The applications of artificial intelligence in the steel industry are mainly artificial intelligence information systems and robots. Artificial intelligence information systems have brought huge economic benefits to the steel industry and are more widely used. In comparison, due to the complexity of the production process, robot development is difficult, but it has a more targeted effect on all aspects of steel production. (What Is The Use Of Robots In the Steel Industry?>>>)
1) Quality control and optimization
The key to production in the steel industry is product quality control and optimization. Traditional quality control mainly relies on manual sampling and empirical adjustment, but this method often has certain limitations. The emergence of artificial intelligence has provided steel companies with more accurate and efficient quality control methods.
First of all, artificial intelligence technology can realize real-time monitoring and prediction of key parameters in the steel production process through data analysis and algorithm models. For example, by analyzing production data and combining it with machine learning algorithms, the probability of defective products under specific production conditions can be predicted, and timely measures can be taken to make adjustments, thereby effectively reducing the rate of defective products. In addition, artificial intelligence can also help companies discover potential problems hidden behind big data through the analysis and mining of product quality information, and provide solutions and optimization suggestions.
2) Intelligence and automation of the production process
The traditional steel production process often requires a lot of manpower input and manual operations, which is not only inefficient but also prone to safety hazards. The application of artificial intelligence can realize the intelligence and automation of the steel production process and improve production efficiency and quality.
Artificial intelligence technology can realize automatic monitoring, adjustment and control of steel production equipment and processes through machine vision, automatic control and unmanned operation. For example, artificial intelligence technology can be used to realize intelligent monitoring and automatic adjustment of steel-making furnace temperature, pressure and other parameters to improve production efficiency and product quality; machine vision and image recognition technology can be used to automatically control the steel-making material dumping process. Control and quality inspection, reducing the need for manual operations; using automated equipment and robotics technology, automated production and handling of steel products can be realized, improving the intelligence level of the production line.
3) Supply chain management and smart logistics
The application of artificial intelligence can also help steel companies optimize supply chain management and intelligent logistics. The supply chain of the steel industry is relatively complex, involving multiple links such as raw material procurement, production scheduling, inventory management, etc., and often needs to face fluctuations and uncertainties in market demand.
Through artificial intelligence technology, real-time analysis and prediction of supply chain data can be achieved, and accurate demand forecasts and material procurement suggestions can be provided, thereby reducing the company’s operating costs and inventory risks. At the same time, artificial intelligence technology can also realize intelligent optimization and scheduling of logistics processes, improve logistics efficiency and accuracy, and reduce transportation costs and risks.
4) Safety production and environmental protection
The production process of the steel industry is often accompanied by a series of safety hazards and environmental problems. The application of artificial intelligence can help enterprises achieve intelligent management of production safety and environmental protection.
Artificial intelligence technology can predict possible safety hazards and environmental problems through real-time monitoring and analysis of production equipment and working environment data, and timely issue early warnings and take measures to reduce the occurrence of accidents and environmental pollution. For example, through intelligent sensors and data analysis technology, the working status of equipment and parameters such as temperature and pressure can be monitored in real-time, and alarms can be promptly issued when abnormal conditions occur; through intelligent control systems, real-time monitoring of environmental indicators such as exhaust emissions and energy consumption can be achieved Monitor and manage to reduce adverse impacts on the environment.
The following introduces the specific application of AI in the steel industry.
1. Application of artificial intelligence technology in raw material process control
At present, in terms of raw material process control, fuzzy control of the clearance of the finishing crusher in the raw material plant, fuzzy control of uniform sintering of the sintering machine, fuzzy control of the returned ore proportion, etc. have been developed.
The sintering operation expert system of Japan’s Kawasaki Mizushima Steel Plant is composed of a computer with EIXAX expert tools connected to the existing DDC and process machine. The computer and process machine are responsible for different aspects of work respectively. Among them, the computer is responsible for burn-through point management, quality management production cycle, etc., and the process machine is responsible for data collection and equipment abnormal management. The burn-through point management is to adjust the trolley speed based on the long-term predicted value of the air permeability of the inlet raw materials to guide production; the production management function is to ensure that the actual productivity reaches the standard; the quality management function is to judge the corresponding parameters to ensure production quality.
2. Application of artificial intelligence technology in blast furnace process control
The reactions in the blast furnace are complex, and the materials are in a three-state coexistence of gas, liquid, and solid. It is difficult to accurately mathematically model the reactions that occur. Many steel plants have developed blast furnace operation expert systems, systems for forecasting or operating guidance of furnace conditions. , involving many aspects such as blast furnace charging distribution control and granular slag bin operation control. In recent years, neural networks have been widely used in blast furnace temperature prediction, air volume guidance, and reaction model establishment.
