WO2022027865A1 - 基于行走无人洗车设备的双视觉反馈运动控制系统 - Google Patents

基于行走无人洗车设备的双视觉反馈运动控制系统 Download PDF

Info

Publication number
WO2022027865A1
WO2022027865A1 PCT/CN2020/129958 CN2020129958W WO2022027865A1 WO 2022027865 A1 WO2022027865 A1 WO 2022027865A1 CN 2020129958 W CN2020129958 W CN 2020129958W WO 2022027865 A1 WO2022027865 A1 WO 2022027865A1
Authority
WO
WIPO (PCT)
Prior art keywords
control system
motion control
car washing
module
camera module
Prior art date
Application number
PCT/CN2020/129958
Other languages
English (en)
French (fr)
Inventor
施恒之
Original Assignee
浙江驿公里智能科技有限公司
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 浙江驿公里智能科技有限公司 filed Critical 浙江驿公里智能科技有限公司
Publication of WO2022027865A1 publication Critical patent/WO2022027865A1/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
    • G05D1/0246—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
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B60—VEHICLES IN GENERAL
    • B60S—SERVICING, CLEANING, REPAIRING, SUPPORTING, LIFTING, OR MANOEUVRING OF VEHICLES, NOT OTHERWISE PROVIDED FOR
    • B60S3/00—Vehicle cleaning apparatus not integral with vehicles
    • B60S3/04—Vehicle cleaning apparatus not integral with vehicles for exteriors of land vehicles
    • 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
    • G05D1/0287—Control of position or course in two dimensions specially adapted to land vehicles involving a plurality of land vehicles, e.g. fleet or convoy travelling
    • 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
    • G05D1/08—Control of attitude, i.e. control of roll, pitch, or yaw
    • G05D1/0891—Control of attitude, i.e. control of roll, pitch, or yaw specially adapted for land vehicles

