[Scientific and Technological Achievements] Putty Automatic Spraying System Based on Vision

With the rapid development of high-speed rail technology in China, the number of EMU trains is increasing, coupled with the promotion of the "Belt and Road Initiative" strategy, the demand for high-speed rail in large and medium-sized cities at home and abroad is increasing rapidly. at the same time, the advanced manufacturing technology of high-speed rail passenger car manufacturing puts forward higher requirements, requiring more stable product quality, more green manufacturing process, higher manufacturing efficiency, more beautiful appearance and so on.

In the past, putty construction mainly used manual scraping brush, but manual scraping brush depends on the experience and state of workers, it is difficult to guarantee the thickness and uniformity of the coating, and it is difficult to quantitatively evaluate the flatness of surface defect repair. Moreover, the coating often contains a large number of organic solvents and toxic low molecular additives, which are harmful to the health of spraying workers.

This issue of "Introduction of Scientific and Technological Achievements" will be introduced by our hospital.Robotics and Automation Equipment Engineering Research Institute(hereinafter referred to as "Robot Institute") independently developedAutomatic spraying system of putty based on vision.

　　

How to achieve automatic spraying?

What changes does a vision-based spray system bring to the job?

What are the advantages compared to the traditional spraying process?

 

Let's get to know each other

Putty automatic spraying system based on vision ~

 

 

science and technology research and development

The automatic spraying system of putty based on vision independently developed by the robot can significantly improve the efficiency and surface paint quality and effectively avoid a series of problems caused by manual operation by using automatic and intelligent technology and equipment and replacing the original scraping brush mode with spraying process. System automation and intelligent technology are highly integrated, the whole spraying process, defect identification, quantitative detection, spraying trajectory planning and spraying operations without manual intervention, the application prospects and future demand market is very broad.

　　

　　Application scenarios and typical metrics

The working object of the system is (3~3.5m)(L)* (3~3.5m)(H) car body test piece made of aluminum alloy (or stainless steel); In the direction perpendicular to the depth of the car body surface, the accuracy of the measurement data is not less than 0.1mm; In the direction parallel to the car body surface, the measurement data can at least detect the depth greater than 0.5mm, and the diameter is greater than 100mm of the recessed area and automatic identification of repair, measurement and control integration.

　　

Main components

 

Automatic Putty Spraying System

Horizontal and vertical motion axis of linkage control

Side wall contour laser scanning system

Lightweight explosion-proof spraying robot

Intelligent off-line programming system

Open Source Robot Control System

　　

Among them, the mechanical body is completely newly designed independently, including linear motion axis, visual camera basic frame and lightweight spraying mechanical arm. The linear motion axis spraying manipulator adopts a 5-axis form. By optimizing the joint axis layout, it not only reduces the number of joints, but also makes the spraying arm lighter in weight and smaller in size, so that it can enter a narrow space more flexibly for spraying operations, and the overall system has a wider and more dexterous operating range. The robot controller adopts open source algorithm technology and is deeply customized and developed according to its own system characteristics.

The motor car body surface pit recognition and measurement system based on machine vision is mainly used to realize the function of car body surface pit recognition and surface reconstruction in the field of putty spraying on the surface of the motor car body. The system is mainly composed of four laser contour sensors. The main functions of the system include: point cloud data acquisition and splicing, 3D reconstruction, base surface extraction and defect point detection, coordinate transformation, etc.

　　

science and technology innovation

Core Technology Introduction

　　1. vision system

The machine vision system is used to realize the three-dimensional topography data acquisition, three-dimensional reconstruction, surface defect point detection and the coordinate unification of the vision system and the robot system.

The 3D profiler is used to scan the surface of the car body, which effectively realizes the collection of the surface topography data of the car body. However, due to the large size and large radian of the car body test piece, the point cloud data obtained by the vision sensor is large and overlapping.

　　

　　Pit extraction

Point cloud block processing, based on random sampling consistency for base surface extraction, while the outside points are saved, after all regional processing, the outside points are clustered and depth, diameter and other analysis, extraction of pits that meet the requirements.

　　three-dimensional reconstruction

Multiple sensors are connected in series and calibrated to realize the sensor coordinate system one, and then realize the splicing of point cloud data, real-time extraction of plane and cylindrical surface, and realize the three-dimensional reconstruction of multi-class surface.

The system uses the form of multiple sensors in series to achieve a wide range of three-dimensional shape measurement, with the help of calibration rod and calibration program to achieve the unity of the coordinate system calibration of multiple sensors, and by solving the position relationship between the sensor coordinate system and the external axis, to achieve the splicing of point cloud data.

