As an important part of smart factories, robots have already played a role in many advanced steel companies.
Robots are becoming an indispensable smart engine for steel companies
Currently, there is still much room for improvement in the robot density of China’s Baowu Steel Industry. China Baowu will strive to realize unmanned and less-manned on-site operations, improve intrinsic safety levels, improve labour efficiency, and promote job changes.
Compared with traditional inspection robots (wheeled and crawler), quadruped robots have discrete footholds and can adapt to different terrains such as stairs, steps, slopes, and mountains. At the same time, they are equipped with GPS (global positioning system), robotic arms, and lidar. and other various instrument testing equipment, which can complete a series of tasks (smart construction, detection and search and rescue, security patrol, etc.), and provide inspection and exploration, material transportation and other work and services for the petrochemical industry, electric power, railway, and mineral collection. Its innovation is reflected in the IP68 protection of the whole machine, which is suitable for indoor and outdoor all-weather scenarios. It is equipped with multiple sets of depth cameras, high-computing power AI (artificial intelligence) chips, and high-torque joints, so it has a longer life, lower cost, and higher production efficiency. The process is simplified and heat dissipation is better.
Promotion in the steel industry can be concentrated in processes such as production line inspections, inspection and sample delivery.
On the one hand, the environmental information collection sensors and cameras carried by the autonomous cruising four-legged robot dog are used to analyze and collect key equipment and environmental parameters of the unit to achieve production line inspections; on the other hand, the four-legged robot dog is used to have certain capabilities. The load-bearing performance and autonomous action capability of planned route autonomous navigation can realize timely delivery of unit samples for inspection.
“The intelligent operation control system in front of the blast furnace is an intelligent production solution in front of the furnace derived from Baoluo’s 3D robot service.
This control system integrates the original operations in front of the furnace such as opening, blocking, mud addition, brazing replacement, temperature measurement and sampling. Through Baoluo Cloud platform data upload, cloud-edge collaboration, model analysis, remote operation and maintenance, etc., the furnace can be Standardization, simplification, and standardization of front-end work, improvement of the working environment, and reduction of work intensity.
Taking the remote control opening and closing as an example, the opening machine and mud cannon are the basic equipment in front of the blast furnace. On the basis of necessary modifications and upgrades such as adding encoders, a PLC (programmable logic controller) control system, The hydraulic solenoid valve station and a wireless remote control transceiver system enable the operator to hold the wireless remote control to open the plug, which enhances the operability and safety of the plug. On this basis, based on the expert opening and closing control system in front of the furnace, and adding remote video monitoring devices, taphole laser marking devices, opening and closing process information feedback systems, fault diagnosis and alarm systems, etc., it can be realized in the remote centralized control room Intelligent one-click opening and closing operations, one-click pre-drilling, secondary mud drilling and other operations.
Based on the China Baowu Industrial Internet Platform, robots and information technology are deeply integrated to realize robot interconnection, so that robots can continue to perform deep learning and adapt to more diverse production environments.
The medium-thick plate finishing and sampling robot system also has a certain influence on the industry.
The system includes a visual dynamic recognition and positioning subsystem, a robot sampling subsystem, an automatic cutting subsystem, and a truss library subsystem.
The spline picking robot system in the robot sampling system uses a high-load model robot, which is used to pick up the spline using the electromagnet at the claw end of the robot and swing the spline smoothly under the guidance of the spline position information recognized by the vision system. Place it in the spline storage station and wait for subsequent actions.
Through the application of the medium-thick plate finishing sampling robot system, the effects of reducing personnel labour intensity, reducing the probability of sampling errors, and improving the efficiency of the plate finishing process have been achieved.
Intelligent systems based on machine vision have become a “good helper” for steel companies to improve efficiency
Machine vision uses machines to replace human eyes for measurement and judgment. In intelligent systems, it often assumes the basic function of data acquisition.
Due to problems such as multiple production lines, dispersed meter inspection equipment, heavy reliance on human judgment, and excess machine resources, there is a huge market demand for centralized meter inspection control and automatic judgment based on 5G+AI for multiple production lines. The 5G+AI meter inspection centralized control system transmits multiple sets of meter inspection data in the production process through the 5G network, performs AI defect identification in the edge processing centre, and implements iterative training of the model in the cloud. On this basis, a multi-process integration is established Judgment, analysis and abnormal traceability of chemical surface quality. By applying this system to a steel company, the number of meter inspection and quality inspection positions was reduced from 18 to 4, the workload of maintenance personnel was reduced by 50%, and server deployment was reduced by 75%, achieving upstream and downstream collaborative digital quality control.
Asymmetric detection and automatic correction control system for hot rolling operation based on machine vision.
This system realizes automatic monitoring of centerline asymmetry during strip operation by installing visual detection equipment for sickle bends and warping heads before and after the roughing mill of the hot rolling production line, and installing visual detection equipment for strip deviation between the finishing rolling stands. On this basis, by improving the theory of asymmetric elastic-plastic deformation of metal, automatic levelling and correction control of the entire line can be achieved, which can replace more than 60% to 80% of manual operations, significantly reduce the frequency of steel stacking and tail flick, and effectively improve the sickle It improves the accuracy of hitting curved and tower-shaped defects, frees operators from heavy, repetitive and tense levelling work, and transforms it into overall control and optimization of production factors.
Detection of end-face defects of cold-rolled steel strip.
The edge quality of chilled steel coils not only affects the production stability and operating efficiency of each cold rolling process but also affects the final quality of the product. For most steel companies, label identification and end-face defect detection of incoming cold-rolled steel coils are confirmed by manual inspection. During the inspection process, there are potential safety hazards for operators such as falling and being hit by objects; manual inspection involves subjective judgment factors, and there may be misjudgments or missed inspections, which will affect the stable operation of the production line.
In order to solve this problem, HBIS Industrial Technology has realized the identification of inner ring labels of steel coils in narrow spaces and large angles by using OCR (optical character recognition) technology, 3D structured light technology, machine vision guided path planning and deep learning algorithms. , High-precision three-dimensional reconstruction and identification of steel coil end faces, automatic planning of scanning paths for steel coil end faces of different coil diameters, self-learning and self-determination of defect categories, etc.
This project was successfully applied to the galvanizing line of Hesteel and Tangshan Iron and Steel Co., Ltd., using a laser to construct an active light source. Two robots were equipped with line-structured lights and cameras at the same time for circular scanning. The end-face 3D reconstruction was performed based on the scan results, and the end-face defects were completed based on the reconstruction results. detection, effectively improving the defect identification rate.
Source of article: “China Metallurgical News”