Definitions

  • the invention relates to the field of unmanned car washing, and more particularly to a dual visual feedback motion control system based on walking unmanned car washing equipment.
  • the unmanned car washing machine refers to the equipment that can automatically complete the vehicle cleaning without the intervention of car washing personnel. Due to its advantages of low cleaning cost and high efficiency, the use of unmanned car washing machines for vehicle cleaning is slowly becoming the first choice for car owners.
  • a common unmanned car washing machine is a gantry car washing machine.
  • the gantry car washing machine includes a fixed walking track, a gantry frame placed on the track, and various types of car washing equipment placed in the gantry frame, which need to be fixed in a specific The place is used by car owners, but this kind of car washing machine is bulky, expensive to build and maintain, and the car washing method is simple and lacks flexibility.
  • the present application provides a dual-visual feedback motion control system based on walking unmanned car washing equipment, which can enable the unmanned car washing equipment under the control of the system to imitate humans to move around the car when washing the car, which can be performed on more occasions Compared with the gantry type car washing machine, it is more flexible and can be stored in the warehouse when it is idle, reducing maintenance costs.
  • a dual visual feedback motion control system based on walking unmanned car washing equipment including a depth camera module, a field camera module, a local motion control system and a cloud control system
  • the depth camera module is installed on the walking unmanned car washing machine and communicates with the local
  • the motion control system is connected by lines
  • the local motion control system is integrated on the electric cabinet of the walking unmanned car washing machine
  • the site camera module is installed around the car wash site
  • the cloud control system is mounted on the cloud server and communicates with the local motion control system. It is wirelessly connected with the field camera module, and the cloud control system communicates with the local motion control system and the field camera module through the first wireless module and the second wireless module respectively.
  • each car washing site is equipped with at least two walking unmanned car washing machines, and each walking unmanned car washing machine has three parts: a depth camera module, a local motion control system and an execution system.
  • the depth camera module and the local motion control system The system is electrically connected, and the depth camera module captures video information from the scene of the car washing site and transmits it to the local motion control system.
  • the local motion control system is electrically connected with the execution system.
  • the information interaction of each walking unmanned car washing machine is a parallel structure, and the local motion
  • the control system and the cloud control system exchange information through the wireless module, and the execution system and the local motion control system can carry out two-way information dissemination.
  • the local motion control system is integrated with a high-performance mainboard, a drive module and a monitoring module
  • the output end of the high-performance mainboard is electrically connected to the input end of the drive module
  • the output end of the drive module is electrically connected to the input end of the execution system
  • the drive module is used for Receive operation instructions and drive each actuator to move
  • the output end of the execution system is electrically connected to the input end of the monitoring module
  • the output end of the monitoring module is electrically connected to the input end of the high-performance motherboard
  • the monitoring module includes other sensors to assist the machine to obtain other information.
  • the mainboard sends the signal given by the cloud control system to the drive module to drive the execution system for execution, while the monitoring module monitors the state parameters of the execution system in real time and transmits them back to the high-performance mainboard.
  • the high-performance motherboard includes an upper computer function module and a lower computer function module
  • the performance of the lower computer function module is lower than that of the upper computer function module
  • the upper computer function module is internally equipped with a vision processing algorithm and a driving instruction
  • the processing algorithm includes Recognition of the vehicle model, calculation of the distance from the machine to the vehicle, recognition of the attitude angle of the vehicle
  • the driving instructions include the movement of the AGV chassis, the movement attitude of the brush, and the opening and closing of the door structure.
  • the output end of the depth camera is electrically connected to the input end of the upper computer function module, and the video information collected by the depth camera is specifically transmitted to the upper computer function module of the high-performance motherboard for advanced algorithm operation processing, and the upper computer function module is connected with the lower computer function module.
  • the computer function module is electrically connected, and the information of the upper computer function module and the lower computer function module can be transmitted to each other, and the transmitted information includes motion task signals, real-time machine status, etc.
  • the output end of the cloud control system is electrically connected to the input end of the lower computer function module, and the cloud control The control signal issued by the system is sent to the lower computer function module.
  • the depth camera module is installed in the depth camera, the depth camera module is directly connected with the high-performance motherboard on the local motion control system through a line, and the depth camera is directly installed under the top cover of the walking unmanned car washing machine to
  • the visual information of the direction facing the vehicle is captured, including model information, machine-to-vehicle angle information, machine-to-vehicle distance information, and vehicle attitude information.
  • the field camera module consists of a plurality of cameras in an array and is evenly installed around the field to monitor the position and attitude of each walking unmanned car washing machine, the position and attitude of the vehicle, the markings and signs on the field, and the People and objects flow information for shooting.
  • the cloud control system is equipped with a visual recognition algorithm and a scheduling and planning algorithm.
  • the visual recognition algorithm includes recognizing the position and attitude of the vehicle, recognizing the position and attitude of each machine, and tracking the movement path of each machine.
  • the scheduling and planning algorithm includes calculating The path that different machines need to travel, the path point that the machine will reach in the next step according to the current position of the machine, the path that the car needs to travel and the attitude that needs to be calculated based on the current position of the car, and the guidance command is issued.
  • the present invention has the following characteristics and beneficial effects:
  • the local motion control system performs algorithm operations according to the information to obtain the actions that the walking unmanned car washing equipment needs to perform at this location.
  • the location and posture of the car on the venue and the position and posture of the multiple walking unmanned car wash equipment are photographed through the venue camera module, and then the information is uploaded to the cloud control system.
  • the cloud control system passes the algorithm Enter the local motion control system of each walking unmanned car washing equipment with scheduling planning and dispatching information to determine the next movement of the walking unmanned car washing equipment;
  • the dual acquisition module the visual information of the vehicle direction (micro) and the complete information of the car wash (macro) can be collected at the same time, so that the current position of the car can be used to calculate the path that the car needs to go through and the attitude that needs to be achieved and issue guidance instructions. , so that it will not collide with the car washing water equipment.
  • the gantry type car washing machine it is more flexible and can be stored in the warehouse when it is idle, reducing maintenance costs.
  • Fig. 1 is the overall block diagram of the dual visual feedback motion system based on the walking unmanned car washing equipment
  • Figure 2 is an information block diagram of the local motion control system.
  • a dual visual feedback motion control system based on walking unmanned car washing equipment including a depth camera module 1, a field camera module 2, a local motion control system 3 and a cloud control system 4, the depth camera module 1 is installed in the walking
  • the unmanned car washing machine is connected to the local motion control system 3 through a line.
  • the local motion control system 3 is integrated on the electric cabinet of the walking unmanned car washing machine.
  • the depth camera module captures video information and enters the local motion control system 3.
  • the local motion control system 3 Algorithm operation is performed according to the information to obtain the actions and movement methods that the unmanned car washing equipment needs to perform in this position; the site camera module 2 is installed around the car washing site, and the cloud control system 4 is mounted on the cloud server and communicates with the local motion control system.
  • the system 3 is wirelessly connected to the site camera module 2, and the site camera module 2 shoots the site information such as the position and attitude of the car on the site and the position and attitude of the plural walking unmanned car washing equipment, and then uploads the information to the cloud control system 4,
  • the cloud control system 4 communicates with the local motion control system 3 and the site camera module 2 through the first wireless module and the second wireless module respectively.
