A Method and System for Controlling Energy Consumption of a Metallurgical Equipment

Abstract

The present invention discloses an energy consumption control method and system for metallurgical equipment, relating to the technical field of energy consumption control of metallurgical equipment, and comprising the following steps: during the actual operation of an electric arc furnace, by installing a variety of sensors around the electrodes, the operation state of the electric arc is monitored in real time; key features reflecting the state of electric arc drift are extracted, and complex original data is converted into a quantitative index reflecting the electric arc drift and its severity; the key quantitative index is used as a feature vector, and a deep learning method is adopted to train the feature vector to construct an electric arc drift prediction model. After the construction is completed, based on the current input feature vector, an electric arc state prediction result is output to identify the electric arc drift. The present invention combines multi-sensors with deep learning to identify electric arc drift in advance, and through the adaptive regulation of the intelligent electrode height and response speed, improves the electric energy utilization rate, reduces the smelting time and equipment loss, and realizes low energy consumption, high efficiency, automation and green smelting.

Description

Technical Field

[0001] The present invention relates to the technical field of energy consumption control of metallurgical equipment, and particularly to an energy consumption control method and system for metallurgical equipment. Background Art

[0002] The energy consumption control of metallurgical equipment refers to the management strategy of realizing the rational utilization of energy and reducing the overall energy consumption by means of optimizing the process flow, improving the energy efficiency of equipment, reducing energy waste, etc. during the metallurgical production process. Specific measures include adopting advanced energy-saving technologies, such as high-efficiency heating systems, waste heat recovery devices, intelligent control systems, etc., to improve the operation efficiency of equipment; optimizing production scheduling to reduce no-load and standby energy consumption; promoting clean energy and new energy to replace traditional high-energy-consuming energy; and implementing energy consumption monitoring and intelligent regulation systems to realize real-time monitoring, analysis, and optimization of energy consumption. By these means, not only can the production cost be reduced, but also carbon emissions can be reduced, promoting the green and low-carbon development of the metallurgical industry.

[0003] As an important metallurgical equipment, an electric arc furnace can real-time monitor the arc state through an intelligent control system, and dynamically adjust the input power according to key parameters such as arc length, electrode position, furnace temperature, melting stage, etc., so that the arc always maintains the optimal power level. The intelligent control system uses sensors and automation algorithms, which can quickly respond to arc fluctuations, optimize electrode lifting, adjust current and voltage, and prevent the increase in energy consumption caused by arc instability. In addition, the system can also combine with electric energy load management, optimize the peak-valley electricity price strategy, reasonably allocate the use of electric energy, and reduce ineffective energy consumption. Through these intelligent control means, the electric arc furnace can not only improve the energy utilization efficiency, reduce the smelting time, but also reduce electrode consumption and equipment wear, thus achieving the goals of energy conservation and consumption reduction and green smelting.

[0004] The prior art has the following deficiencies:

[0005] During the actual operation of the electric arc furnace, the phenomenon of arc drift often occurs in the arc state. Although the prior art can dynamically adjust the electrode height, due to the inability of the electrode response speed to be adaptively adjusted, when the adjustment speed cannot meet the requirements, the arc drift cannot be corrected in time, resulting in abnormal fluctuations in the arc length. In this case, arc extinction (arc break) or excessive arc length (electric energy diffusion) may be caused, seriously reducing the electric energy utilization efficiency. More seriously, the arc cannot effectively transfer heat to the molten pool, and a large amount of input electric energy is consumed ineffectively, resulting in a significant increase in the unit energy consumption (kWh / t), directly pushing up the smelting cost and reducing the production efficiency.

[0006] The above information disclosed in the background art section is only used to strengthen the understanding of the background of the present disclosure, so it may include information that does not constitute the prior art known to those of ordinary skill in the art. Summary of the Invention

[0007] The object of the present invention is to provide an energy consumption control method and system for metallurgical equipment, which can achieve precise monitoring and data acquisition of the arc operation state through multi-sensor fusion technology. Combining with a deep learning prediction model, it can identify and predict the severity of arc drift before or just when it occurs, ensuring the timeliness and accuracy of adjustment response. At the same time, by introducing intelligent electrode height adjustment and adaptive response speed optimization, the lifting rate of the electrode can be dynamically adjusted according to the degree of arc drift, enabling the arc to quickly return to the ideal state, preventing the arc from being too long or too short, thereby improving the power utilization rate and reducing problems such as extended smelting time, furnace lining damage, and abnormal electrode consumption caused by arc instability, so as to solve the problems in the above-mentioned background technology.

