WO2022252220A1 - 一种多轴线平板车精准停靠系统及方法 - Google Patents

一种多轴线平板车精准停靠系统及方法 Download PDF

Info

Publication number
WO2022252220A1
WO2022252220A1 PCT/CN2021/098375 CN2021098375W WO2022252220A1 WO 2022252220 A1 WO2022252220 A1 WO 2022252220A1 CN 2021098375 W CN2021098375 W CN 2021098375W WO 2022252220 A1 WO2022252220 A1 WO 2022252220A1
Authority
WO
WIPO (PCT)
Prior art keywords
axis
flatbed
module
vehicle
information
Prior art date
Application number
PCT/CN2021/098375
Other languages
English (en)
French (fr)
Inventor
鲁守银
李志鹏
高焕兵
王涛
张涛
Original Assignee
山东建筑大学
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 山东建筑大学 filed Critical 山东建筑大学
Priority to AU2021448614A priority Critical patent/AU2021448614A1/en
Publication of WO2022252220A1 publication Critical patent/WO2022252220A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G05—CONTROLLING; REGULATING
    • G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
    • G05D1/00—Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots
    • G05D1/02—Control of position or course in two dimensions
    • G05D1/021—Control of position or course in two dimensions specially adapted to land vehicles
    • G05D1/0231—Control of position or course in two dimensions specially adapted to land vehicles using optical position detecting means
    • G05D1/0242—Control of position or course in two dimensions specially adapted to land vehicles using optical position detecting means using non-visible light signals, e.g. IR or UV signals
    • G—PHYSICS
    • G05—CONTROLLING; REGULATING
    • G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
    • G05D1/00—Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots
    • G05D1/02—Control of position or course in two dimensions
    • G05D1/021—Control of position or course in two dimensions specially adapted to land vehicles
    • G05D1/0231—Control of position or course in two dimensions specially adapted to land vehicles using optical position detecting means
    • G05D1/0246—Control of position or course in two dimensions specially adapted to land vehicles using optical position detecting means using a video camera in combination with image processing means
    • G05D1/0253—Control of position or course in two dimensions specially adapted to land vehicles using optical position detecting means using a video camera in combination with image processing means extracting relative motion information from a plurality of images taken successively, e.g. visual odometry, optical flow
    • G—PHYSICS
    • G05—CONTROLLING; REGULATING
    • G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
    • G05D1/00—Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots
    • G05D1/02—Control of position or course in two dimensions
    • G05D1/021—Control of position or course in two dimensions specially adapted to land vehicles
    • G05D1/0276—Control of position or course in two dimensions specially adapted to land vehicles using signals provided by a source external to the vehicle