  • the cloud control system 4 dispatches the scheduling information through the algorithm to dispatch the scheduling information into each walking unmanned car washing equipment.
  • the local motion control system 3 is used to determine the next movement of the walking unmanned car washing equipment, and at the same time, collect and analyze the monitoring information sent in the local motion system and the video stream information collected by the site camera module 2, and report to the local motion.
  • the control system 3 sends task information.
  • each car wash site is equipped with at least 2 walking unmanned car washing machines, and each walking unmanned car washing machine has three parts: depth camera module 1, local motion control system 3 and execution system 5, depth camera module 1 It is electrically connected with the local motion control system 3, and the depth camera module 1 captures video information from the scene of the car washing site and transmits it to the local motion control system 3.
  • the local motion control system 3 is electrically connected with the execution system 5, and the information of each walking unmanned car washing machine is The interaction is a parallel structure.
  • the local motion control system 3 and the cloud control system 4 exchange information through wireless modules, and the execution system 5 and the local motion control system 3 can carry out two-way information dissemination, and the execution system 5 can move and operate the machine by Further change the real-time car wash scene to form a control closed-loop system.
  • the local motion control system 3 is integrated with a high-performance mainboard 6 , a drive module 7 and a monitoring module 8 .
  • the output end of the high-performance mainboard 6 is electrically connected to the input end of the drive module 7
  • the output end of the drive module 7 is electrically connected to the input end of the execution system 5 .
  • the drive module 7 is used to receive operation instructions and drive each actuator to move
  • the output end of the execution system 5 is electrically connected to the input end of the monitoring module 8
  • the output end of the monitoring module 8 is electrically connected to the input end of the high-performance motherboard 6, and the monitoring module 8 Including other sensors to assist the machine to obtain other information
  • the high-performance motherboard 6 sends the signal given by the cloud control system 4 to the drive module 7 to drive the execution system 5 for execution
  • the monitoring module 8 monitors the state parameters of the execution system 5 in real time, and Transfer it back to the high-performance motherboard 6.
  • the high-performance motherboard 6 includes an upper computer functional module 9 and a lower computer functional module 10.
  • the upper computer functional part is used to receive the video information captured by the depth camera and perform identification processing to obtain the operation instructions that the machine needs to execute.
  • Sent to the functional part of the lower computer; the functional part of the lower computer is used to receive the motion command issued by the functional part of the main board upper computer and the cloud control system 4, and process it as the operation command that needs to be executed by different drive mechanisms in the drive module 7, And send it to the drive module 7, the upper computer function module 9 has more excellent processing performance than the lower computer function module 10, if the performance of the main board is insufficient due to the pursuit of a higher precision algorithm, you can consider the upper computer function module.
  • the upper computer function module 9 is separated from the lower computer function module 10, the upper computer function module 9 uses a CPU with better performance for data processing, and the lower computer function module 10 uses ordinary PLC or single chip, which can greatly save manufacturing costs, the upper computer function module 9
  • the interior is equipped with vision processing algorithms and driving instructions; the processing algorithms include the recognition of vehicle models, the calculation of the distance from the machine to the vehicle, and the recognition of the vehicle attitude angle.
  • the driving instructions include the movement of the AGV chassis, the movement attitude of the brush, and the door structure. opening and closing, etc.
  • the output end of the depth camera is electrically connected to the input end of the upper computer function module 9, and the video information collected by the depth camera is specifically transmitted to the upper computer function module 9 of the high-performance mainboard 6 for advanced algorithm operation processing, and the upper computer function module 9. It is connected with the lower computer function module 10 so that the information can be transmitted to each other, and the transmitted information includes motion task signals, machine real-time status, etc.
  • the output end of the cloud control system 4 is electrically connected to the input end of the lower computer function module 10, and the control issued by the cloud control system 4 The signal is sent to the lower computer function module 10, and the lower computer function module 10 re-coordinates the collected signal, sends the drive signal to the drive module 7 and drives the execution system 5, and the monitoring module 8 monitors the execution system 5. The state parameters are sent back to the lower computer function module 10.
  • the depth camera module 1 is installed in the depth camera, the depth camera module 1 is directly connected with the high-performance motherboard 6 on the local motion control system 3 through the line, and the depth camera is directly installed under the top cover of the walking unmanned car washing machine, so as to The visual information of the direction facing the vehicle is captured, including model information, machine-to-vehicle angle information, machine-to-vehicle distance information, and vehicle attitude information.
  • the site camera module 2 consists of a plurality of cameras in an array and is evenly installed around the site to ensure that the complete information of the car wash site can be captured.
  • the shooting information includes the position and attitude of each walking unmanned car washing machine, and the position and attitude of the vehicle. , markings and signs on the venue, and information on the flow of people and objects on the venue.
  • the cloud control system 4 is equipped with a visual recognition algorithm and a scheduling planning algorithm.
  • the visual recognition algorithm includes recognizing the position and attitude of the vehicle, recognizing the position and attitude of each machine, and tracking the movement path of each machine.
  • the scheduling and planning algorithm includes calculating different The path that the machine needs to travel, the path point that the machine will reach in the next step according to the current position of the machine, the path that the car needs to travel through and the attitude that the car needs to reach based on the current position of the car, and the guidance command is issued.
  • the present application relates to a dual visual feedback motion control system based on a walking unmanned car washing machine.
  • a depth camera installed under the top cover of the walking unmanned car washing machine captures visual information facing the direction of the vehicle, including model information, Machine-to-vehicle angle information, machine-to-vehicle distance information, and vehicle attitude information, etc., and at the same time, the collected video information is transmitted to the upper computer function module 9 of the high-performance motherboard 6 in the local motion control system 3 for advanced algorithm calculation processing,
  • the processing algorithm includes the identification of the vehicle model, the calculation of the distance from the machine to the vehicle, and the identification of the vehicle attitude angle, to obtain the operation instructions that the machine needs to execute and send the instructions to the lower computer function module 10.
  • the site cameras evenly installed around the site take pictures of the car's position, attitude and other information on the site, as well as the position, attitude and other information of multiple car wash robots, and upload the information to the cloud control system 4.
  • the cloud control system 4 passes the visual recognition algorithm. Identify the position and attitude of the vehicle, identify the position and attitude of each machine, track the movement path of each machine, and use the scheduling planning algorithm to calculate the path that different machines need to travel, and calculate the path point that the machine will reach in the next step according to the current position of the machine.
  • Re-coordination processing process it into the operation instructions that different drive mechanisms in the drive module 7 need to execute, and then send it to the drive module 7 and drive the execution system 5, so that the execution system 5 drives each execution mechanism to perform movements such as car washing, moving, etc., and at the same time
  • the monitoring module 8 transmits the monitored state parameters of the execution system 5 back to the lower-level computer function module 10, so that the execution system 5 and the local motion control system 3 carry out two-way information dissemination, and the execution system 5 further moves and operates the machine through the movement and operation of the machine.
  • the real-time car washing scene is changed to form a control closed-loop system.
  • the present invention adopts dual acquisition modules to collect the microscopic visual information of the vehicle direction and the macroscopic information of the complete car washing site, and calculates the path that the car needs to pass and the attitude that needs to be achieved by the current position of the car. And issue guidance instructions so that it will not collide with the car washing water equipment.
  • the car washing system can first use software/electronic screen/voice and other methods to guide the vehicle to enter a suitable area with a suitable attitude , and then dispatch the walking unmanned car wash equipment to carry out the coordinated car wash operation of the vehicle.