At the same time, based on the PCL(Point Cloud Library) point cloud database, the method of point cloud data segmentation processing and the idea of random sampling consistency are used to extract the base surface of the car body. On this basis, the defect points extracted by the segmentation are clustered and analyzed and integrated to provide data support for trajectory planning.

　　

　　2. robot control system

The distributed control system based on high-speed bus is designed independently, and the robot control system, offline programming software and monitoring software are communicated and controlled by industrial standard network, and the signal interaction and control are carried out by special connection with the putty spraying system. The system technology can monitor and control the vision system, putty spraying system, robot control system and offline programming software in real time, and provide important algorithms and control functions such as parameter adjustment, model matching and intelligent optimization.

　　

　　3. Automatic Trajectory Planning Process Execution Technology Based on Spray Model

Based on the depression contour point cloud information obtained by the vision system, the peripheral contour of each layer of spray putty is obtained by point cloud slicing technology, the contour of each layer to be sprayed is obtained by contour bias, and then the spraying path and process parameters are obtained by contour filling technology. Finally, the path output data code of the spraying algorithm is optimized by using the traveling salesman problem model. The generated code can be converted into a robot language for the robot to call and execute.

The process execution system of putty for high-viscosity slurry spraying can establish the corresponding defect model library according to different scenes and equipment parameters, and can match the typical defect model library according to the point cloud data, generate the spraying process parameters and output the execution code. With adaptive process parameter adjustment capability.

　　

Technical advantages

1. Wide applicability, can be applied to a variety of material surface automatic three-dimensional reconstruction and defect detection and process planning, to achieve a high viscosity putty special spraying.

2. Improve the automation and intelligence of the system, change the existing manual identification method, the entire identification, detection, spraying trajectory planning and spraying operations are free of manual interference, and improve the integration and intelligence level of the system.

3. Based on the exquisite design of the mechanical body, the automatic defect identification and repair are realized, and the subsystems are highly integrated to automatically realize the demand process.

4. The measurement system has a simple structure and can realize surface data splicing without pasting markers on the surface of the measured object, which is suitable for spraying and other industries.

5. The user interface is friendly, simple, based on the parameterization of the model, and does not require personnel with special professional skills to use.

6. The use of self-designed integrated control system and putty spraying robot to complete the "defect spraying" process.

　　

application value

　　improve construction efficiency

Using manual operation, a carriage needs 12 people to complete in 3 days. After using automatic spraying equipment, the number of personnel is reduced to two, which can be completed in 3 hours. The production efficiency is 90 times that of manual operation.

　　

　　Reduce raw material loss

The use of continuous automatic spraying can save at least 10%-20% of raw material loss.

　　

　　Cost reduction

Reduce the number of personnel used, greatly reduce labor costs, according to the same production capacity, the use of automated spraying equipment, each year for enterprises to save 50% of the original labor costs.

　　

　　enhance competitiveness

For enterprises with defect filling and large area putty spraying demand, it can significantly improve the intelligent degree of enterprise equipment, improve the working environment and improve the competitiveness of enterprises.

　　

　　

　　Robotics and Automation Equipment Engineering Research Institute

The Institute of Robotics and Automation equipment Engineering was formally established in 2015. It is the intelligent robot and automation equipment achievement and technology transformation and industrial incubation base of the excellent team of Professor Chen Ken of Tsinghua University in Tsinghua Tianjin High-end Institute. It is an innovative intelligent robot and automation equipment technology engineering center integrating development, design, production and sales. Facing the major national civil and national defense science and technology and model engineering requirements for intelligent robots and automation equipment in aerospace, advanced manufacturing, national defense science and technology, medical and health, social life services, and other industrial fields, the team has carried out theoretical and new technology research in the fields of special operation and mobile intelligent robots, humanoid and bionic service robots, and intelligent robot systems, the field of aviation digital manufacturing and assembly ranks among the best in China. It has a wealth of non-standard automation, intelligent equipment research and development and engineering implementation capabilities, and has developed various special-purpose intelligent robot equipment and automation systems.

　　

　　Technical docking and business negotiation

Contact: Rampo

Telephone: 18522169653

E-mail: lanbo@tsinghua-tj.org

　　

　　

　　

Source | Robotics and Automation Equipment Engineering Research Institute

Editor | Zhang Shu

The final interpretation right of the above related technologies belongs to the Institute of Robotics and Automation Equipment Engineering.