Control method of oxygen lance nitrogen sealing system of converter
Through intelligent monitoring and automatic adjustment of the oxygen and nitrogen supply, combined with flame state analysis, the problems of waste of resources and insufficient pressure in the converter oxygen gun nitrogen sealing system are solved, and a more efficient and safer smelting process is achieved.
Patent Information
- Application Number
- CN202410836927.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-06-26
- Publication Date
- 2024-11-12
AI Technical Summary
The existing converter oxygen gun nitrogen sealing system has a continuous supply of gas sources when there is no large injection, resulting in waste of resources, and the nitrogen sealing pressure during large injection is insufficient, which affects the stability of the smelting process and equipment safety.
By setting target parameters, the state in the furnace is monitored in real time, the oxygen supply and nitrogen enclosure are automatically adjusted using intelligent algorithms and PLC control systems, and combined with flame image sample feature vectors and cluster recognition learning algorithms, intelligent flame control and oxygen-to-nitrogen ratio are achieved. Adjustment, supports remote monitoring and operation.
It effectively avoids energy waste, improves smelting efficiency and product quality, reduces accident risks, improves the safety of equipment and staff, optimizes flame control and oxygen-nitrogen ratio, and improves system stability and control accuracy.
Smart Images

Figure CN118932134A_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the technical field of iron and steel metallurgy, and in particular to a control method for an oxygen lance nitrogen sealing system of a converter. Background Art
[0002] During converter steelmaking operations at a certain steel plant, a large amount of slag suddenly splashed, causing smoke and flames to escape through the gap under the muzzle of the oxygen lance on the converter platform located at a height of 22.5 meters. This caused damage to some equipment on the platform, deformation of the rails, and affected the surrounding environment. As an important link in steelmaking production, the stability and safety of the converter process are of vital importance, and the normal discharge of slag is an indispensable part of the steelmaking process.
[0003] In the prior art, a nitrogen-sealed enclosure is currently used to control flame and flue gas with a constant nitrogen gas of 0.5 MPa (manual nitrogen control by DCS). This control method has disadvantages: when there is no large-scale blowdown, if the gas source is continuously supplied but not utilized, it will lead to a waste of gas resources and increase production costs. In addition, insufficient nitrogen sealing pressure during large-scale blowdown will lead to an inability to control the atmosphere in the furnace, affecting the stability and quality of the smelting process. At the same time, insufficient nitrogen sealing pressure will lead to unstable flame control, increase the risk of equipment damage, and even cause equipment failure or accidents, posing a threat to equipment safety.
[0004] Therefore, those skilled in the art provide a control method for an oxygen lance nitrogen sealing system of a converter to solve the problems raised in the above background technology. Summary of the invention
[0005] In view of the deficiencies in the prior art, the present invention provides a control method for the oxygen lance nitrogen sealing system of a converter, which solves the problem that when there is no large-scale blowing, if the gas source is continuously supplied but not utilized, it will lead to waste of gas resources and increase production costs; and insufficient nitrogen sealing pressure during large-scale blowing will lead to the inability to control the atmosphere in the furnace, affecting the stability and quality of the smelting process; at the same time, insufficient nitrogen sealing pressure will lead to unstable flame control, increase the risk of equipment damage, and even cause equipment failure or accidents, posing a threat to equipment safety.
[0006] To achieve the above objectives, the present invention is implemented through the following technical solutions: A control method for an oxygen lance nitrogen sealing system of a converter comprises the following steps:
[0007] Step 1: Set target parameters: Set target parameters according to smelting requirements and process flow;
[0008] Step 2: Real-time monitoring of system status: real-time monitoring of furnace temperature, pressure, oxygen and nitrogen flow through sensors;
[0009] Step 3: Intelligently adjust oxygen supply and nitrogen sealing: According to real-time monitoring data, the oxygen supply and nitrogen sealing amount are automatically adjusted through intelligent algorithms;
[0010] Step 4: PLC control system: adjust the oxygen-nitrogen ratio in real time according to the situation in the furnace;
[0011] Step 5: Intelligent feedback adjustment: Automatically make feedback adjustments based on real-time monitoring data and preset algorithms;
[0012] Step 6. Remote monitoring and operation: Support remote monitoring and operation functions, which facilitate operators to monitor the operating status of equipment in real time;
[0013] Step 7: Selecting flame image sample feature vectors: Obtaining and converting the image into image format through an image acquisition card, and completing image enhancement, restoration, and denoising before extracting the image sample feature vectors;
[0014] Step 8: Flame status monitoring and analysis: Monitor the flame status, analyze the flame characteristics, and identify the flame status through the flame clustering recognition learning algorithm;
[0015] Step 9: Intelligent flame control: According to the flame status monitoring results, the control system intelligently adjusts the oxygen and nitrogen supply;
[0016] Step 10: Data integration and comprehensive analysis: Integrate the flame state monitoring data with the traditional control parameter data and conduct a comprehensive analysis.