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Aviation & Aerospace Engineering (AREA)
  • Radar, Positioning & Navigation (AREA)
  • Remote Sensing (AREA)
  • General Physics & Mathematics (AREA)
  • Automation & Control Theory (AREA)
  • Mechanical Engineering (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Multimedia (AREA)
  • Electromagnetism (AREA)
  • Control Of Position, Course, Altitude, Or Attitude Of Moving Bodies (AREA)

Abstract

一种基于行走无人洗车设备的双视觉反馈运动控制系统，涉及无人洗车领域，其可模仿人类绕车移动的洗车方式，能够在更多的场合进行洗车操作，且相较于龙门式洗车机备更加灵活，能够在闲置的时候入库保存，降低维护成本；该控制系统包括深度摄像头模块（1）、场地摄像头模块（2）、本地运动控制系统（3）以及云端控制系统（4），深度摄像头模块（1）安装在行走无人洗车机器上并与本地运动控制系统（3）通过线路连接，本地运动控制系统（3）集成在行走无人洗车机器的电柜上，场地摄像头模块（2）安装在洗车场地的周边，云端控制系统（4）搭载在云端服务器上并与本地运动控制系统（3）和场地摄像头模块（2）无线连接。

Description

基于行走无人洗车设备的双视觉反馈运动控制系统 技术领域
本发明涉及无人洗车领域，更具体地涉及基于行走无人洗车设备的双视觉反馈运动控制系统。
背景技术
无人洗车机，指的是无需洗车人员介入即可全自动完成车辆清洗的设备，由于其具有清洗成本低效率高等优势，采用无人洗车机进行车辆清洗正慢慢成为有车一族的首选。
市面上常见的无人洗车机其本身结构存在一定的限制，导致其在实际应用中依旧存在局限。具体的，常见的无人洗车机为龙门式洗车机，龙门式洗车机包括固定的行走轨道，置于轨道上的龙门架以及置于龙门架内的各类洗车设备，其需要被固定在特定场所供有车一族使用，但是这种洗车机体积庞大，建造成本和维护成本高昂，洗车方式单一，缺少灵动性。
发明内容
基于此，本申请提供一种基于行走无人洗车设备的双视觉反馈运动控制系统，其可使系统控制下的无人洗车设备在洗车时模仿人类进行绕车移动，能够在更多的场合进行洗车操作，且相较于龙门式洗车机备更加灵活，能够在闲置的时候入库保存，降低维护成本。
一种基于行走无人洗车设备的双视觉反馈运动控制系统，包括深度摄像头模块、场地摄像头模块、本地运动控制系统以及云端控制系统，所述深度摄像头模块安装在行走无人洗车机器上并与本地运动控制系统通过线路连接，本地运动控制系统集成在行走无人洗车机器的电柜上，所述场地摄像头模块安装在洗车场地的周边，所述云端控制系统搭载在云端服务器上并与本地运动控制系统和场地摄像头模块无线连接，云端控制系统分别通过第一无线模块和第二无线模块与本地运动控制系统和场地摄像头模块进行通讯。
优选的，所述每个洗车场地至少配备2台行走无人洗车机器，每台行走无 人洗车机器均有深度摄像头模块、本地运动控制系统和执行系统三个部分，深度摄像头模块与本地运动控制系统电连接，深度摄像头模块从洗车场地的场景摄取视频信息传递给本地运动控制系统，本地运动控制系统与执行系统电连接，所述每台行走无人洗车机器的信息交互为并联结构，本地运动控制系统与云端控制系统通过无线模块进行信息的往来，执行系统与本地运动控制系统可以进行双向的信息传播。
优选的，所述本地运动控制系统集成有高性能主板、驱动模块和监测模块，高性能主板输出端与驱动模块输入端电连接，驱动模块输出端与执行系统输入端电连接，驱动模块用于接收操作指令并且驱动各个执行机构进行运动，执行系统输出端与监测模块输入端电连接，监测模块输出端与高性能主板输入端电连接，监测模块包括其他传感器以辅助机器获取其他信息，高性能主板将云端控制系统给予的信号下发至驱动模块驱动执行系统进行执行，而监测模块则实时监测执行系统的状态参量，并将其传送回高性能主板。
优选的，所述高性能主板包括上位机功能模块和下位机功能模块，下位机功能模块性能低于上位机功能模块，所述上位机功能模块内部搭载有视觉处理算法和驱动指令，处理算法包括对车辆型号的识别、对机器到车辆距离的计算、对车辆姿态角度的识别，驱动指令包括AGV底盘的运动，毛刷的运动姿态，门结构的开合等。
优选的，所述深度摄像头输出端与上位机功能模块输入端电连接，深度摄像头所采集的视频信息具体传递给了高性能主板的上位机功能模块进行高级算法运算处理，上位机功能模块与下位机功能模块电连接，上位机功能模块与下位机功能模块的信息能够相互传递，传递信息包括运动任务信号、机器实时状态等，云端控制系统输出端与下位机功能模块输入端电连接，云端控制系统下达的控制信号发送到下位机功能模块。
优选的，所述深度摄像头模块安装在深度摄像头内，深度摄像头模块与本地运动控制系统上的高性能主板通过线路直接连接，所述深度摄像头直接安装在行走无人洗车机器的顶盖下方，以拍摄到正对车辆方向的视觉信息，包括有车型信息、机器到车辆的角度信息，机器到车辆的距离信息、车辆的姿态信息。