[0008] To achieve the above object, the present invention provides the following technical solution: An energy consumption control method for metallurgical equipment, comprising the following steps:

[0009] During the actual operation of the electric arc furnace, by installing a variety of sensors around the electrode, the arc operation state is monitored in real time;

[0010] Extract the key features reflecting the arc drift state, and convert the complex original data into quantitative indicators reflecting the arc drift and its severity;

[0011] Use the key quantitative indicators as feature vectors, and adopt deep learning methods to train the feature vectors to construct an arc drift prediction model. After the arc drift prediction model is constructed, based on the current input feature vector, output the arc state prediction result to identify the arc drift;

[0012] When the arc drift prediction model detects the risk of arc drift, adjust the electrode height according to the preset adjustment strategy, dynamically change the height of the electrode, so that the arc quickly returns to the ideal position, correct the arc position in real time, and at the same time, intelligently adjust the electrode response speed for different severities of arc drift.

[0013] Preferably, extract the key features reflecting the arc drift state, wherein the extracted features include the deviation between the actual arc position and the ideal position and the temperature distribution around the arc. Within the monitoring window, through feature processing technology, the deviation between the actual arc position and the ideal position and the temperature distribution around the arc are converted into an arc offset index and an arc hot spot index, and the arc drift and its severity are reflected through the arc offset index and the arc hot spot index.

[0014] Preferably, use the arc offset index and the arc hot spot index as feature vectors, and adopt a deep learning algorithm to train the feature vectors to construct an arc drift prediction model. After the arc drift prediction model is constructed, output a drift index based on the current input feature vector, and use the drift index as the arc state prediction result to identify the arc drift.

[0015] Preferably, when predicting the arc state through the arc drift prediction model, the drift index generated is compared and analyzed with a preset drift index reference threshold to identify arc drift. The specific steps are as follows:

[0016] If the drift index is greater than the drift index reference threshold, it indicates that there is a risk of arc drift in the electric arc furnace. If the drift index is less than or equal to the drift index reference threshold, it indicates that the arc in the electric arc furnace operates stably and there is no risk of arc drift.

[0017] Preferably, within the monitoring window, the deviation between the actual arc position and the ideal position is converted into an arc offset index through feature processing technology. The specific steps are as follows:

[0018] To quantify the degree of arc offset, first establish a spatial deviation model between the actual arc position and the ideal arc position. Let the arc position be represented in three-dimensional space as and the ideal arc position be Then the spatial deviation vector D of the arc is defined as: ;

[0019] The non-linear normalization method is used to map the spatial offset to an exponential scale. The specific expression is: , where represents the normalized spatial offset vector;

[0020] After obtaining the normalized spatial offset vector , an arc offset index is constructed by the weighted geometric mean method. The constructed expression is: , where: is the arc offset index, , , respectively represent the weight coefficients in the x, y, and z axis directions, which are used to adjust the offset contributions in different directions.

[0021] Preferably, within the monitoring window, the temperature distribution around the arc is converted into an arc hot spot index through feature processing technology. The specific steps are as follows:

[0022] Within the monitoring window, use a high-resolution thermal imaging sensor to capture the temperature field data around the arc to form a spatial temperature distribution matrix , where represents the pixel coordinates in the thermal image. To accurately identify the arc hot spot area, a temperature gradient aggregation function is used to calculate the spatial change rate of the temperature field around the arc. The calculated expression is: , where, represents the global gradient aggregation value of the temperature field, and Calculate the second-order derivatives of temperature in the horizontal and vertical directions to reflect the regions with drastic changes in the temperature field. is a weighting function that gives a higher weight to the central region of the arc to enhance the signal response within the arc influence range.

[0023] Use the global gradient aggregation value of the temperature field to calculate the arc hot spot index. The calculation formula is: , where: is the arc hot spot index, which is used to measure the arc drift and its severity. is an adjustment coefficient, which is used to control the influence degree of the hot spot density on the final arc hot spot index. is the arc hot spot density index, which is defined as: , where is the temperature gradient, is the area of the region around the arc, which is used for normalization. The arc hot spot density index is used to measure the local hot spot density of the temperature field.