[0017] Step 1: Set parameters to achieve the best smelting effect; Step 2: Provide the data monitored by the sensor to the control system for subsequent intelligent adjustment and analysis; Step 3: Intelligently adjust the oxygen supply and nitrogen sealing to ensure the balance and stability of the atmosphere in the furnace and improve the smelting efficiency and product quality; Step 4: PLC control system, through the feedback mechanism of the control system and the content of oxygen and nitrogen in the furnace, timely adjust the oxygen and nitrogen ratio to meet the requirements of smelting; Step 5: Intelligent feedback adjustment, by analyzing the monitoring data, the control system can timely identify abnormal conditions and make corresponding adjustments to ensure the stable operation of the converter; Step 6: Remote monitoring and operation can improve work efficiency and safety; Step 7: Select the feature vector of the flame image sample to improve the image quality and accuracy; Step 8: Through the flame clustering recognition learning algorithm, it can help judge the combustion situation and process status in the furnace; Step 9: By automatically adjusting the gas supply, the combustion process can be optimized, and the smelting efficiency and product quality can be improved; Step 10: Through the comprehensive analysis of various data, potential associations and laws can be found to provide a basis for optimization and improvement.
[0018] Preferably, in step one, the target parameters include oxygen supply, nitrogen sealing amount and oxygen-nitrogen ratio, the oxygen supply range is 5%-30%, the nitrogen sealing amount range is 2%-15%, and the oxygen-nitrogen ratio range is 1:1-10:1.
[0019] Preferably, in step 2, the temperature range of the furnace is 1500°C-2000°C, the pressure range of the furnace is 0.1MPa-0.5MPa, and the oxygen flow rate range is 100Nm 3 / h-1000Nm 3 / h, nitrogen flow range is 50Nm 3 / h-500Nm 3 / h.
[0020] In step 2, the temperature, pressure, oxygen and nitrogen flow parameters in the furnace are monitored in real time. The optimal range can be provided to the control system for subsequent intelligent adjustment and analysis. The monitoring of parameters can help operators understand the status of the converter in a timely manner, detect abnormal conditions and take corresponding measures.
[0021] Preferably, in step 3, the intelligent algorithm formula is: [new oxygen supply = old oxygen supply + K 氧气 ×Δ oxygen], [new nitrogen sealing volume = old nitrogen sealing volume + K 氮气 ×Δ nitrogen],
[0022] Where, old oxygen supply is oxygen supply, K 氧气 is the proportional coefficient used to adjust the oxygen supply, ΔO2 is the oxygen adjustment amount,
[0023] Among them, the old nitrogen sealing amount is the nitrogen sealing amount, K 氮气 It is the proportional coefficient used to adjust the nitrogen sealing amount, and Δnitrogen is the nitrogen adjustment amount.
[0024] In step three, an intelligent algorithm is used to adjust the oxygen supply and nitrogen sealing amount. In the intelligent algorithm formula, by setting the proportional coefficient and the adjustment amount, the oxygen and nitrogen supply can be automatically adjusted according to the real-time monitoring data, which can ensure the balance and stability of the furnace atmosphere and improve the smelting efficiency and product quality.
[0025] Preferably, in step 4, the specific implementation steps of the PLC control system are:
[0026] Programming: Develop PLC program to adjust the oxygen-nitrogen ratio in real time according to the real-time monitoring data of furnace temperature, pressure, oxygen-nitrogen flow rate and nitrogen flow rate;
[0027] Data collection: Set up the data collection module to obtain sensor monitoring data in real time;
[0028] Control logic: adjust the oxygen-nitrogen ratio according to monitoring data and preset algorithms;
[0029] Communication module: configure the communication module to realize data exchange and communication between the PLC system and the monitoring system and data integration system;
[0030] Fault handling: Set up fault detection and handling mechanisms so that the system can respond promptly when abnormal situations occur.
[0031] The purpose of programming is to combine monitoring data with control logic to achieve automated control operations; the data acquisition module can be connected to sensors or signal input devices to transmit real-time monitoring data to the PLC system; the control logic can be designed according to specific control requirements and algorithms to ensure that the oxygen-nitrogen ratio is within the set range and dynamically adjusted according to real-time monitoring data; the communication module can transmit the data acquired by the PLC system to the monitoring system or data integration system through the network or other communication methods, and receive instructions or control signals from these systems; fault handling may include detection and processing of abnormal conditions such as sensor failure and communication interruption to ensure the stability and reliability of the system; the PLC control system helps to improve the smelting efficiency and product quality of the converter and reduce the need for manual intervention.