优选的，所述场地摄像头模块由多个摄像头组成阵列并均匀安装在场地周边，对各个行走无人洗车机器的位置与姿态、车辆的位置与姿态、场地上的标线与标识、场地上的人员物体流动信息进行拍摄。
优选的，所述云端控制系统上搭载有视觉识别算法与调度规划算法，视觉识别算法包括识别车辆的位置和姿态、识别各个机器的位置和姿态、跟踪各个机器的运动路径，调度规划算法包括计算不同机器需要走过的路径、根据机器当前的位置计算机器下一步所要到达的路径点、通过汽车当前的位置计算汽车需要经过的路径和需要达到的姿态并下发引导指令。
相较现有技术，本发明具有以下的特点和有益效果：
1、通过行走无人洗车设备上的深度摄像机模块拍摄其周围视频信息并传送给本地运动控制系统，本地运动控制系统根据信息进行算法运算以获得在该位置上行走无人洗车设备需要执行的动作以及运动的方式，同时通过场地摄像头模块对汽车在场地上的位置、姿态以及复数行走无人洗车设备的位置、姿态等场地信息进行拍摄，然后将该信息上传至云端控制系统，云端控制系统通过算法以调度规划，并下发调度信息进入各个行走无人洗车设备的本地运动控制系统，以决定行走无人洗车设备下一步的运动；
2、通过采用双采集模块，同时采集车辆方向视觉信息(微观)和洗车场地的完整信息(宏观)，从而可通过汽车当前的位置计算汽车需要经过的路径和需要达到的姿态并下发引导指令，使其不会和洗车水设备冲撞，相较于龙门式洗车机备更加灵活，能够在闲置的时候入库保存，降低维护成本。
附图说明
为了更清楚地说明本申请实施例或相关技术中的技术方案，下面将对实施例描述中所需要使用的附图作简要介绍，显而易见地，下面描述中的附图仅仅是本申请的一些实施例，对于本领域普通技术人员来讲，在不付出创造性劳动的前提下，还可以根据这些附图获得其他的附图。
图1为基于行走无人洗车设备的双视觉反馈运动系统整体框图；
图2为本地运动控制系统的信息框图。
图中标号说明：1、深度摄像头模块；2、场地摄像头模块；3、本地运动控 制系统；4、云端控制系统；5、执行系统；6、高性能主板；7、驱动模块；8、监测模块；9、上位机功能模块；10、下位机功能模块。
具体实施方式
为了使本申请的目的、技术方案及优点更加清楚明白，以下结合附图及实施例，对本申请进行进一步详细说明。应当理解，此处描述的具体实施例仅仅用以解释本申请，并不用于限定本申请。基于本申请中的实例，本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其他实例，都属于本申请保护的范围。
本领域技术人员应理解的是，除非另作定义，权利要求书和说明书中使用的技术术语或者科学术语应当为本申请所属技术领域内具有一般技能的人士所理解的通常意义。“连接”或者“相连”等类似的词语并非限定于物理的或者机械的连接，而是可以包括电气的连接，不管是直接的还是间接的；“和/或”，描述关联对象的关联关系，表示可以存在三种关系；术语“纵向”、“横向”、“上”、“下”、“前”、“后”、“左”、“右”、“竖直”、“水平”、“顶”、“底”“内”、“外”等指示的方位或位置关系是基于附图所示的方位或位置关系，其仅是为了便于描述本实用新型和简化描述，而不是指示或暗示所指的装置或元件必须具有特定的方位、以特定的方位构造和操作，因此上述术语不能理解为对本实用新型的限制。
显而易见地，下面描述中的附图仅仅是本申请的一些示例或实施例，对于本领域的普通技术人员来讲，在不付出创造性劳动的前提下，还可以根据这些附图将本申请应用于其他类似情景。此外，还可以理解的是，虽然这种开发过程中所作出的努力可能是复杂并且冗长的，然而对于与本申请公开的内容相关的本领域的普通技术人员而言，在本申请揭露的技术内容的基础上进行的一些设计，制造或者生产等变更只是常规的技术手段，不应当理解为本申请公开的内容不充分。
请参照图1，一种基于行走无人洗车设备的双视觉反馈运动控制系统，包括深度摄像头模块1、场地摄像头模块2、本地运动控制系统3以及云端控制系统4，深度摄像头模块1安装在行走无人洗车机器上并与本地运动控制系统3通过 线路连接，本地运动控制系统3集成在行走无人洗车机器的电柜上，深度摄像机模块拍摄视频信息进入本地运动控制系统3，本地运动控制系统3根据信息进行算法运算以获得在该位置上行走无人洗车设备需要执行的动作以及运动的方式；场地摄像头模块2安装在洗车场地的周边，云端控制系统4搭载在云端服务器上并与本地运动控制系统3和场地摄像头模块2无线连接，场地摄像头模块2对汽车在场地上的位置、姿态以及复数行走无人洗车设备的位置、姿态等场地信息进行拍摄，然后将该信息上传至云端控制系统4，云端控制系统4分别通过第一无线模块和第二无线模块与本地运动控制系统3和场地摄像头模块2进行通讯，云端控制系统4通过算法以调度规划，下发调度信息进入各个行走无人洗车设备的本地运动控制系统3，以决定行走无人洗车设备下一步的运动，同时进行收集并分析处理本地运动系统中所发送的监测信息与场地摄像头模块2采集到的视频流信息，并向本地运动控制系统3发送任务信息。
请参照图2，每个洗车场地至少配备2台行走无人洗车机器，每台行走无人洗车机器均有深度摄像头模块1、本地运动控制系统3和执行系统5三个部分，深度摄像头模块1与本地运动控制系统3电连接，深度摄像头模块1从洗车场地的场景摄取视频信息传递给本地运动控制系统3，本地运动控制系统3与执行系统5电连接，每台行走无人洗车机器的信息交互为并联结构，本地运动控制系统3与云端控制系统4通过无线模块进行信息的往来，执行系统5与本地运动控制系统3可以进行双向的信息传播，并且执行系统5通过进行机器的运动和操作进一步地改变实时洗车场景，以形成一个控制闭环系统。