Definitions

  • the invention relates to the field of path planning and positioning of mobile robots, in particular to a system and method for precise docking of multi-axis flatbed vehicles based on infrared road signs.
  • multi-axis flatbed trucks are widely used in various large-scale engineering transportation environments. They are the main equipment for transporting segmented hulls in shipyards, and are also suitable for the transportation of super-large concrete prefabricated parts in steel mills and roads. Moreover, the multi-axis flatbed truck needs to perform frequent docking actions, and the docking accuracy has an extremely important impact on the completion of the task.
  • the purpose of the present invention is to solve the problems existing in the prior art, and provide a method for accurate docking of multi-axis flatbed vehicles based on infrared road signs.
  • the present invention provides a precise parking system for a multi-axis flatbed vehicle, including a vehicle-mounted device and an infrared road sign, the infrared road mark is installed at a stop, and the vehicle-mounted device includes a starting module, a path acquisition module, and a driving module , steering module, trajectory tracking control module, visual perception module and embedded host computer;
  • the starting module is used to start the multi-axis flatbed vehicle, load and read the global map containing coordinate information;
  • the path acquisition module which determines the moving path of the multi-axis flatbed vehicle according to the initial point and the stop point of the multi-axis flatbed vehicle;
  • the drive module is used to control the multi-axis flatbed vehicle to travel according to the trajectory generated by the path acquisition module;
  • the steering module is used to judge the steering mode by collecting path information, analyze the expected rotation angle of each wheel train according to the steering kinematics model of the multi-axis flatbed truck established by the embedded host computer, and receive the latest actual wheel train through the CAN bus Rotation angle, and then use closed-loop control to obtain the control output of each loop, and send corresponding output instructions to each node, so as to control each wheel train to achieve the desired rotation angle, to ensure that the wheels of each combination rotate according to the predetermined angle during the steering process and maintain each combination Coherence between the wheels;
  • the multi-axis flatbed truck has various steering modes such as figure-of-eight steering, conventional steering, and center rotation during driving;
  • the track tracking control module performs real-time positioning on the actual position of the multi-axis flatbed vehicle when it travels, and when the deviation between the actual position and the target position is greater than a set threshold, it adjusts the pose of the flatbed vehicle so that the flatbed vehicle gradually returns to the target path;
  • the visual perception module is installed on the front part of the multi-axis flat car body, recognizes the infrared road signs installed at the stop, and transmits the recognized information to the embedded host computer;
  • the embedded host computer is used to load the steering kinematics model of the wheel system, and complete the positioning of the multi-axis flatbed vehicle and the trajectory planning for precise parking according to the information identified by the visual perception module, and control the drive module of the multi-axis flatbed vehicle to complete the precise dock.
  • the visual perception module is composed of a camera and an image information processing card
  • the camera is used to collect the environmental information in front of the multi-axis flatbed vehicle, and the camera is installed under the front part of the multi-axis flatbed vehicle body, and an optical filter is installed for Accurate recognition of infrared light
  • the image information processing card is used to receive the environmental information sent by the camera, and the environmental information captured by the camera recognizes the information of the infrared road sign in front of the parking space at the stop, and transmits the information to the embedded host computer.
  • the present invention also proposes a method for precise parking of multi-axis flatbed vehicles based on infrared road signs, including the following steps:
  • Step 1 Obtain the global map, initial point and stop point, and add the initial point and stop point to the global map;
  • Step 2 Plan the shortest path required from the initial point to the stop point
  • Step 3 Drive along the shortest path and obtain current location information in real time
  • Step 4 When reaching the vicinity of the stop, decelerate and recognize the infrared road signs. According to the image information of the infrared road signs, the host computer calculates the relative pose of the multi-axis flatbed car and the infrared road signs, and then completes the body of the multi-axis flatbed car. Adjust and realize the docking at the target point.
  • the present invention realizes the precise docking task of the multi-axis flatbed vehicle through trajectory planning and autonomous positioning of the multi-axis flatbed vehicle, combined with the recognition of infrared road signs, and has high parking accuracy; through trajectory planning and motion construction of the virtual multi-axis flatbed vehicle
  • the model makes the docking of the multi-axis flatbed truck more smooth and improves the docking accuracy.
  • Fig. 1 is a system structure diagram
  • Fig. 2 is a general flowchart
  • Fig. 3 is a schematic diagram of steering module control
  • FIG. 4 (b) conventional kinematics model diagram of multi-axis flatbed truck
  • Fig. 4 (c) the kinematics model diagram of the center rotation of the multi-axis flatbed car
  • Figure 5 is a flow chart of precise docking based on the visual perception module
  • Fig. 6 is a flow chart of multi-axis flatbed vehicle motion control
  • Fig. 7 is a schematic diagram of the precise parking motion track of the multi-axis flatbed truck.
  • the multi-axis flatbed vehicle precision parking system in the present invention includes a vehicle-mounted device and an infrared road sign.
  • the infrared road mark is installed at a stop, and the vehicle-mounted device includes a starting module, a path acquisition module, a drive module, Steering module, trajectory tracking control module and visual perception module;
  • the above-mentioned starting module is used to start the multi-axis flatbed truck, load and read the global map containing coordinate information;
  • the above path acquisition module which determines the moving path of the multi-axis flatbed vehicle according to the initial point and the stop point of the multi-axis flatbed vehicle;
  • the above-mentioned drive module is used to control the multi-axis flatbed vehicle to travel according to the trajectory generated by the path acquisition module;
  • the above-mentioned steering module is used to judge the steering mode by collecting path information, analyze the expected rotation angle of each wheel train according to the steering kinematics model of the multi-axis flatbed truck established by the embedded host computer, and receive the latest actual wheel train rotation angle through the CAN bus , and then use the closed-loop control to obtain the control output of each loop, and send corresponding output commands to each node, so as to control each wheel train to achieve the desired rotation angle, and ensure that each combined wheel rotates according to a predetermined angle during the steering process and maintains each combined wheel. coordination between
  • the multi-axis flatbed truck has various steering modes such as figure-of-eight steering, conventional steering, and center rotation during driving;
  • the above-mentioned track tracking control module performs real-time positioning on the actual position of the multi-axis flatbed vehicle when it travels, and when the deviation between the actual position and the target position is greater than the set threshold, it adjusts the pose of the flatbed vehicle so that the flatbed vehicle gradually returns to the target path;
  • the above-mentioned visual perception module is installed on the front of the multi-axis flatbed car body, and recognizes the infrared road signs installed at the stop, and transmits the recognized information to the embedded host computer;
  • the above-mentioned embedded host computer is used to load the steering kinematics model of the wheel system, complete the positioning of the multi-axis flatbed vehicle and the trajectory planning for precise parking according to the information recognized by the visual perception module, and control the drive module of the multi-axis flatbed vehicle to complete the precise docking .
  • the multi-axis flatbed is started by the starting module, the global map with coordinate information is loaded and read, the moving path of the multi-axis flatbed is determined by the path acquisition module according to the initial point and the stop point of the multi-axis flatbed, and the multi-axis flatbed is controlled by the driving module.
  • the axis flatbed car drives according to the trajectory generated by the path acquisition module.
  • the steering module collects path information, judges the steering mode, and analyzes the expected rotation angle of each wheel system according to the multi-axis flatbed car steering kinematics model established by the embedded host computer.
  • the actual position of the axis flatbed is positioned in real time.
  • the trajectory tracking control module adjusts the pose of the flatbed to make the flatbed gradually return to the target path.
  • Infrared road signs are installed in front of the parking spaces at the stops, and a visual perception module is installed at the front of the multi-axis flatbed.
  • the multi-axis flatbed moves to the vicinity of the stop