[0032] Preferably, in step 5, the preset algorithm formula is:
[0033]
[0034] Among them, K P .K I .K D are the proportional, integral and differential coefficients, used to adjust the proportional, integral and differential effects of the feedback control. The error is the deviation between the monitored value and the target value. The ∑ error is the cumulative sum of the errors, used for integral control. is the rate of change of error and is used for derivative control.
[0035] Preferably, in step seven, the flame image sample feature vector selection includes:
[0036] Average light intensity (Ie), if p(i) is the probability of the i-th grayscale, where 0≤i≤L, L is the maximum grayscale value of the image, and m is the actual maximum grayscale level of the image, then
[0037] Light intensity variance (σ), if M, N are the height and width of the image, both in pixels, f(i, j) is the gray value at (i, j), and f is the average light intensity, then
[0038] Light intensity information entropy (H), if the grayscale amplitude of the digital flame image has x1, x2, x3, ... xm types with probabilities of P1, P2, P3, ... Pm respectively, then
[0039] Flame radiation intensity (E), for random image signal field, if f(i, j) is used to represent it, F(u, v) is the discrete Fourier transform of flame image The flame radiation intensity
[0040] Flame occupancy (A), if the area occupied by the pixels of the M×N digital flame image is s, then the area of the entire image is: Stotal = M×N×s. After segmenting the image and removing the background, assuming that the image area of the flame is Sflame, and the total number of pixels is ψ, then: Sflame = ψ×s, and the flame abundance is defined as: Reflects the degree of flame fullness.
[0041] In step seven, the selection of flame image sample feature vectors includes the following indicators: average light intensity: calculate the average value of image grayscale to describe the overall brightness of the flame; light intensity variance: calculate the variance of image grayscale to indicate the degree of change in flame brightness; light intensity information entropy: calculate information entropy based on the probability distribution of image grayscale to describe the complexity of the grayscale distribution of the flame; flame radiation intensity: calculate the radiation intensity of the flame by performing discrete Fourier transform on the image; flame occupancy: calculate the flame occupancy based on the number of pixels and total area of the flame image to describe the proportion of the flame in the image; the above-mentioned feature vectors can be obtained by processing and analyzing the flame image, and are used to describe the brightness, grayscale distribution, radiation intensity and occupancy characteristics of the flame, which is convenient for further analysis and judgment of the state and characteristics of the flame, and using the flame image sample feature vector to describe the characteristics of the flame is helpful to realize intelligent regulation and control of the flame, and improve the stability and efficiency of the smelting process.
[0042] Preferably, in step eight, the flame clustering recognition learning algorithm formula is:
[0043]
[0044] ||U k ||=[U k1 +U k2 +U k3 +∧+U kn ] 1 / 2 ,
[0045] The evaluation function is the shortest Euclidean distance between vectors.
[0046] d j =||Uk -W j ||,d g =min)d j )(j=1,2,∧,M),
[0047] In the process of self-organizing learning, the connection weight of the network needs to be based on the vector U K Adjustment is made, that is, the connection weight is adjusted using the following equation:
[0048] The vector W of the connection j (t+1) is normalized:
[0049]
[0050] Update learning efficiency η(t) and neighborhood N g (t):
[0051]
[0052] In the formula, η is the initial value of learning efficiency; t is the number of learning times; T is the total number of learning times; INT[x] means rounding x.
[0053] Preferably, in step nine, the intelligent flame control step is:
[0054] Flame status monitoring: obtain real-time flame status data based on the flame monitoring system;
[0055] Data analysis: Use flame clustering recognition learning algorithm to analyze flame state data;
[0056] Determine the adjustment strategy: Determine the adjustment strategy based on the analysis results of the flame state;
[0057] Intelligent regulation: intelligently adjust the oxygen and nitrogen supply through the PLC control system;
[0058] Real-time monitoring and adjustment: Continuously monitor flame status and system parameters in real time and adjust oxygen and nitrogen supply;
[0059] Feedback mechanism: A feedback mechanism is established to feed back the effect of the adjusted oxygen and nitrogen supply to the system.