[0024] Preferably, when the arc drift prediction model detects the arc drift risk, adjust the electrode height according to the preset adjustment strategy, and intelligently adjust the electrode response speed for different severities of arc drift. The specific steps are as follows:

[0025] When the arc drift prediction model detects that the arc drift index exceeds the reference threshold, immediately adjust the electrode height. The electrode height adjustment keeps the arc length within the optimal range. The adjustment calculation formula is as follows: , where: is the adjusted electrode height, is the current electrode height, is the electrode height adjustment coefficient, which determines the sensitivity of the electrode adjustment after the drift index exceeds the drift index reference threshold. is the drift index predicted in real time by the arc drift prediction model, which is used to quantify the current arc drift degree. is the drift index reference threshold, which is used to judge the safety limit of the drift risk. is the arc length change trend, is the sign function, which is used to determine the adjustment direction.

[0026] Preferably, on the basis of completing the electrode height adjustment, dynamically adjust the electrode response speed to adapt to different severities of arc drift. The calculation formula for the dynamic adjustment of the electrode response speed is as follows: , where: is the dynamically adjusted electrode adjustment speed, is the basic electrode adjustment speed, which is used for the electrode lifting and lowering rate under normal conditions. is the speed gain coefficient, which is used to adjust the increase amplitude of the speed with the change of the drift index. is the degree to which the drift index exceeds the drift index reference threshold, which is used to dynamically calculate the additional acceleration ratio.

[0027] An energy consumption control system for a metallurgical device, including an arc state monitoring module, an arc drift feature extraction module, an arc drift prediction and identification module, and an electrode intelligent adjustment and dynamic response module;

[0028] The arc state monitoring module, during the actual operation of the electric arc furnace, installs various sensors around the electrode to monitor the arc operation state in real time;

[0029] The arc drift feature extraction module extracts the key features reflecting the arc drift state, and converts the complex original data into a quantitative index reflecting the arc drift and its severity;

[0030] The arc drift prediction and identification module uses the key quantitative index as a feature vector, and adopts a deep learning method to train the feature vector to construct an arc drift prediction model. After the arc drift prediction model is constructed, it outputs the arc state prediction result based on the current input feature vector to identify the arc drift;

[0031] The electrode intelligent adjustment and dynamic response module, when the arc drift prediction model detects the arc drift risk, adjusts the electrode height according to the preset adjustment strategy, and changes the electrode height dynamically to quickly restore the arc to the ideal position, correct the arc position in real time. At the same time, it intelligently adjusts the electrode response speed for different severities of arc drift.

[0032] In the above technical solution, the technical effects and advantages provided by the present invention are:

[0033] The present invention realizes the accurate monitoring and data collection of the arc operation state through the multi-sensor fusion technology. Combined with the deep learning prediction model, it can identify and predict the severity of the arc drift before or just after the arc drift occurs, ensuring the timeliness and accuracy of the adjustment response. At the same time, the introduction of intelligent electrode height adjustment and adaptive response speed optimization can dynamically adjust the lifting rate of the electrode according to the degree of arc drift, quickly restore the arc to the ideal state, prevent the arc from being too long or too short, thereby improving the electric energy utilization rate and reducing problems such as extended smelting time, furnace lining damage, and abnormal electrode consumption caused by arc instability. In addition, through the real-time adjustment of the intelligent control system, the dependence on manual intervention is reduced, the automation and intelligence level of the smelting process are improved, enabling metallurgical enterprises to reduce production costs while improving the overall production efficiency and equipment safety, and promoting the development of the metallurgical industry towards energy conservation, intelligence, and greenness. Brief Description of the Drawings

[0034] To more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments recorded in the present invention. For those of ordinary skill in the art, other drawings can also be obtained based on these drawings.

[0035] Figure 1 It is a method flow chart of an energy consumption control method for a metallurgical device of the present invention.

[0036] Figure 2 It is a module schematic diagram of an energy consumption control system for a metallurgical device of the present invention. Specific implementation manners

[0037] Now, the exemplary embodiments will be described more fully with reference to the accompanying drawings. However, the exemplary embodiments can be implemented in various forms and should not be construed as limited to the examples set forth herein; rather, these exemplary embodiments are provided so that the present disclosure will be more thorough and complete, and will fully convey the concept of the exemplary embodiments to those skilled in the art.

[0038] The present invention provides an energy consumption control method for a metallurgical device as shown in Figure 1 and includes the following steps:

[0039] During the actual operation of the electric arc furnace, by installing various sensors (such as current, voltage sensors, optical cameras, temperature sensors, vibration monitoring devices, etc.) around the electrodes, the operating state of the electric arc is monitored in real time;

[0040] The sensors will transmit the information such as the arc voltage, arc current, arc light intensity, temperature distribution in the furnace, and furnace wall vibration collected to the data acquisition system. Through the real-time acquisition of this multi-source data, the dynamic changes of the electric arc in the electric arc furnace can be comprehensively reflected, providing a basis for subsequent data analysis and prediction.