  • the multi-axis flatbed slows down and rotates in situ by the visual perception module.
  • Search after searching the infrared road signs, calculate the image position information of the infrared road signs, and transmit the recognized information to the embedded host computer of the multi-axis flatbed truck.
  • the embedded host computer of the multi-axis flatbed car calculates the relative pose of the multi-axis flatbed car and the infrared road sign, and realizes the motion planning of the multi-axis flatbed car, and then the drive module controls the multi-axis flatbed car to achieve precise parking.
  • the visual perception module is composed of a camera and an image information processing card
  • the camera is used to collect the environmental information in front of the multi-axis flatbed vehicle, and the camera is installed under the front part of the multi-axis flatbed vehicle body, and an optical filter is installed for Accurately identify infrared light
  • the image information processing card is used to receive the environmental information sent by the camera, and further complete the processing of image information.
  • the shape of the infrared road sign is an isosceles right triangle with the length of two sides being 50 cm, and infrared light-emitting diodes are installed at the three vertices of the infrared road sign.
  • Step 1.1 Start the multi-axis flatbed truck by the startup module, determine the global information of the job, and rasterize the global scene of the job to obtain the global grid map.
  • Step 1.2 Obtain the initial positioning information (X 0 , Y 0 , ⁇ 0 ) of the multi-axis flatbed truck in the global map coordinate system, as well as the target stop information, and store them in the database.
  • Step 2.1 In the global grid map in step 1.1, the path acquisition module calculates the shortest path from the initial position to the vicinity of the target stop by the shortest path algorithm.
  • the A* algorithm is used as the shortest path algorithm.
  • the specific implementation process of the algorithm is as follows:
  • the A* algorithm uses the combination of heuristic search and breadth-first algorithm, and is the most effective direct search algorithm for solving the optimal path in a static environment.
  • the A* algorithm uses a cost function F(n), selects the search direction to expand from the starting point to the surrounding, calculates the cost value of each surrounding node through the heuristic function H(n), selects the minimum cost value as the next expansion point, and repeats This process, until the end point is reached, generates a path from the start point to the end point. In the search process, since each node on the path is a node with the minimum cost, the path cost obtained is the minimum.
  • F(n) is the evaluation function of the current node
  • G(n) is the actual path cost from the starting point to the current node
  • H(n) is the minimum estimated cost from the current node to the target point.
  • the abscissa of the starting point is x i
  • the ordinate of the starting point is y i
  • the abscissa of the end point is x n
  • the ordinate of the end point is y n .
  • Step 3.1 Under the control of the drive module, the multi-axis flatbed vehicle walks according to the shortest path planned in step 2. During the steering process, the path information is collected through the steering module, and the steering mode is judged. According to the multi-axis flatbed vehicle steering mathematics established by the embedded host computer The model calculates the expected angle of each wheel, and receives the latest actual angle of rotation of each wheel through the CAN bus, and then uses closed-loop control to obtain the control output of each loop, and sends corresponding output instructions to each node, so as to control each wheel train to achieve the desired angle .
  • the figure-of-eight steering of the multi-axis flatbed truck is taken as an example.
  • the vertical line of each wheel intersects with the steering center line at a point P, which is recorded as the steering point of the multi-axis flatbed truck.
  • the steering angle of each wheel on the right side is ⁇ i
  • the steering angle of each wheel on the left side is ⁇ i
  • the distance from the steering point P to the center line of the flatbed is R.
  • the distance between the axles is equal, and set as S 1
  • the distance between the two axles 3 and 4 is S 2 , taking the multi-axis flatbed truck turning right as an example to establish the mathematical model of the deflection angle:
  • the embedded host computer calculates the steering target value ⁇ 0 of each wheel set through the steering mathematical model of the multi-axis flatbed truck, and then compares it with the actual steering deflection angle ⁇ t fed back by the angular displacement sensor of each wheel set to calculate the steering
  • the angle error ⁇ is transmitted to the proportional amplifier as an output signal to obtain the current signal I, which is used to control the electro-hydraulic proportional valve.
  • the change in the opening of the electro-hydraulic proportional valve is then converted into flow Q output, and then the movement of the hydraulic cylinder is controlled to realize each wheel. Steer to the steering target value ⁇ 0 , and then realize the cooperative work between the various wheel sets.
  • Step 3.2 The trajectory tracking control module obtains the positioning information of the multi-axis flatbed in real time. When the deviation between the actual position and the target position is greater than a threshold, the trajectory tracking control module adjusts the pose of the flatbed to make the flatbed gradually return to the target path.
  • trajectory tracking control module in this embodiment uses the dead reckoning algorithm to realize the real-time positioning of the multi-axis flatbed vehicle, and the specific implementation process of the algorithm is as follows:
  • dead reckoning uses a certain point on the earth's surface as the origin of the local coordinate system, and calculate the relative position of the moving object at the next moment according to the change in velocity direction and speed of the moving object at the current moment, and repeat this process continuously , the trajectory of the moving object can be calculated.
  • ⁇ i represents the heading of the moving object at point P i at time t i. (where 0 ⁇ i ⁇ n).
  • Step 4.1 Collect the environmental information in front of the multi-axis flatbed vehicle through the visual perception module, and process the image information, so as to obtain the position information of the infrared road sign in the image;
  • Step 4.2 Send the position information of the infrared road sign in the image to the embedded host computer, and the embedded host computer calculates the relative pose of the multi-axis flatbed truck and the infrared road sign through the point perspective problem analysis method;
  • Step S1 create a world coordinate system on the infrared road sign, obtain the position information of the infrared road mark under this coordinate system through measurement, the camera is fixed on the front part of the multi-axis flat panel body, and the internal parameters of the camera are calibrated simultaneously;
  • the internal parameter matrix of the camera is:
  • f u and f v are the focal lengths in u and v directions respectively
  • ⁇ is the parameter of the skewness of the two coordinate axes of the image
  • u 0 and v 0 are the image coordinates of the principal point respectively.
  • Step S2 use the perspective n-point problem (PnP) analysis method to calculate the external parameters of the camera;
  • PnP perspective n-point problem
  • the external parameter matrix of the camera is:
  • c P is the translation matrix of the origin of the world coordinates and the origin of the camera coordinates.
  • Step S3 Calibrate the transformation matrix of the multi-axis flatbed vehicle coordinate system and the camera coordinate system, and then calculate the pose of the multi-axis flatbed vehicle in the world coordinate system according to the internal and external parameters of the camera.
  • Step 4.3 According to the path planning during precise parking, the drive module controls the movement of the multi-axis flatbed, and uses the closed-loop control algorithm to complete precise parking.
  • Step A1 Assuming that the body length of the multi-axis flatbed is L 1 , the width of the body is L 2 , and the initial coordinates of point A are (x A y A ⁇ A ) T , the coordinates of point B can be calculated through geometric relations.
  • the coordinates of point B are:
  • the feedback error of the closed-loop system can be obtained through calculation.
  • the closed-loop feedback error is:
  • the multi-axis flatbed is controlled by the closed-loop drive module to move from A to B. During the process, the multi-axis flatbed is gradually approaching the parking space.
  • Step A2 After the multi-axis flatbed truck arrives at point B, adjust the vehicle body so that the pose of the vehicle body is (x B y B ⁇ B ) T .
  • ⁇ B is:
  • Step A3 The initial position of a virtual multi-axis flatbed is at point C, the virtual multi-axis flatbed travels along the straight line CD to point D at a constant speed, calculates and collects the position information of the virtual multi-axis flatbed at each moment, and uses the virtual multi-axis flatbed Calculate the feedback error with the current position and attitude information of the multi-axis flatbed car, and the drive module controls the flatbed to move steadily to point D, gradually reducing the feedback error in the X W direction.
  • Step A4 After the multi-axis flatbed vehicle moves to point D, the visual perception module recognizes the infrared landmarks again to obtain the current position information of the multi-axis flatbed vehicle, so as to eliminate the cumulative error when using dead reckoning for real-time positioning.