[0060] In step nine, the specific steps of intelligently adjusting flame control are as follows: obtaining real-time flame images, temperature, and light intensity parameters through the flame monitoring system, and analyzing the flame state data using the flame clustering recognition learning algorithm to understand the state, change trend, and characteristics of the flame. At the same time, according to the parameters that need to be adjusted and the state characteristics of the flame, the PLC control system automatically adjusts the control parameters according to the preset adjustment strategy and control algorithm; real-time monitoring and adjustment are compared with the target value, and adjustments are made according to feedback information to ensure the stability and accuracy of flame control; by monitoring the changes in flame state and control parameters, the adjustment effect is evaluated, and corrections and optimizations are made in time to achieve closed-loop control; intelligently adjusting flame control can determine the appropriate adjustment strategy based on real-time flame state data using data analysis and clustering algorithms, and perform intelligent adjustment through the PLC control system. Continuous real-time monitoring and adjustment and the establishment of a feedback mechanism can ensure the stability and accuracy of flame control and improve the efficiency and quality of the smelting process.
[0061] Preferably, in step 10, the specific implementation method of the data integration and comprehensive analysis is:
[0062] Data integration: Integrate the flame state data obtained in the flame state monitoring and analysis steps with the data obtained in the traditional control parameter monitoring into the data platform;
[0063] Data cleaning and preprocessing: Clean and preprocess the integrated data;
[0064] Data analysis and modeling: Use data mining and machine learning techniques to analyze and model the integrated data and explore the relationships and patterns between data;
[0065] Comprehensive analysis and decision support: Based on the results of data analysis, conduct comprehensive analysis to find key information and trends in the data.
[0066] The purpose of data integration is to integrate scattered data into the platform, provide a complete data foundation for subsequent analysis and decision-making, and help the system fully understand the production process; cleaning and preprocessing data help eliminate noise and errors in the data, make the data more credible, and provide a high-quality data foundation for subsequent analysis and modeling; through data analysis and modeling, we can deeply understand the characteristics of the data, discover the valuable information behind the data, and provide decision support and predictive capabilities for optimizing the control system; the purpose of comprehensive analysis is to transform the results of data analysis into practical actions, provide decision support for managers, help them formulate effective control strategies, optimize the production process, and improve production efficiency and product quality.
[0067] The present invention provides a control method for an oxygen lance nitrogen sealing system of a converter, which has the following beneficial effects:
[0068] 1. The present invention sets target parameters and monitors the system status in real time, so that the system can intelligently adjust the oxygen supply and nitrogen sealing amount according to actual conditions, optimize the oxygen-nitrogen ratio, improve smelting efficiency and production quality, and adjust the oxygen-nitrogen ratio in real time through the PLC control system. Intelligent feedback adjustment and remote monitoring can timely discover problems and make adjustments, reduce accident risks, and improve the safety of equipment and personnel.
[0069] 2. The present invention intelligently adjusts the oxygen and nitrogen supply and flame control, performs intelligent adjustments based on real-time monitoring data, effectively controls the supply of oxygen and nitrogen, avoids energy waste, and reduces production costs. At the same time, through flame state monitoring and analysis as well as data integration and comprehensive analysis, the system can have a more thorough understanding of the relationship between flame state and control parameters, provide a comprehensive information basis for the control system, optimize control decisions, and improve system stability and control accuracy.
[0070] 3. The present invention adopts intelligent algorithms and learning algorithms for adjustment and monitoring, combined with remote monitoring and operation functions, so that the system has a high level of intelligence, can automatically control and adjust, reduce human intervention, and improve work efficiency. Through flame state monitoring and analysis, the system can more accurately control the flame state, which helps to improve product quality, optimize the production process, reduce the defective rate, and enhance the competitiveness of the enterprise. BRIEF DESCRIPTION OF THE DRAWINGS
[0071] Figure 1 The figure is a flow chart of the method of the present invention. DETAILED DESCRIPTION
[0072] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.
[0073] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.
[0074] The present invention is described in detail below in conjunction with the accompanying drawings:
[0075] Embodiment 1: Traditional control method
[0076] Please refer to the attached Figure 1 The embodiment of the present invention provides a control method for an oxygen lance nitrogen sealing system of a converter, comprising the following steps:
[0077] 1. Set target parameters:
[0078] Oxygen supply range: 10%-25%
[0079] Nitrogen sealing range: 3%-12%
[0080] Oxygen-nitrogen ratio range: 2:1-8:1
[0081] 2. Implementation steps:
[0082] a. Target parameter setting: The operator manually sets the target parameters according to the smelting requirements and process flow;
[0083] b. System status monitoring: Sensors monitor the following parameters in real time:
[0084] Furnace temperature range: 1500℃-2000℃
[0085] Furnace pressure range: 0.1MPa-0.5MPa
[0086] Oxygen flow range: 100Nm3 / h-1000Nm3 / h
[0087] Nitrogen flow range: 50Nm3 / h-500Nm3 / h
[0088] c. Manual adjustment: The operator manually adjusts the oxygen supply and nitrogen sealing amount according to the monitoring data to maintain the set target parameters;
[0089] d.PLC control system: Based on manual intervention, the PLC system adjusts the oxygen and nitrogen ratio in real time;
[0090] e. Data integration and analysis: Integrate monitoring data into the data platform and manually analyze the data for decision support.