具体的，本地运动控制系统3集成有高性能主板6、驱动模块7和监测模块8，高性能主板6输出端与驱动模块7输入端电连接，驱动模块7输出端与执行系统5输入端电连接，驱动模块7用于接收操作指令并且驱动各个执行机构进行运动，执行系统5输出端与监测模块8输入端电连接，监测模块8输出端与高性能主板6输入端电连接，监测模块8包括其他传感器以辅助机器获取其他信息，高性能主板6将云端控制系统4给予的信号下发至驱动模块7驱动执行系统5进行执行，而监测模块8则实时监测执行系统5的状态参量，并将其传送回高性能主板6。
另外，高性能主板6包括上位机功能模块9和下位机功能模块10，上位机功能部分用于接收深度摄像头拍摄到的视频信息并进行识别处理，得到机器所需要执行的操作指令，并将指令下发至下位机功能部分；下位机功能部分用于接收主板上位机功能部分和云端控制系统4下发的运动指令，并将其处理为驱动模块7中不同驱动机构所需要执行的操作指令，并将其下发至驱动模块7，上位机功能模块9相较于下位机功能模块10的处理性能更加卓越，若因追求精度更高的算法而导致主板性能不足，可以考虑将上位机功能模块9与下位机功能模块10拆分开，上位机功能模块9使用性能更加优秀的CPU进行数据处理，下位机功能模块10使用普通PLC或者单片机，这样可以大幅度节约制造成本，上位机功能模块9内部搭载有视觉处理算法和驱动指令；处理算法包括对车辆型号的识别、对机器到车辆距离的计算、对车辆姿态角度的识别，驱动指令包括AGV底盘的运动，毛刷的运动姿态，门结构的开合等。
进一步的，深度摄像头输出端与上位机功能模块9输入端电连接，深度摄像头所采集的视频信息具体传递给了高性能主板6的上位机功能模块9进行高级算法运算处理，上位机功能模块9与下位机功能模块10连接使其信息能够相互传递，传递信息包括运动任务信号、机器实时状态等，云端控制系统4输出端与下位机功能模块10输入端电连接，云端控制系统4下达的控制信号发送到下位机功能模块10，下位机功能模块10将收集到的信号进行再协调处理，将驱动信号下发至驱动模块7并驱动执行系统5，同时监测模块8将监测到的执行系统5的状态参量传送回下位机功能模块10。
具体的，深度摄像头模块1安装在深度摄像头内，深度摄像头模块1与本地运动控制系统3上的高性能主板6通过线路直接连接，深度摄像头直接安装在行走无人洗车机器的顶盖下方，以拍摄到正对车辆方向的视觉信息，包括有车型信息、机器到车辆的角度信息，机器到车辆的距离信息、车辆的姿态信息。
具体的，场地摄像头模块2由多个摄像头组成阵列并均匀安装在场地周边，以确保能够拍摄到洗车场地的完整信息，拍摄信息包括各个行走无人洗车机器的位置与姿态、车辆的位置与姿态、场地上的标线与标识、场地上的人员物体流动信息。
具体的，云端控制系统4上搭载有视觉识别算法与调度规划算法，视觉识别算法包括识别车辆的位置和姿态、识别各个机器的位置和姿态、跟踪各个机器的运动路径，调度规划算法包括计算不同机器需要走过的路径、根据机器当前的位置计算机器下一步所要到达的路径点、通过汽车当前的位置计算汽车需要经过的路径和需要达到的姿态并下发引导指令。
本申请涉及的一种基于行走无人洗车设备的双视觉反馈运动控制系统，使用时通过安装在行走无人洗车机器顶盖下方的深度摄像头拍摄到正对车辆方向的视觉信息，包括车型信息、机器到车辆的角度信息、机器到车辆的距离信息及车辆的姿态信息等，同时将采集的视频信息传递给了本地运动控制系统3中高性能主板6的上位机功能模块9进行高级算法运算处理，处理算法包括对车辆型号的识别、对机器到车辆距离的计算、对车辆姿态角度的识别，得到机器所需要执行的操作指令并将指令下发至下位机功能模块10，同时在上述过程中，均匀安装在场地周边的场地摄像机对汽车在场地上的位置、姿态等信息以及复数洗车机器人的位置、姿态等信息进行拍摄，并将该信息上传至云端控制系统4，云端控制系统4通过视觉识别算法识别车辆的位置和姿态、识别各个机器的位置和姿态、跟踪各个机器的运动路径，并利用调度规划算法计算不同机器需要走过的路径、根据机器当前的位置计算机器下一步所要到达的路径点、通过汽车当前的位置计算汽车需要经过的路径和需要达到的姿态并下发引导指令，下位机功能模块10在接收到上位机功能模块9和云端控制系统4下发的指令后进行再协调处理，将其处理为驱动模块7中不同驱动机构所需要执行的操作指令，然后将其下发至驱动模块7并驱动执行系统5，使执行系统5驱动各个执行机构进行洗车、移动等运动，同时监测模块8将监测到的执行系统5的状态参量传送回下位机功能模块10，使得执行系统5与本地运动控制系统3进行双向的信息传播，并且执行系统5通过进行机器的运动和操作进一步地改变实时洗车场景，以形成一个控制闭环系统，本发明采用双采集模块对车辆方向视觉信息微观和洗车场地的完整信息宏观进行采集，通过汽车当前的位置计算汽车需要经过的路径和需要达到的姿态并下发引导指令，使其不会和洗车水设备冲撞，在上述洗车流程中，可先通过洗车系统利用软件/电子屏/语音等方式，引导车辆以 一个合适的姿态驶入一个合适的区域，随后再调度行走无人洗车设备对车辆进行联动协同洗车操作。
以上实施例的各技术特征可以进行任意的组合，为使描述简洁，未对上述实施例中的各个技术特征所有可能的组合都进行描述，然而，只要这些技术特征的组合不存在矛盾，都应当认为是本说明书记载的范围。
以上所述实施例仅表达了本申请的几种实施方式，其描述较为具体和详细，但并不能因此而理解为对本申请专利范围的限制。应当指出的是，对于本领域的普通技术人员来说，在不脱离本申请构思的前提下，还可以做出若干变形和改进，这些都属于本申请的保护范围。因此，本申请专利的保护范围应以所附权利要求为准。

