As an important hub for logistics connection and production rhythm control in the steel production process, the warehouse area is the basis for the unmanned and intelligent construction of the factory. The traditional steel rolling production workshop warehouse operations mainly rely on manual work. The level of automation is low, the labour capacity is large, the operating environment is harsh, the safety production and occupational health risks are high, the operations are standardized and the production efficiency is low, which seriously restricts the intelligent unmanned steel rolling production line. development bottleneck. The realization of unmanned operation of the crane and intelligent dispatching of the reservoir area will fundamentally change the operation mode of the reservoir area and is of great significance for improving production efficiency, reducing production costs and improving product quality.
From 2013 to 2018, Tangshan Iron and Steel, Baowu, Shougang and other companies carried out research and development of unmanned overhead cranes and intelligent warehouse management technology and achieved application results for steel coil warehouses. However, for a long time, the long steel field, which accounts for more than 45% of total steel production, has never been able to break through the bottleneck of intelligent unmanned technology in the reservoir area.
To achieve unmanned lifting of long materials, electromagnetic spreaders are the most common way. However, it is difficult to stack rods and wires strictly according to the stacking position, and rolling displacement is prone to occur. At this time, according to the positioning method of the steel coil warehouse, it is easy to cause a “virtual hanging” phenomenon where the centre of gravity of the steel is outside the edge of the spreader due to insufficient positioning accuracy. This leads to lifting failure or steel falling accidents during lifting. Therefore, it is necessary to carry out key technical research on unmanned overhead crane lifting of long products.
On the other hand, the intelligent warehouse management system should be able to adapt to the characteristics of long steel lifting, have the ability to automatically adjust and adapt to changes in warehouse locations, and should meet the global dynamic optimal dispatch of overhead cranes, ground logistics equipment, shipping equipment, etc. in large logistics warehouses. demand; at the same time, for the in-factory reshipping, storage, transfer and sales functions of the logistics warehouse area, it is necessary to meet the needs of fully automatic unloading and automatic loading under unmanned operation conditions, and to achieve docking and collaboration with the production and sales systems.
To this end, the Institute of Engineering Technology of the University of Science and Technology Beijing (hereinafter referred to as “Beijing Institute of Engineering Technology”) has innovatively developed a system for the rod and wire storage area that combines storage area environment perception and three-dimensional reconstruction, machine vision and overhead crane control with deep integration. Multi-agent collaborative optimization of intelligent warehouse management scheduling, warehouse area centralized control and smart factory collaborative optimization is the iconic second-generation unmanned overhead cranes and intelligent warehouse management technology and realizes the intelligent warehouse for bars and high-line intelligent warehouses. First demonstration application.
1. How to realize intelligent management and control of long steel (rod and wire) storage areas?
In response to the technical difficulties faced by the long steel intelligent warehouse area, this plan starts from the long steel warehouse area perception reconstruction technology and equipment, machine vision precise positioning control technology and equipment, the warehouse area production logistics optimization scheduling model, and the long steel warehouse area wisdom collection. Starting from several aspects of the control system, a complete set of intelligent management and control technology and equipment for the long steel storage area has been established.
1) Three-dimensional reconstruction of multi-source information and environmental perception technology in the long steel storage area
In view of the characteristics and problems of the metallurgical long steel warehouse area with a large field of view, densely stacked materials in the warehouse area, and diverse delivery vehicles, the institute has developed vehicle and vehicle-borne material identification, warehouse area map construction and feature recognition technology to realize the long steel warehouse area rod The precise positioning of materials, high wire coils, transportation vehicles and onboard materials provides comprehensive and accurate material spatial location information for precise control of unmanned overhead cranes.
(1) Vehicle and vehicle-borne material identification based on non-repetitive scanning.
A fixed detection system based on non-repetitive scanning was built, and a vehicle feature recognition model based on probability density, a bar position and shape recognition model based on projection dimensionality reduction rasterization features, and a dimensionality reduction circular fitting based on the principal component direction was developed. The coil position recognition model enables accurate identification of vehicles and other vehicles as well as vehicle cargo.
(2) Bar storage area map construction and feature recognition based on binocular vision.
A three-dimensional detection model of the bar storage area based on binocular stereo vision, a multi-pose measurement model of the bar storage area image based on visual and inertial information, and a bar position identification model based on the reduced-dimensional grid mean curvature characteristics are proposed to achieve Accurate identification of bar bundle stacking in the bar warehouse area.
(3) High-line reservoir area map construction and feature recognition based on line scanning.
A dynamic scanning three-dimensional point cloud map construction model and a line roll point cloud segmentation model in the high-line reservoir area were proposed to achieve accurate identification of high-line roll stacking positions in the high-line reservoir area.
2) Precise positioning control technology of unmanned overhead crane integrated with machine vision warehouse location recognition
In view of the characteristics of high positioning accuracy requirements, easy extrusion deformation, easy position changes, and high lifting stability requirements faced by the long steel lifting process, the institute innovatively developed a precise positioning control and operation process based on machine vision storage location recognition. Key technologies such as spreader anti-sway control, multi-section path collaborative transition control, four-axis linkage control and multi-car dynamic anti-collision control have broken through the bottleneck of long product unmanned overhead crane technology and achieved high efficiency and high stability in the long product intelligent warehouse area. Trouble-free operation.
(1) Precise positioning control based on machine vision location recognition.