[0091] Embodiment 2: Intelligent control method
[0092] Please refer to the attached Figure 1 The embodiment of the present invention provides a control method for an oxygen lance nitrogen sealing system of a converter, comprising the following steps:
[0093] 1. Set target parameters:
[0094] Oxygen supply range: 5%-30%
[0095] Nitrogen sealing range: 2%-15%
[0096] Oxygen-nitrogen ratio range: 1:1-10:1
[0097] 2. Implementation steps:
[0098] a. Target parameter setting: Set target parameters according to smelting requirements and process flow;
[0099] b. System status monitoring: Sensors monitor the following parameters in real time:
[0100] Furnace temperature range: 1500℃-2000℃
[0101] Furnace pressure range: 0.1MPa-0.5MPa
[0102] Oxygen flow range: 100Nm3 / h-1000Nm3 / h
[0103] Nitrogen flow range: 50Nm3 / h-500Nm3 / h
[0104] c. Intelligent regulation: Based on real-time monitoring data, the intelligent algorithm automatically adjusts the oxygen supply and nitrogen sealing amount;
[0105] d.PLC control system: According to the flame status monitoring results, the PLC system intelligently adjusts the oxygen-nitrogen ratio;
[0106] e. Flame status monitoring and analysis: Intelligently adjust flame control through flame image sample feature vector selection, flame status monitoring and analysis;
[0107] f. Data integration and comprehensive analysis: Integrate flame status monitoring data and traditional control parameter data into the data platform, and use machine learning technology for comprehensive analysis and decision support.
[0108] Example 3: Control of converter oxygen lance nitrogen sealing system based on model predictive control
[0109] Please refer to the attached Figure 1 The embodiment of the present invention provides a control method for an oxygen lance nitrogen sealing system of a converter, comprising the following steps:
[0110] 1. Set target parameters:
[0111] Oxygen supply range: 8%-28%
[0112] Nitrogen sealing range: 4%-14%
[0113] Oxygen-nitrogen ratio range: 1.5:1-9:1
[0114] 2. Implementation steps:
[0115] a. Target parameter setting: Set target parameters according to process requirements and optimize them using model predictive control methods;
[0116] b. System status monitoring: Sensors monitor key parameters in real time, including furnace temperature, pressure, and oxygen and nitrogen flow rates;
[0117] c. Model predictive control: Predict system behavior based on mathematical models and optimize oxygen and nitrogen supply to achieve more precise control;
[0118] d. Real-time adjustment: adjust the oxygen-nitrogen ratio in real time according to the prediction results to ensure stable operation of the system;
[0119] e. Data integration and analysis: Integrate real-time monitoring data and model prediction results, and conduct comprehensive analysis to improve control strategies;
[0120] Example 4: Control of converter oxygen lance nitrogen sealing system based on reinforcement learning
[0121] Please refer to the attached Figure 1 The embodiment of the present invention provides a control method for an oxygen lance nitrogen sealing system of a converter, comprising the following steps:
[0122] 1. Set target parameters:
[0123] Oxygen supply range: 6%-32%
[0124] Nitrogen sealing range: 3%-16%
[0125] Oxygen-nitrogen ratio range: 1.2:1-9.5:1
[0126] 2. Implementation steps:
[0127] a. Target parameter setting: Set target parameters and introduce reinforcement learning algorithm to optimize oxygen and nitrogen supply strategy;
[0128] b. System status monitoring: Sensors monitor key parameters in real time, including furnace temperature, pressure, and oxygen and nitrogen flow rates;
[0129] c. Reinforcement learning control: The system learns the optimal control strategy based on the reward mechanism and achieves adaptive adjustment;
[0130] d. Intelligent adjustment: The system intelligently adjusts the oxygen and nitrogen supply according to the learning results to improve control accuracy;
[0131] e. Data integration and analysis: Integrate the data in the reinforcement learning process, analyze the effects and optimize the control strategy;
[0132] Example 5: Control of converter oxygen lance nitrogen sealing system based on fuzzy logic control
[0133] Please see attached Figure 1 The embodiment of the present invention provides a control method for an oxygen lance nitrogen sealing system of a converter, comprising the following steps:
[0134] 1. Set target parameters:
[0135] Oxygen supply range: 7%-27%
[0136] Nitrogen sealing range: 3.5%-13.5%
[0137] Oxygen-nitrogen ratio range: 1.3:1-9:1
[0138] 2. Implementation steps:
[0139] a. Target parameter setting: Set target parameters and use fuzzy logic control methods to deal with fuzziness and uncertainty;
[0140] b. System status monitoring: Sensors monitor key parameters in real time, including furnace temperature, pressure, and oxygen and nitrogen flow rates;
[0141] c. Fuzzy logic control: establish a fuzzy rule base and adjust the oxygen and nitrogen supply according to fuzzy reasoning;
[0142] d. Real-time adjustment: adjust the oxygen-nitrogen ratio in real time according to fuzzy rules to adapt to system changes;
[0143] e. Data integration and analysis: Integrate the data in the fuzzy logic control process to conduct performance evaluation and control strategy optimization.