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Aviation & Aerospace Engineering (AREA)
  • Radar, Positioning & Navigation (AREA)
  • Remote Sensing (AREA)
  • Automation & Control Theory (AREA)
  • Electromagnetism (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Multimedia (AREA)
  • Traffic Control Systems (AREA)
  • Automobile Manufacture Line, Endless Track Vehicle, Trailer (AREA)
  • Control Of Position Or Direction (AREA)
  • Steering Control In Accordance With Driving Conditions (AREA)

Abstract

一种多轴线平板车精准停靠系统及方法，包括：启动模块，用于启动多轴线平板车，加载并读取含有坐标信息的全局地图；路径获取模块，其根据多轴线平板车的初始点和停靠点确定多轴线平板车的移动路径，并生成轨迹；驱动模块，用于控制多轴线平板车按照路径获取模块生成的轨迹行驶；轨迹跟踪控制模块，对多轴线平板车行进时的实际位置进行实时定位和调整；视觉感知模块，对安装在停靠点的红外路标进行识别，并将识别的信息传送到嵌入式上位机；嵌入式上位机，根据视觉感知模块识别的信息完成对多轴线平板车的定位以及精准停靠时的轨迹规划，并控制多轴线平板车的驱动模块完成精准停靠。

Description

一种多轴线平板车精准停靠系统及方法 技术领域
本发明涉及移动机器人路径规划以及定位领域，特别地涉及一种基于红外路标的多轴线平板车精确停靠系统及方法。
背景技术
随着智能制造领域的发展，多轴线平板车被广泛应用各种大型的工程运输环境之中，是造船厂运输分段船体的主要设备，也适应于钢厂和公路特大型混凝土预制件的运输，并且多轴线平板车需要进行频繁的停靠动作，并且停靠的精度对任务的完成度有着极其重要的影响。
对于多轴线平板车完成停靠任务，当前有许多不同的解决办法，各个方案都存在一些缺点，有些方案需要在路面事先铺设好磁条，通过磁条感应信号实现引导，但磁条容易破损并且路径变更需要重新铺设磁条。有些方案需要在停靠点附近装设传感器等辅助设备，但是受制于传感器自身的原因造成多轴线平板车不能完成精准停靠，本方案通过路径规划使多轴线平板车行驶最短路径达到停靠点附近，并在停靠点附近安装红外路标实现多轴线平板车在停靠点附近的定位，进一步实现多轴线平板车的精准停靠。
发明内容
本发明的目的在于解决现有技术中存在的问题，并提供一种基于红外路标的多轴线平板车精确停靠的方法。
本发明采用的技术方案如下：
第一方面，本发明中提供了一种多轴线平板车精准停靠系统，包括车载装 置和红外路标，所述的红外路标安装在停靠点，所述的车载装置包括启动模块、路径获取模块、驱动模块、转向模块、轨迹跟踪控制模块和视觉感知模块和嵌入式上位机；
所述的启动模块，用于启动多轴线平板车，加载并读取含有坐标信息的全局地图；
所述的路径获取模块，其根据多轴线平板车的初始点和停靠点确定多轴线平板车的移动路径；
所述的驱动模块，用于控制多轴线平板车按照路径获取模块生成的轨迹行驶；
所述的转向模块，用于通过采集路径信息，判断转向模式，根据嵌入式上位机建立的多轴线平板车转向运动学模型解析出各轮系的期望转角，并通过CAN总线接收最新的实际轮系转角，再利用闭环控制得到各回路的控制量输出，向各节点发送相应的输出指令，从而控制各轮系达到期望转角，确保在转向过程中各个组合车轮间按照预定的角度回转并保持各个组合车轮间的协调一致；
其中，多轴线平板车在行驶过程中存在八字转向、常规转向、中心回转等多种转向模式；
所述的轨迹跟踪控制模块，对多轴线平板车行进时的实际位置进行实时定位,当实际位置与目标位置偏差大于设定的阈值时，其调整平板车的位姿，使平板车逐渐回到目标路径；
所述的视觉感知模块，安装在多轴线平板车体前部，对安装在停靠点的红外路标进行识别，并将识别的信息传送到嵌入式上位机；
所述的嵌入式上位机，用于加载车轮系转向运动学模型，并根据视觉感知 模块识别的信息完成对多轴线平板车的定位以及精准停靠时的轨迹规划，控制多轴线平板车的驱动模块完成精准停靠。
优选的，视觉感知模块由一个相机、一个图像信息处理卡组成，相机用于采集多轴线平板车前方的环境信息，并且相机安装在多轴线平板车体的前部下方，并安装滤光片用来准确的识别红外光；图像信息处理卡用来接收相机传送来的环境信息，由相机拍摄的环境信息识别出停靠点车位前方红外路标的信息，并将信息传送到嵌入式上位机。
第二方面，本发明还提出了一种基于红外路标的多轴线平板车精确停靠的方法，包括如下步骤：
步骤1：获取全局地图、初始点与停靠点，并将初始点与停靠点添加在全局地图中；
步骤2：规划从初始点到停靠点所需要的最短路径；
步骤3：沿最短路径进行行驶，并实时获取当前的定位信息；
步骤4：当达到停靠点附近时，进行减速，并对红外路标进行识别，根据红外路标的图像信息，由上位机计算出多轴线平板车与红外路标的相对位姿，进而完成多轴线平板车的车身调整，并实现目标点的停靠。
本发明的有益效果如下：