Claims (8)

  1. 基于行走无人洗车设备的双视觉反馈运动控制系统，其特征在于：包括深度摄像头模块(1)、场地摄像头模块(2)、本地运动控制系统(3)以及云端控制系统(4)，所述深度摄像头模块(1)和本地运动控制系统(3)均安装在行走无人洗车机器上且二者通过线路连接，场地摄像头模块(2)安装在洗车场地周边，所述云端控制系统(4)搭载在云端服务器上并与本地运动控制系统(3)和场地摄像头模块(2)无线连接。
  2. 根据权利要求1所述的基于行走无人洗车设备的双视觉反馈运动控制系统，其特征在于：所述每个洗车场地至少配备2台行走无人洗车机器，每台行走无人洗车机器均有深度摄像头模块(1)、本地运动控制系统(3)和执行系统(5)三个部分，深度摄像头模块(1)与本地运动控制系统(3)电连接，本地运动控制系统(3)与执行系统(5)电连接。
  3. 根据权利要求2所述的基于行走无人洗车设备的双视觉反馈运动控制系统，其特征在于：所述本地运动控制系统(3)集成有高性能主板(6)、驱动模块(7)和监测模块(8)，高性能主板(6)输出端与驱动模块(7)输入端电连接，驱动模块(7)输出端与执行系统(5)输入端电连接，执行系统(5)输出端与监测模块(8)输入端电连接，监测模块(8)输出端与高性能主板(6)输入端电连接。
  4. 根据权利要求3所述的基于行走无人洗车设备的双视觉反馈运动控制系统，其特征在于：所述高性能主板(6)包括上位机功能模块(9)和下位机功能模块(10)，且下位机功能模块(10)性能低于上位机功能模块(9)，所述上位机功能模块(9)内部搭载有用于对车辆型号进行识别、对机器到车辆距离进行计算、对车辆姿态角度进行识别的视觉处理算法和驱动指令。
  5. 根据权利要求4所述的基于行走无人洗车设备的双视觉反馈运动控制系统，其特征在于：所述深度摄像头输出端与上位机功能模块(9)输入端电连接，上位机功能模块(9)与下位机功能模块(10)电连接，云端控制系统(4)输出端与下位机功能模块(10)输入端电连接。
  6. 根据权利要求1所述的基于行走无人洗车设备的双视觉反馈运动控制系统，其特征在于：所述深度摄像头模块(1)安装在深度摄像头内，深度摄像头 直接安装在行走无人洗车机器的顶盖下方。
  7. 根据权利要求1所述的基于行走无人洗车设备的双视觉反馈运动控制系统，其特征在于：所述场地摄像头模块(2)由多个摄像头组成阵列并均匀安装在场地周边。
  8. 根据权利要求1所述的基于行走无人洗车设备的双视觉反馈运动控制系统，其特征在于：所述云端控制系统(4)上搭载有视觉识别算法与调度规划算法。
PCT/CN2020/129958 2020-08-06 2020-11-19 基于行走无人洗车设备的双视觉反馈运动控制系统 WO2022027865A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202010781444.9 2020-08-06
CN202010781444.9A CN111752284B (zh) 2020-08-06 2020-08-06 基于行走无人洗车设备的双视觉反馈运动控制系统

Publications (1)

Publication Number Publication Date
WO2022027865A1 true WO2022027865A1 (zh) 2022-02-10

Family

ID=72713152

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2020/129958 WO2022027865A1 (zh) 2020-08-06 2020-11-19 基于行走无人洗车设备的双视觉反馈运动控制系统

Country Status (2)

Country Link
CN (1) CN111752284B (zh)
WO (1) WO2022027865A1 (zh)