[0144] Summary of the embodiments:
[0145] Function of Embodiment 1: The traditional control method relies on manual intervention and experience, and maintains the target parameters by manually adjusting the oxygen and nitrogen supply, which is suitable for simple control requirements and operation processes.
[0146] Function of the second embodiment: The intelligent control method introduces intelligent algorithms and flame state monitoring technology to achieve automatic regulation and intelligent smelting process control, which is suitable for complex control requirements and real-time monitoring.
[0147] Function of Embodiment 3: The model predictive control method uses a mathematical model to predict system behavior, optimize the oxygen and nitrogen supply, and improve control accuracy and stability. It is suitable for high-demand processes that require precise control.
[0148] Function of Example 4: The reinforcement learning control method learns the optimal control strategy through a reward mechanism to achieve adaptive adjustment and optimization, and is suitable for control scenarios that require dynamic adjustment and strong adaptability.
[0149] Function of the fifth embodiment: The fuzzy logic control method handles fuzziness and uncertainty, and realizes the regulation of oxygen and nitrogen supply through the fuzzy rule base, which is suitable for complex systems and frequently changing control environments.
[0150] Experimental design:
[0151] Comparative experiment 1:
[0152] Experimental steps:
[0153] 1. Determine the initial nitrogen and oxygen supply before the experiment begins;
[0154] 2. Start the deep learning control method and record the output at each time point;
[0155] 3. Start the fuzzy logic control method and record the output at each time point;
[0156] 4. Monitor the actual nitrogen and oxygen supply in real time and record the data;
[0157] 5. Analyze the control effects of deep learning methods and fuzzy logic control methods;
[0158] 6. Evaluate the control accuracy and stability of deep learning-based control methods and fuzzy logic control methods based on experimental data;
[0159] 7. Compare the actual nitrogen and oxygen supply with the output of the deep learning control method and the fuzzy logic control method, and analyze the deviation;
[0160] 8. End the experiment and summarize the results, and make suggestions for improvement.
[0161] Experimental data table:
[0162] Time (min) Deep learning method output Fuzzy logic control method output Actual nitrogen and oxygen supply 0 0.5 0.3 0.4 5 0.6 0.4 0.55 10 0.7 0.45 0.65 15 0.8 0.5 0.75
[0163] Parameter settings:
[0164] Deep learning method: Use a neural network model, set the learning rate to 0.001, and the training cycle to 100 times;
[0165] Fuzzy logic control method: setting the fuzzy rule base, including fuzzy sets and membership functions;
[0166] Actual nitrogen oxygen supply range: 0~1.
[0167] Comparative experiment 2:
[0168] Experimental steps:
[0169] 1. Set the model parameters and prediction period of the model predictive control method;
[0170] 2. Start the model predictive control method and record the output at each time point;
[0171] 3. Set the reward mechanism and learning rate of the reinforcement learning control method;
[0172] 4. Start the reinforcement learning control method and record the output at each time point;
[0173] 5. Monitor the actual nitrogen and oxygen supply in real time and record the data;
[0174] 6. Compare the effects of model predictive control method and reinforcement learning control method;
[0175] 7. Analyze the control performance and stability of model predictive control method and reinforcement learning control method;
[0176] 8. End the experiment and summarize the results, and make suggestions for improvement.
[0177] Experimental data table:
[0178]
[0179]
[0180] Parameter settings:
[0181] Model predictive control method: Use the ARIMA model for prediction and set the prediction period to 5 minutes;
[0182] Reinforcement learning control method: Use Q-learning algorithm, set reward function and learning rate;
[0183] Actual nitrogen oxygen supply range: 0~1.