本发明通过多轴线平板车的轨迹规划和自主定位，以及结合对红外路标的识别，实现了多轴线平板车的精准停靠任务，且拥有较高的停靠精度；通过轨迹规划和虚拟多轴线平板车的运动建模，使得多轴线平板车的停靠更加流畅，并且提高了停靠精度。
附图说明
图1为系统结构图；
图2为总流程图；
图3为转向模块控制原理图；
图4(a)多轴线平板车八字转向运动学模型图；
图4(b)多轴线平板车常规运动学模型图；
图4(c)多轴线平板车中心回转运动学模型图；
图5为基于视觉感知模块的精准停靠流程图；
图6为多轴线平板车运动控制流程图；
图7为多轴线平板车精准停靠运动轨迹示意图。
具体实施方式
下面结合附图和具体实施方式对本发明做进一步阐述，以便更好地理解本发明。
如图1所示，本发明中的多轴线平板车精准停靠系统，包括车载装置和红外路标，所述的红外路标安装在停靠点，所述的车载装置包括启动模块、路径获取模块、驱动模块、转向模块、轨迹跟踪控制模块和视觉感知模块；
上述的启动模块，用于启动多轴线平板车，加载并读取含有坐标信息的全局地图；
上述的路径获取模块，其根据多轴线平板车的初始点和停靠点确定多轴线平板车的移动路径；
上述的驱动模块，用于控制多轴线平板车按照路径获取模块生成的轨迹行驶；
上述的转向模块，用于通过采集路径信息，判断转向模式，根据嵌入式上位机建立的多轴线平板车转向运动学模型解析出各轮系的期望转角，并通过CAN总线接收最新的实际轮系转角，再利用闭环控制得到各回路的控制量输出，向各节点发送相应的输出指令，从而控制各轮系达到期望转角，确保在转向过程中各个组合车轮间按照预定的角度回转并保持各个组合车轮间的协调一致；
其中，多轴线平板车在行驶过程中存在八字转向、常规转向、中心回转等多种转向模式；
上述的轨迹跟踪控制模块，对多轴线平板车行进时的实际位置进行实时定位,当实际位置与目标位置偏差大于设定的阈值时，其调整平板车的位姿，使平板车逐渐回到目标路径；
上述的视觉感知模块，安装在多轴线平板车体前部，对安装在停靠点的红外路标进行识别，并将识别的信息传送到嵌入式上位机；
上述的嵌入式上位机，用于加载车轮系转向运动学模型，并根据视觉感知模块识别的信息完成对多轴线平板车的定位以及精准停靠时的轨迹规划，控制多轴线平板车的驱动模块完成精准停靠。
本发明首先通过启动模块启动多轴线平板车，加载并读取有坐标信息的全局地图，根据多轴线平板车的初始点和停靠点由路径获取模块确定多轴线平板车的移动路径，再由驱动模块控制多轴线平板车按照路径获取模块生成的轨迹行驶，在行驶过程中，通过转向模块采集路径信息，判断转向模式，根据嵌入式上位机建立的多轴线平板车转向运动学模型解析出各轮系的期望转角，并通过CAN总线接收最新的实际轮系转角，再利用闭环控制得到各回路的控制量输出，向各节点发送相应的输出指令，从而控制各轮系达到期望转角；再由轨迹 跟踪控制模块对多轴线平板车的实际位置进行实时定位，当实际位置与目标位置偏差大于一个阈值时，轨迹跟踪控制模块调整平板车的位姿，使平板车逐渐回到目标路径。在停靠点车位前方安装红外路标，在多轴线平板车身前部安装视觉感知模块，当多轴线平板车移动到停靠点附近时，多轴线平板车进行减速并通过原地旋转由视觉感知模块对红外路标进行搜寻，搜寻到红外路标后，计算得到红外路标的图像位置信息，并将识别到的信息传输给多轴线平板车的嵌入式上位机。多轴线平板车的嵌入式上位机计算出多轴线平板车与红外路标的相对位姿，并实现多轴线平板车的运动规划，再由驱动模块控制多轴线平板车实现精准停靠。
优选的，视觉感知模块由一个相机和一个图像信息处理卡组成，相机用于采集多轴线平板车前方的环境信息，并且相机安装在多轴线平板车体的前部下方，并安装滤光片用来准确的识别红外光；图像信息处理卡用来接收相机传送来的环境信息，进一步的完成对图像信息的处理。
优选的，红外路标的形状为等腰直角三角形且两边边长为50cm，且红外路标的三个顶点处安装有红外发光二极管。
如图2、3、5所示，多轴线平板车精准停靠的方法，具体实施步骤如下：
步骤1.1：由启动模块启动多轴线平板车，确定作业全局信息，对作业全局的场景进行栅格化后得到全局栅格地图。
步骤1.2：在全局地图坐标系下获取多轴线平板车初始定位信息(X 0,Y 0,θ 0)，以及目标停靠点信息，并存储在数据库中。
步骤2.1在步骤1.1的全局栅格地图中，由路径获取模块通过最短路径算法计算出初始位置到目标停靠点附近的最短路径。
其中本实施例中采用A*算法作为最短路径算法。该算法的具体实现过程如下：
A*算法采用启发式搜索与广度优先算法结合，是静态环境中用于求解最优路径最有效的直接搜索算法。A*算法通过一个代价函数F(n)，选择搜索方向从起点开始向周围扩展，通过启发函数H(n)计算得到周围每个节点的代价值，选择最小代价值作为下一个扩展点，重复这个过程，直到到达终点，生成从起点到终点的路径。在搜索过程中，由于路径上的每一个节点都是具有最小代价节点，得到的路径代价是最小的。
A*算法的代价函数为：
F(n)＝G(n)+H(n)
Figure PCTCN2021098375-appb-000001
其中F(n)为当前节点的估价函数，G(n)为起始点到当前节点的实际路径代价，H(n)为当前节点到目标点的最小估计代价。
起始点横坐标为x i，起始点纵坐标为y i，终点横坐标为x n，终点纵坐标为y n。
步骤3.1：多轴线平板车在驱动模块控制下按照步骤2中规划出的最短路径行走，在转向过程中，通过转向模块采集路径信息，判断转向模式，根据嵌入式上位机建立的多轴线平板车转向数学模型计算出各轮的期望角度，并通过CAN总线接收各轮最新的实际转角，再利用闭环控制得到各回路的控制量输出，向各节点发送相应的输出指令，从而控制各轮系达到期望转角。
