可重构无人车三段式对接控制方法，其特征在于，所述传感模块还包括两个以上力传感器；两个以上所述力传感器安装在六自由度平台活动端端面上，且沿周向间隔分布；所述力传感器与所述控制模块相连；8. The three-stage docking control method for reconfigurable unmanned vehicles according to claim 6, wherein the sensing module further comprises two or more force sensors; The movable end face of the degree platform is distributed at intervals along the circumferential direction; the force sensor is connected with the control module;两个无人车单元对接时，所述力传感器与所述对接导向块对接面接触，向所述控制模块反馈对接导向块对接面与所述锁定芯对接面的应力。When the two unmanned vehicle units are docked, the force sensor is in contact with the docking surface of the docking guide block, and feeds back to the control module the stress on the docking surface of the docking guide block and the docking surface of the locking core.

Images