[0184] Comparative experiment three:
[0185] Experimental steps:
[0186] 1. Determine the initial nitrogen and oxygen supply before the experiment begins;
[0187] 2. Start the intelligent control method and record the output at each time point;
[0188] 3. Start the traditional control method and record the output at each time point;
[0189] 4. Monitor the actual nitrogen and oxygen supply in real time and record the data;
[0190] 5. Compare the effects of intelligent control methods and traditional control methods;
[0191] 6. Analyze the control performance and stability of intelligent control methods and traditional control methods;
[0192] 7. Check the difference between the actual nitrogen and oxygen supply and the output of the intelligent control method and the traditional control method;
[0193] 8. End the experiment and summarize the results, and make suggestions for improvement
[0194] Experimental data table:
[0195]
[0196]
[0197] Parameter settings:
[0198] Intelligent control method: combining neural network and fuzzy logic control to achieve adaptive regulation;
[0199] Traditional control method: manually adjust the nitrogen and oxygen supply;
[0200] Actual nitrogen oxygen supply range: 0~1.
[0201] Comparative experiment summary:
[0202] Deep learning-based control method VS fuzzy logic control method:
[0203] Results analysis: The deep learning method showed higher accuracy and stability in controlling the nitrogen and oxygen supply, and the deviation from the actual value was smaller; the fuzzy logic control method could achieve control, but with slightly lower accuracy and larger deviation than the deep learning method;
[0204] Conclusion: In this experiment, the control method based on deep learning performed better than the fuzzy logic control method and had better control effect and stability.
[0205] Model predictive control method VS reinforcement learning control method:
[0206] Result analysis:
[0207] Both the model predictive control method and the reinforcement learning control method can effectively control the nitrogen and oxygen supply; the output of the model predictive control method is closer to the actual value;
[0208] Conclusion: In this experiment, the model predictive control method can show control accuracy and stability compared with the reinforcement learning control method.
[0209] Intelligent control method VS traditional control method:
[0210] Result analysis:
[0211] Intelligent control methods are more effective in controlling nitrogen and oxygen supply, with smaller deviations from actual values and more stable control effects. Although traditional control methods can achieve control, they are limited by manual intervention and experience, resulting in reduced control accuracy.
[0212] Conclusion: In this experiment, the intelligent control method is significantly superior to the traditional control method and has higher control accuracy and stability.
[0213] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.
Claims
1. A control method for an oxygen lance nitrogen sealing system of a converter, characterized in that: The following steps are involved: Step 1: Set target parameters: Set target parameters according to smelting requirements and process flow; Step 2: Real-time monitoring of system status: real-time monitoring of furnace temperature, pressure, oxygen and nitrogen flow through sensors; Step 3: Intelligently adjust oxygen supply and nitrogen sealing: According to real-time monitoring data, the oxygen supply and nitrogen sealing amount are automatically adjusted through intelligent algorithms; Step 4: PLC control system: adjust the oxygen-nitrogen ratio in real time according to the situation in the furnace; Step 5: Intelligent feedback adjustment: Automatically make feedback adjustments based on real-time monitoring data and preset algorithms; Step 6. Remote monitoring and operation: Support remote monitoring and operation functions, which facilitate operators to monitor the operating status of equipment in real time; Step 7: Selecting flame image sample feature vectors: Obtaining and converting the image into image format through an image acquisition card, and completing image enhancement, restoration, and denoising before extracting the image sample feature vectors; Step 8: Flame status monitoring and analysis: Monitor the flame status, analyze the flame characteristics, and identify the flame status through the flame clustering recognition learning algorithm; Step 9: Intelligent flame control: According to the flame status monitoring results, the control system intelligently adjusts the oxygen and nitrogen supply; Step 10: Data integration and comprehensive analysis: Integrate the flame state monitoring data with the traditional control parameter data and conduct a comprehensive analysis.
2. The control method of the oxygen lance nitrogen sealing system of a converter according to claim 1, characterized in that: In the step 1, the target parameters include oxygen supply, nitrogen sealing amount and oxygen-nitrogen ratio. The oxygen supply range is 5%-30%, the nitrogen sealing amount range is 2%-15%, and the oxygen-nitrogen ratio range is 1:1-10:
1.
3. The control method of the oxygen lance nitrogen sealing system of a converter according to claim 1, characterized in that: In step 2, the temperature range in the furnace is 1500°C-2000°C, the pressure range in the furnace is 0.1MPa-0.5MPa, and the flow rate range of oxygen is 100Nm 3 / h-1000Nm 3 / h, the flow range of nitrogen is 50Nm 3 / h-500Nm 3 / h.
4. The control method of the oxygen lance nitrogen sealing system of a converter according to claim 1, characterized in that: In step 3, the intelligent algorithm formula is: [new oxygen supply = old oxygen supply + K 氧气 ×Δ oxygen], [new nitrogen sealing volume = old nitrogen sealing volume + K 氮气 ×Δ nitrogen], Where, old oxygen supply is oxygen supply, K 氧气 is the proportional coefficient used to adjust the oxygen supply, ΔO2 is the oxygen adjustment amount, Among them, the old nitrogen sealing amount is the nitrogen sealing amount, K 氮气 It is the proportional coefficient used to adjust the nitrogen sealing amount, and Δnitrogen is the nitrogen adjustment amount.