Establish accurate and safe positioning control technology based on machine vision warehouse location recognition. During the entire positioning control process, we can realize primary and secondary lifting positioning control based on warehouse management targets and scanning corrections and lifting and weighing based on weighing and scanning verification. +Position verification control, primary and secondary positioning control based on warehouse management targets and scanning corrections, and position verification control based on visual scanning.
(2) Spreader anti-sway control technology during operation.
Establish an anti-swing open-loop speed setting control and closed-loop acceleration control model. By setting a multi-section speed curve between the start and end positions, and based on the measured crane swing angle, actual position and speed of the crane, etc., the crane acceleration is adjusted in real-time to maintain The swing of the spreader within a certain limit range.
(3) Four-axis linkage and multi-path collaborative control technology.
Through the development of four-axis linkage control technology for large vehicles, trolleys, hoisting and rotation, as well as multi-path collaborative transition control technology, collaborative optimization operations of hoisting, lowering and avoiding dangerous areas can be achieved, effectively shortening the crane operating time and improving the overall reservoir area. operating efficiency.
3) Multi-agent-based efficient optimization and scheduling technology for logistics in the warehouse area
Considering the collaborative scheduling of multiple logistics equipment under the conditions of complex time and space constraints, the institute has established a multi-agent-based efficient production optimization scheduling technology for warehouse area logistics. At the same time, the scheduling process is based on dynamic production tasks, delivery plans and multiple Dynamically updated scheduling strategies based on the real-time status of the equipment.
(1) Dynamic dispatch model of crane based on multi-car collaborative optimization.
Establish a multi-task and multi-crane crane cluster scheduling model, consider the collaborative work of multiple shipping equipment and cranes, and implement dynamic scheduling of cranes in response to equipment status changes to improve the operating efficiency of cranes in the reservoir area.
(2) Intelligent optimization model for warehouse entry and exit stacking locations.
Comprehensively consider the real-time inventory situation, production plan, shipping plan, lifting capacity and other factors in the warehouse area, and establish an intelligent decision-making model for material inbound and outbound stacking positions to reduce the amount of stacking in the warehouse area, reduce the degree of irregularity in stacking locations, and reduce logistics waiting. time and improve work efficiency in the reservoir area.
(3) Dynamic optimization model of crane path in unmanned reservoir area.
In view of the complex environment and dynamic changes in the layout of the actual reservoir area of the steel plant, an unmanned overhead crane path planning model based on reinforcement learning was developed. Through real-time interaction with the environment and autonomous perception, it can solve the problem of unmanned overhead cranes in complex reservoir area maps and uncertainties. Path planning problems in deterministic environments.
4) Centralized control of long steel intelligent warehouse area based on digital twins and industrial 5G
We have developed key technologies such as digital twins and cluster dispatching simulation systems based on key elements of the warehouse operation, warehouse logistics design and analysis optimization, industrial 5G networks for online control and monitoring, and visual AI-based security monitoring of the warehouse. We have built a new model of intelligent centralized control of long steel storage areas featuring unmanned operation on-site in the storage area, efficient coordinated operation of logistics equipment, and remote centralized visualization and digital monitoring.
(1) Cluster-type integrated intelligent centralized control centre in the reservoir area.
Adopt a centralized and flat control model to realize real-time centralized monitoring of in-warehouse dispatch management, equipment inspection, and receipt and delivery, centrally and quickly handle on-site abnormalities and faults, streamline and integrate on-site operation positions, and improve the efficiency of multi-professional collaborative work.
(2) Digital twin modelling and analysis system based on key elements of reservoir operation.
Use digital twin technology to truly display the scene in the reservoir area, drive the 3D scene in real-time through production data, and present the information that production operations, process technology and management personnel care about to the corresponding personnel in a timely, accurate and clear manner, improving execution, process analysis and The efficiency of implementation of management decisions.
(3) 5G network technology for online control and monitoring.
Combined with the new generation of 5G network technology, it achieves full coverage of 5G signals in high-line reservoir areas. It uses 5G’s low latency and large bandwidth characteristics to realize wireless and highly reliable transmission of surveillance videos, 3D point cloud data, and PLC signals, completing vehicle-mounted and ground-based transmission. Communication to ensure unmanned driving requirements for overhead cranes.
(4) Reservoir area safety monitoring based on visual AI.
Artificial intelligence recognition technology is used through monitoring terminals + intelligent AI terminals to complete intelligent supervision and automatic recording of abnormalities in the reservoir area, as well as early warning prompts and quick review.
2. Application and effects
This achievement has been applied in the bar intelligent library of Zhuhai Yue Yufeng Steel Co., Ltd., the special steel high-line intelligent library of Maanshan Iron and Steel Co., Ltd., and the bar and wire multi-functional logistics library of Hebei Logistics Group, realizing fully unmanned steel in the long steel warehouse area. Automatic lifting, as well as efficient, stable and intelligent coordinated dispatching. In June 2021, the overall online operation was completed, and the successful application performance of the first set of unmanned bar warehouses and the first set of high-line unmanned warehouses in China was achieved, and the 7*24 hours fully unmanned automatic offline transfer and transfer in the warehouse area were successfully completed. For functions such as car shipment and train shipment, the entire reservoir area only requires 1 to 2 people for remote monitoring, which solves the problem of manual operation in heavy and harsh environments in the traditional reservoir area model, significantly improves the operational efficiency of the reservoir area, and effectively releases production bottlenecks by breaking through the reservoir area. production capacity.
Source of the article: “Intelligent unmanned management and control technology for long products (rods and wires) in the reservoir area”