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111752284B (zh) * 2020-08-06 2021-09-07 浙江驿公里智能科技有限公司 基于行走无人洗车设备的双视觉反馈运动控制系统
CN114185350B (zh) * 2021-12-06 2022-09-23 浙江驿公里智能科技有限公司 基于角度编码器的连轴洗车设备的路径控制方法及系统

Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8696827B2 (en) * 2010-12-01 2014-04-15 Whirlpool Corporation Dishwasher with imaging device for measuring load characteristics and a method for controlling same
CN203780498U (zh) * 2014-01-29 2014-08-20 郑州大学 基于双目相机三维成像的自动洗车设备
CN205292570U (zh) * 2015-11-26 2016-06-08 张致琛 一种智能洗车设备
CN106563658A (zh) * 2015-10-09 2017-04-19 苏州宝时得电动工具有限公司 自动清洗设备
WO2018007846A1 (en) * 2016-07-04 2018-01-11 Montaseri Poorya Green automated robotic personal car wash
CN207029127U (zh) * 2017-06-20 2018-02-23 艾合麦提·麦麦提 一种基于图像识别的移动式自动洗车机器人
CN108490830A (zh) * 2018-03-16 2018-09-04 合肥工业大学 一种基于机器视觉的自动洗车的控制装置及其控制方法
CN110216690A (zh) * 2019-05-31 2019-09-10 广州大学 一种洗车机器人
CN110315021A (zh) * 2019-04-22 2019-10-11 中信戴卡股份有限公司 一种车轮锻造模具自适应自动清理装置
CN111752284A (zh) * 2020-08-06 2020-10-09 浙江驿公里智能科技有限公司 基于行走无人洗车设备的双视觉反馈运动控制系统

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110901597A (zh) * 2019-11-14 2020-03-24 安徽新博普曼智能科技股份有限公司 一种基于机器视觉的智能无人洗车方法
CN211075841U (zh) * 2019-12-10 2020-07-24 浙江驿公里智能科技有限公司 一种利用超声波传感技术的防撞无人洗车机

Patent Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8696827B2 (en) * 2010-12-01 2014-04-15 Whirlpool Corporation Dishwasher with imaging device for measuring load characteristics and a method for controlling same
CN203780498U (zh) * 2014-01-29 2014-08-20 郑州大学 基于双目相机三维成像的自动洗车设备
CN106563658A (zh) * 2015-10-09 2017-04-19 苏州宝时得电动工具有限公司 自动清洗设备
CN205292570U (zh) * 2015-11-26 2016-06-08 张致琛 一种智能洗车设备
WO2018007846A1 (en) * 2016-07-04 2018-01-11 Montaseri Poorya Green automated robotic personal car wash
CN207029127U (zh) * 2017-06-20 2018-02-23 艾合麦提·麦麦提 一种基于图像识别的移动式自动洗车机器人
CN108490830A (zh) * 2018-03-16 2018-09-04 合肥工业大学 一种基于机器视觉的自动洗车的控制装置及其控制方法
CN110315021A (zh) * 2019-04-22 2019-10-11 中信戴卡股份有限公司 一种车轮锻造模具自适应自动清理装置
CN110216690A (zh) * 2019-05-31 2019-09-10 广州大学 一种洗车机器人
CN111752284A (zh) * 2020-08-06 2020-10-09 浙江驿公里智能科技有限公司 基于行走无人洗车设备的双视觉反馈运动控制系统

Also Published As

Publication number Publication date
CN111752284B (zh) 2021-09-07
CN111752284A (zh) 2020-10-09

Similar Documents

Publication Publication Date Title
WO2022027865A1 (zh) 基于行走无人洗车设备的双视觉反馈运动控制系统
CN104111655B (zh) 一种基于远程控制的智能家居服务机器人系统
CN102280826B (zh) 变电站智能机器人巡检系统及巡检方法
CN102354174B (zh) 基于变电站移动式巡检装置的巡检系统及其巡检方法
CN108297098A (zh) 人工智能驱动的机器人控制系统及方法
CN109807887B (zh) 基于深度神经网络的柔性臂智能感知与控制方法和系统
CN205080433U (zh) 一种实时远程移动的智能机器人监控系统
CN104820418A (zh) 一种针对机械臂的嵌入式视觉系统及其使用方法
CN204856222U (zh) 一种基于物联网技术的全向移动平台控制系统
Dumbre et al. Robotic vehicle control using internet via webpage and keyboard
CN112123338A (zh) 一种支持深度学习加速的变电站智能巡检机器人系统
CN208522891U (zh) 一种区域监控全景摄像系统
CN205693804U (zh) 一种无线网络视频监控装置
CN202261627U (zh) 智能跟踪网络球型摄像机
CN203984582U (zh) 一种基于变电站作业人员图像捕捉定位网络传输装置
CN116476074A (zh) 基于混合现实技术的远程机械臂操作系统及人机交互方法
CN107121988A (zh) 一种机械臂抓取无人机电池的方式
CN107089483A (zh) 智能上料系统及智能上料方法
CN111522347A (zh) 一种远程信息交互机器人
CN206578822U (zh) 一种子母构型的索并联机器人
Babu et al. Subject Tracking with Camera Movement Using Single Board Computer
CN206773458U (zh) 多智能体协调控制实验平台
CN113325815B (zh) 用于监测诊断工业现场环境和设备的dcs系统
CN202143149U (zh) 空中监视系统
Ye et al. The Application of Artificial Intelligence in Indoor Positioning and Scene Reconstruction of Robots

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: 20948843

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 20948843

Country of ref document: EP

Kind code of ref document: A1