5. The control method of the oxygen lance nitrogen sealing system of a converter according to claim 1, characterized in that: In step 4, the specific implementation steps of the PLC control system are: Programming: Develop PLC program to adjust the oxygen-nitrogen ratio in real time according to the real-time monitoring data of furnace temperature, pressure, oxygen-nitrogen flow rate and nitrogen flow rate; Data collection: Set up the data collection module to obtain sensor monitoring data in real time; Control logic: adjust the oxygen-nitrogen ratio according to monitoring data and preset algorithms; Communication module: configure the communication module to realize data exchange and communication between the PLC system and the monitoring system and data integration system; Fault handling: Set up fault detection and handling mechanisms so that the system can respond promptly when abnormal situations occur.
6. The control method of the oxygen lance nitrogen sealing system of a converter according to claim 1, characterized in that: In step 5, the preset algorithm formula is: Among them, K P .K I .K D are the proportional, integral and differential coefficients, used to adjust the proportional, integral and differential effects of the feedback control. The error is the deviation between the monitored value and the target value. The ∑ error is the cumulative sum of the errors, used for integral control. is the rate of change of error and is used for derivative control.
7. The control method of the oxygen lance nitrogen sealing system of a converter according to claim 1, characterized in that: In the step 7, the flame image sample feature vector selection includes: Average light intensity (Ie), if p(i) is the probability of the i-th grayscale, where 0≤i≤L, L is the maximum grayscale value of the image, and m is the actual maximum grayscale level of the image, then Light intensity variance (σ), if M, N are the height and width of the image, both in pixels, f(i, j) is the gray value at (i, j), and f is the average light intensity, then Light intensity information entropy (H), if the grayscale amplitude of the digital flame image has x1, x2, x3, ... xm types with probabilities of P1, P2, P3, ... Pm respectively, then Flame radiation intensity (E), for random image signal field, if f(i, j) is used to represent it, F(u, v) is the discrete Fourier transform of flame image The flame radiation intensity Flame occupancy (A), if the area occupied by the pixels of the M×N digital flame image is s, then the area of the entire image is: Stotal = M×N×s. After segmenting the image and removing the background, assuming that the image area of the flame is Sflame, and the total number of pixels is ψ, then: Sflame = ψ×s, and the flame abundance is defined as: Reflects the degree of flame fullness.
8. The control method of the oxygen lance nitrogen sealing system of a converter according to claim 1, characterized in that: In step eight, the flame clustering recognition learning algorithm formula is: ∥U k ∥=[U k1 +U k2 +U k3 +∧+U kn ] 1 / 2 , then the evaluation function is the shortest Euclidean distance between vectors, d j =||U k -W j ||,d g =min(d j )(j=1,2,∧,M), the network is in the process of self-organizing learning, and its connection weight needs to be based on the vector U K Adjustment is made, that is, the connection weight is adjusted using the following equation: The vector W of the connection j (t+1) is normalized: Update learning efficiency η(t) and neighborhood N g (t): Among them, η is the initial value of learning efficiency; t is the number of learning times; T is the total number of learning times; INT[x] means rounding x.
9. The control method of the oxygen lance nitrogen sealing system of a converter according to claim 1, characterized in that: In step nine, the intelligent flame adjustment control step is: Flame status monitoring: obtain real-time flame status data based on the flame monitoring system; Data analysis: Use flame clustering recognition learning algorithm to analyze flame state data; Determine the adjustment strategy: Determine the adjustment strategy based on the analysis results of the flame state; Intelligent regulation: intelligently adjust the oxygen and nitrogen supply through the PLC control system; Real-time monitoring and adjustment: Continuously monitor flame status and system parameters in real time and adjust oxygen and nitrogen supply; Feedback mechanism: A feedback mechanism is established to feed back the effect of the adjusted oxygen and nitrogen supply to the system.
10. The control method of the oxygen lance nitrogen sealing system of a converter according to claim 1, characterized in that: In step 10, the specific implementation of the data integration and comprehensive analysis is as follows: Data integration: Integrate the flame state data obtained in the flame state monitoring and analysis steps with the data obtained in the traditional control parameter monitoring into the data platform; Data cleaning and preprocessing: Clean and preprocess the integrated data; Data analysis and modeling: Use data mining and machine learning techniques to analyze and model the integrated data and explore the relationships and patterns between data; Comprehensive analysis and decision support: Based on the results of data analysis, conduct comprehensive analysis to find key information and trends in the data.
Citation Information
Cited By
Visual management method and system for auxiliary transportation line of single-roadway driving face
CN122242954A