其中多轴线平板车转向运动学模型图中，以多轴线平板车八字转向为例，各车轮垂直线与转向中心线交于一点P，该点记为多轴线平板车转向点。设转向时右侧各车轮转向角为α i，左侧各车轮转向角为β i，转向点P到平板车中心线的距离为R，1、2、3和4、5、6两组轮轴中各轮轴之间的距离相等，并设为S 1,3、4两 轮轴之间的距离为S 2,以多轴线平板车右转为例建立偏转角度的数学模型：
当i＝1、2、3轮轴时
Figure PCTCN2021098375-appb-000002
Figure PCTCN2021098375-appb-000003
所以有
Figure PCTCN2021098375-appb-000004
Figure PCTCN2021098375-appb-000005
当i＝4、5、6轮轴时
α 7-i＝-α i
β 7-i＝-β i
其中闭环控制中，嵌入式上位机通过多轴线平板车转向数学模型计算出各轮组的转向目标值θ 0，再与各轮组角位移传感器反馈的实际转向偏转角度θ t进行比较，计算出转向角度误差ε，作为输出信号传输给比例放大器，获得电流信号I，用其控制电液比例阀，电液比例阀开度的变化再转化成流量Q输出，进而控制液压油缸的运动，实现各车轮转向到转向目标值θ 0，进而实现各轮组间的协同工作。
步骤3.2由轨迹跟踪控制模块实时获取多轴线平板车的定位信息，当实际位置与目标位置偏差大于一个阈值时，轨迹跟踪控制模块调整平板车的位姿，使平板车逐渐回到目标路径。
其中本实施例中轨迹跟踪控制模块采用航迹推算法实现多轴线平板车的实时定位，该算法的具体实现过程如下：
航迹推算的基本思想是以地球表面某点作为当地坐标系的原点，根据移动物体当前时刻的速度方向变化量和速度大小，推算出移动物体下一时刻的相对位置，不断地重复这一过程，便可以推算出移动物体的运动轨迹。
设t i时刻移动物体位于已知点P i(x i，y i)，θ i表示t i时刻，移动物体在点P i时的航向。(其中0≤i≥n)。
Figure PCTCN2021098375-appb-000006
由此可知，若已知t 0时刻移动物体的位置P 0(x 0，y 0)和t 0时刻之后任意时刻速度的大小和方向，则可以推算出t 0时刻之后任意时刻t i移动物体的位置P i(x i，y i)。
步骤4.1：通过视觉感知模块对多轴线平板车前方的环境信息进行采集，并进行图像信息处理，以此得到红外路标在图像中的位置信息；
步骤4.2：将红外路标在图像中的位置信息发送给嵌入式上位机，嵌入式上位机通过点透视问题分析方法计算出多轴线平板车与红外路标的相对位姿；
其中多轴线平板车相对于红外路标的位姿计算方法步骤如下：
步骤S1:在红外路标上创建世界坐标系，经过测量获取在此坐标系下红外路标的位置信息，相机固定于多轴线平板车身前部上，同时标定相机的内部参数；
其中相机的内部参数矩阵为：
Figure PCTCN2021098375-appb-000007
f u、f v分别为u、v两个方向的焦距，γ为图像两坐标轴偏斜度的参数，u 0、v 0分 别为主点的图像坐标。
步骤S2：利用透视n点问题(PnP)分析方法计算出相机的外部参数；
其中相机的外部参数矩阵为：
Figure PCTCN2021098375-appb-000008
Figure PCTCN2021098375-appb-000009
为世界坐标相对于相机坐标的旋转矩阵， c P为世界坐标原点与相机坐标原点的平移矩阵。
步骤S3:标定多轴线平板车坐标系和相机坐标系的变换矩阵，再根据相机的内外部参数计算得到多轴线平板车在世界坐标系下的位姿。
步骤4.3：依据精准停靠时的路径规划，由驱动模块控制多轴线平板车运动，利用闭环控制算法完成精准停靠。
如图6、7所示，多轴线平板车精准停靠运动控制方法，具体过程如下：
步骤A1:假设多轴线平板车车身长为L 1,车身宽为L 2，A点的初始坐标为(x A y A θ A) T，通过几何关系，可以计算出B点的坐标。
其中B点的坐标为：
Figure PCTCN2021098375-appb-000010
Figure PCTCN2021098375-appb-000011
假设在t时刻，多轴线平板车的坐标为(x t y t θ t) T，可以通过计算得到闭环系统反馈误差。
其中闭环反馈误差为：
e x＝x t-x B
e y＝y t-y B
Figure PCTCN2021098375-appb-000012
依据闭环反馈误差e＝(e x e y e t) T，通过驱动模块闭环控制多轴线平板车由A到B运动，在此过程中多轴线平板车渐渐接近停靠车位。
步骤A2：多轴线平板车到达B点后，进行车身的调整，使车身的位姿为(x B y B θ B) T。
其中θ B为：
Figure PCTCN2021098375-appb-000013
步骤A3：虚拟一个多轴线平板车其初始位置在C点，虚拟的多轴线平板车沿直线CD向D点做匀速行驶，计算并采集每一时刻虚拟多轴线平板车的位置信息，利用虚拟的多轴线平板车的位姿信息计算出与当前多轴线平板车位姿信息的反馈误差，由驱动模块控制平板车平稳的向D点运动，逐渐减小X W方向的反馈误差。
步骤A4：当多轴线平板车运动到D点后，再次由视觉感知模块对红外路标进行识别得到多轴线平板车当前的位置信息，以此来消除用航迹推算进行实时定位时的累计误差。
以上所述仅为本发明的优选实施例而已，并不用于限制本发明，对于本领域的技术人员来说，本发明可以有各种更改和变化。凡在本发明的精神和原则之内，所作的任何修改、等同替换、改进等，均应包含在本发明的保护范围之内。