For example, the expert system of a steel plant is divided into two parts: furnace heat control and abnormal diagnosis. It uses a combination of process machines and special computers to process and analyze the conditions in the furnace. The process machine analyzes and processes data, models, etc. Special computer knowledge uses three expression methods: rules, frameworks, and LISP, and adopts multi-level grouping structures and forward reasoning. Some knowledge generated based on production experience forms corresponding rules; static knowledge such as air supply temperature and humidity in the furnace is stored in a frame form; LISP functions are used to represent procedural knowledge. The entire knowledge system consists of 1,500 rules and 150 frameworks, which can well guide operations, maintain furnace temperature, improve thermal efficiency, and ensure the quality of molten iron. In addition, the blast furnace operation expert system expresses knowledge with rules and frameworks, and uses forward reasoning to diagnose daily management of furnace conditions, abnormal forecasts, forward rules, etc., and the hit rate of forecasts can be as high as 94%.
The furnace conditions in the blast furnace production process are relatively complex and cannot be represented by a simple “IF A, THEN B” method. Use neural networks to identify furnace conditions and guide operations, or combine them with expert systems to achieve good results. Effective, widely used in blast furnaces.
3. Application of artificial intelligence technology in steelmaking process control
Artificial intelligence technology also has important applications in the steelmaking process. At present, converter blowing control, solvent and coolant loading control systems, and electric arc furnace regulation neural network systems have been developed.
The expert system has three new models: static slag splash prediction, dynamic slag splash prediction, and tapping determination. The static slag splash model evaluates the slag situation and splash possibility based on the initial static blowing data. , select the blowing control mode in each period, and feedback to the operator, who will choose the specific operation; the dynamic slag splash prediction model is based on the sensor data and exhaust model calculation results to evaluate the slag situation and splash possibility. Make optimal choices for the air blowing volume, ore input speed, etc.; the tapping model determines whether to tap based on the data. The application of artificial intelligence technology improves the level of automatic blowing, reduces the occurrence rate of splashing, and improves work efficiency.
4. Application of artificial intelligence technology in continuous casting process control
At this stage, in continuous casting process control, people have invented intelligent robots, liquid level control expert systems, continuous casting breakout predictions, and slab printing text quality identification systems. The intelligent robot in the steel plant consists of three parts: a sensor system, a judgment system and a motion system. It can identify the liquid level of the crystallizer and spread the initial protective slag; identify the amount and edge of the protective slag and protect the parts where the protective slag is insufficient. Add slag and remove the slag edge; identify the formed anti-slag and press it into the molten steel.
The steel plant breakout prediction system is a multi-level neural network composed of a time sequence grid and a spatial control grid. The time sequence network identifies the thermocouple temperature, and the space sequence network identifies the movement pattern of the temperature in the crystallizer. Practice shows that the accuracy of this steel breakage prediction system is almost 100%.
5. Application of artificial intelligence technology in steel rolling process control
Expert systems such as heating furnace combustion control, tapping rhythm control, and electric welded pipe rolling rhythm control have been developed for steel rolling process control. Fuzzy control has been applied to many aspects of the steel rolling process.
The heating furnace combustion control expert system is composed of a mathematical model and an expert system. The mathematical model is used to calculate the relationship between steel temperature and time, and the expert system is used to select the objective function and related constraints to determine the optimal temperature. Improve production efficiency and product quality.
6. Application of artificial intelligence technology in product design
Product design must consider many factors. There are great difficulties in optimizing the product design process using artificial intelligence technology. Research in this area has been ongoing in recent years and has achieved certain results.
According to the user’s ordering specifications, an expert system for the material design of large-diameter steel pipes has been developed to select the most suitable production plan. The system converts product parameters into intermediate product parameters through 300 rules and corresponding data operations, stipulates constraint conditions, tentatively determines steel types and manufacturing methods, then accurately matches basic data, compares existing data, and predicts this design. Finally, the characteristic values of each influencing factor are adjusted until the specifications are met. This system is simple to operate and improves production efficiency.
7. Application of artificial intelligence technology in production planning and scheduling
Formulating production plans and scheduling production is also an important part of the steel industry. In recent years, there have been many studies on how to apply artificial intelligence technology to the formulation and scheduling of production plans, such as raw material plant operation plans, steelmaking batch combination plans, and coal mixing plans. , hot rolling billet distribution, cold rolling rolling sequence preparation, cold rolling tin plating scheduling plan, etc.
The steel plant’s raw material plant optimal operation planning system is a large-scale system that applies artificial intelligence technology to support operations, including raw material loading and unloading, raw material plant distribution, coal transportation, coal mining, ore homogenization, coke oven operations, and dry coke quenching operations. , Sintering operation expert system.
The specific operation method takes coal mixing as an example. Through the analysis of various coals, artificial intelligence and mathematical model analysis and calculation, an optimal coal mixing plan is formulated. The specific method is: CRT designs relevant specifications, taking into account the supply and demand situation, determines the order and proportion of achieving the goals, generates a plan, then analyses and calculates the plan, evaluates and verifies, and finally revises the plan and repeat calculations until a satisfactory plan is generated.
Source of article: “Application of Artificial Intelligence in the Steel Industry”