Claims (9)

  1. 一种多轴线平板车精准停靠系统，其特征在于，包括车载装置和红外路标，所述的红外路标安装在停靠点，所述的车载装置包括启动模块，用于启动多轴线平板车，加载并读取含有坐标信息的全局地图；路径规划模块，其根据多轴线平板车的初始点和停靠点确定移动路径；驱动模块，用于控制多轴线平板车按照路径获取模块生成的轨迹行驶；轨迹跟踪控制模块，对多轴线平板车行进时的实际位置进行实时定位和调整；视觉感知模块，对安装在停靠点的红外路标进行识别，并将识别的信息传送到嵌入式上位机；嵌入式上位机，用于加载车轮系转向运动学模型，并根据视觉感知模块识别的信息完成对多轴线平板车的定位以及精准停靠时的轨迹规划，控制多轴线平板车的驱动模块完成精准停靠。
  2. 如权利要求1所述的多轴线平板车精准停靠系统，其特征在于，所述的车载装置还包括转向模块，用于通过采集路径信息，判断转向模式，根据嵌入式上位机建立的多轴线平板车转向运动学模型解析出各轮系的期望转角，并通过总线接收最新的实际轮系转角，再利用闭环控制得到各回路的控制量输出，向各节点发送相应的输出指令。
  3. 如权利要求1所述的多轴线平板车精准停靠系统，其特征在于，所述的视觉感知模块包括相机、图像信息处理装置，所述相机用于采集多轴线平板车前方的环境信息，并且相机安装在多轴线平板车体的前部下方，并安装滤光片用来准确的识别红外光；图像信息处理装置用来接收相机传送来的环境信息，由相机拍摄的环境信息识别出停靠点车位前方红外路标的信息，并将信息传送到嵌入式上位机。
  4. 一种多轴线平板车精确停靠的方法，其特征在于，包括如下步骤：
    步骤1：获取全局地图、初始点与停靠点，并将初始点与停靠点添加在全局地图中；
    步骤2：规划从初始点到停靠点所需要的最短路径；
    步骤3：沿最短路径进行行驶，并实时获取当前的定位信息；
    步骤4：当达到停靠点附近时，进行减速，并对红外路标进行识别，根据红外路标的图像信息，由上位机计算出多轴线平板车与红外路标的相对位姿，进而完成多轴线平板车的车身调整，并实现目标点的停靠。
  5. 如权利要求4所述的多轴线平板车精确停靠的方法，其特征在于，步骤1中将全局地图进行格栅化后得到全局栅格地图，在全局栅格地图中获取多轴线平板车初始定位信息，以及目标停靠点信息，并存储在数据库中。
  6. 如权利要求5所述的多轴线平板车精确停靠的方法，其特征在于，步骤2中，在全局栅格地图中，由路径获取模块通过最短路径算法计算出初始位置到目标停靠点附近的最短路径。
  7. 如权利要求4所述的多轴线平板车精确停靠的方法，其特征在于，步骤3中，多轴线平板车在驱动模块控制下按照步骤2中规划出的最短路径行走，在转向过程中，通过转向模块采集路径信息，判断转向模式，根据嵌入式上位机建立的多轴线平板车转向数学模型计算出各轮的期望角度，并通过总线接收各轮最新的实际转角，再利用闭环控制得到各回路的控制量输出，向各节点发送相应的输出指令，从而控制各轮系达到期望转角。
  8. 如权利要求4所述的多轴线平板车精确停靠的方法，其特征在于，步骤3中，由轨迹跟踪控制模块实时获取多轴线平板车的定位信息，当实际位置与目标位置偏差大于一个阈值时，轨迹跟踪控制模块调整平板车的位姿，使平板车 逐渐回到目标路径。
  9. 如权利要求4所述的多轴线平板车精确停靠的方法，其特征在于，步骤3中，多轴线平板车相对于红外路标的位姿计算方法如下：
    步骤S1:在红外路标上创建世界坐标系，经过测量获取在此坐标系下红外路标的位置信息，相机固定于多轴线平板车身前部上，同时标定相机的内部参数；
    步骤S2：利用透视n点问题分析方法计算出相机的外部参数；
    步骤S3:标定多轴线平板车坐标系和相机坐标系的变换矩阵，再根据相机的内外部参数计算得到多轴线平板车在世界坐标系下的位姿。
PCT/CN2021/098375 2021-06-01 2021-06-04 一种多轴线平板车精准停靠系统及方法 WO2022252220A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
AU2021448614A AU2021448614A1 (en) 2021-06-01 2021-06-04 Precise stopping system and method for multi-axis flatbed vehicle

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202110608144.5A CN113341968A (zh) 2021-06-01 2021-06-01 一种多轴线平板车精准停靠系统及方法
CN202110608144.5 2021-06-01

Publications (1)

Publication Number Publication Date
WO2022252220A1 true WO2022252220A1 (zh) 2022-12-08

Family

ID=77474070

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2021/098375 WO2022252220A1 (zh) 2021-06-01 2021-06-04 一种多轴线平板车精准停靠系统及方法

Country Status (3)

Country Link
CN (1) CN113341968A (zh)
AU (1) AU2021448614A1 (zh)
WO (1) WO2022252220A1 (zh)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114633801B (zh) * 2022-04-22 2023-07-21 武汉科技大学 一种运梁车隧道行走自动纠偏系统及方法

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107422730A (zh) * 2017-06-09 2017-12-01 武汉市众向科技有限公司 基于视觉导引的agv运输系统及其驾驶控制方法
CN107450540A (zh) * 2017-08-04 2017-12-08 山东大学 基于红外路标的室内移动机器人导航系统及方法
CN110673610A (zh) * 2019-10-11 2020-01-10 天津工业大学 一种基于ros的工厂agv路径规划方法
CN111596657A (zh) * 2020-05-09 2020-08-28 浙江工业大学 一种agv及其轨道运动方法
WO2020210808A1 (en) * 2019-04-12 2020-10-15 Continental Automotive Systems, Inc. Autonomous vehicle-trailer maneuvering and parking
CN111976718A (zh) * 2020-07-13 2020-11-24 浙江大华汽车技术有限公司 自动泊车的控制方法和系统

Family Cites Families (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104382710B (zh) * 2014-11-24 2017-02-22 中国科学院自动化研究所 一种智能轮椅床系统的床椅自动对接方法
CN205375187U (zh) * 2015-12-15 2016-07-06 北京智行者科技有限公司 一种基于多轴无轨电车循迹跟踪系统及具有其的电车
CN106681320A (zh) * 2016-12-15 2017-05-17 浙江大学 一种基于激光数据的移动机器人导航控制方法
CN108388244A (zh) * 2018-01-16 2018-08-10 上海交通大学 基于人工路标的移动机器人系统、停靠方法及存储介质
CN111986506B (zh) * 2020-07-20 2022-04-01 苏州易航远智智能科技有限公司 基于多视觉系统的机械式停车位泊车方法
CN112506195B (zh) * 2020-12-02 2021-10-29 吉林大学 基于视觉和底盘信息的车辆自主定位系统及定位方法

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107422730A (zh) * 2017-06-09 2017-12-01 武汉市众向科技有限公司 基于视觉导引的agv运输系统及其驾驶控制方法
CN107450540A (zh) * 2017-08-04 2017-12-08 山东大学 基于红外路标的室内移动机器人导航系统及方法
WO2020210808A1 (en) * 2019-04-12 2020-10-15 Continental Automotive Systems, Inc. Autonomous vehicle-trailer maneuvering and parking
CN110673610A (zh) * 2019-10-11 2020-01-10 天津工业大学 一种基于ros的工厂agv路径规划方法
CN111596657A (zh) * 2020-05-09 2020-08-28 浙江工业大学 一种agv及其轨道运动方法
CN111976718A (zh) * 2020-07-13 2020-11-24 浙江大华汽车技术有限公司 自动泊车的控制方法和系统

Also Published As

Publication number Publication date
AU2021448614A9 (en) 2024-06-20
CN113341968A (zh) 2021-09-03
AU2021448614A1 (en) 2023-06-15

Similar Documents

Publication Publication Date Title
CN110262508B (zh) 应用于封闭场地无人驾驶货运车辆上的自动引导系统及方法
CN106950964B (zh) 无人电动大学生方程式赛车及其控制方法
CN107272008A (zh) 一种带惯性补偿的agv激光导航系统
CN110307850A (zh) 航迹推算定位方法及自动泊车系统
US20080059015A1 (en) Software architecture for high-speed traversal of prescribed routes
CN107992054A (zh) 一种机器人的定位的方法及系统
CN108345005A (zh) 巷道掘进机的实时连续自主定位定向系统及导航定位方法
CN104914865A (zh) 变电站巡检机器人定位导航系统及方法
CN106123908A (zh) 车辆导航方法和系统
CN105955257A (zh) 基于固定路线的公交车自动驾驶系统及其驾驶方法
CN109900273B (zh) 一种室外移动机器人导引方法及导引系统
CN110907891B (zh) 一种agv定位方法及装置
CN114200945B (zh) 一种移动机器人的安全控制方法
CN203490506U (zh) 一种车载激光自动导向控制系统
CN202166895U (zh) 变电站智能巡检机器人的激光导航系统
WO2022252220A1 (zh) 一种多轴线平板车精准停靠系统及方法
Beliveau et al. Autonomous vehicle navigation with real-time 3D laser based positioning for construction
CN111596666A (zh) 一种基于agv运动预测的障碍物碰撞威胁的检测方法
CN112710301B (zh) 一种自动驾驶车辆高精度定位方法和系统
CN106052660A (zh) 与全站仪交互的自动锚点标识小车
CN113341999A (zh) 一种基于优化d*算法的叉车路径规划方法及装置
CN211427151U (zh) 一种应用于封闭场地无人驾驶货运车辆上的自动引导系统
TWI770966B (zh) 無人自走車之導引控制方法
CN109828569A (zh) 一种基于2d-slam导航的智能agv叉车
TWI770965B (zh) 無人自走車之導引控制方法

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 21943581

Country of ref document: EP

Kind code of ref document: A1

ENP Entry into the national phase

Ref document number: 2021448614

Country of ref document: AU

Date of ref document: 20210604

Kind code of ref document: A

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 21943581

Country of ref document: EP

Kind code of ref document: A1