[0041] Extract the key features reflecting the arc drift state, and convert the complex original data into quantitative indicators that can reflect the arc drift and its severity;

[0042] Extract the key features reflecting the arc drift state. Among them, the extracted features include the deviation between the actual arc position and the ideal position and the temperature distribution around the arc. Within the monitoring window, through feature processing technology, the deviation between the actual arc position and the ideal position and the temperature distribution around the arc are converted into an arc offset index and an arc hot spot index, and the arc drift and its severity are reflected through the arc offset index and the arc hot spot index.

[0043] When the deviation between the actual arc position and the ideal position is large, it usually indicates that the arc in the arc furnace may have deviated from the target smelting area, presenting a potential risk of arc drift. The ideal position is typically the optimal smelting area set according to process requirements and equipment structure, which can ensure that heat is concentrated and effectively transferred to the molten bath. When the deviation expands, the arc may approach the furnace wall or deviate from the center of the molten bath, resulting in reduced thermal efficiency, extended smelting time, and potentially excessive lining wear or other instability factors in the furnace. Therefore, by monitoring and analyzing the deviation between the actual arc position and the ideal position, the phenomenon of arc drift can be identified and prevented early, and adjustment measures can be taken in a timely manner.

[0044] In the monitoring window, the deviation between the actual arc position and the ideal position is converted into an arc offset index through feature processing technology. The specific steps are as follows:

[0045] To quantify the degree of arc offset, first, a spatial deviation model between the actual arc position and the ideal arc position needs to be established. Suppose the arc position is represented in a three-dimensional space as , and the ideal arc position is , then the spatial deviation vector D of the arc is defined as: ;

[0046] A non-linear normalization method is used to map the spatial offset to an exponential scale. The specific expression is: , where represents the normalized spatial offset vector, and the exponential transformation in the form of , and can enhance the sensitivity to small offsets, so that small offsets will not have too much impact, while large offsets can be significantly amplified, thereby improving the accuracy of arc drift detection;

[0047] The function of this step is to convert the spatial information of arc offset into a normalized feature, making subsequent calculations more robust.

[0048] After obtaining the normalized spatial offset vector , an arc offset index is constructed through the weighted geometric mean method. The constructed expression is: , where: is the arc offset index, , , respectively represent the weight coefficients along the x, y, and z axes, which are used to adjust the offset contributions in different directions. They can be adaptively adjusted according to factors such as the magnetic field in the furnace and the shape of the molten bath. The +1 ensures that even if the offset value in a certain direction is 0, the calculation of the arc offset index is still valid. Taking the geometric mean can reduce the influence of outliers in a single direction and improve the stability of the arc offset index. The final -1 is used to restore the original scale, so that when the offset is zero ;

[0049] The function of this step is to aggregate the arc offset data into an overall offset index, which can quantify the degree of arc drift. The larger the value, the more serious the offset. Due to the use of exponential mapping and weighted geometric mean, this arc offset index is more tolerant of small offsets and more sensitive to severe drifts, which helps the intelligent control system make accurate judgments.

[0050] As can be seen from the arc offset index, within the monitoring window, the deviation between the actual arc position and the ideal position is converted into an arc offset index through feature processing technology. The larger the performance value of the arc offset index, the greater the deviation between the actual arc position and the ideal position, indicating a higher risk of arc drift, which may lead to problems such as uneven heat distribution, decreased power utilization efficiency, and even damage to the furnace lining. On the contrary, when the arc offset index is small, it indicates that the arc position is relatively stable and the drift risk is low, indicating that the electric arc furnace is in a good operating state. Therefore, the size of the arc offset index can be used as an important basis for the intelligent control system to identify the arc stability, so as to achieve precise energy consumption optimization and smelting control.

[0051] When the temperature distribution around the arc is uneven and hot spots are obvious, it usually indicates that there is a potential risk of arc drift in the electric arc furnace. During normal operation of the arc, its heat transfer should be relatively uniform and mainly concentrated in the molten pool area to ensure uniform heating and rapid melting of the metal. However, when arc drift occurs, local high-temperature hot spots (thermal spots) may form in some areas due to heat accumulation, while in other areas, insufficient melting may occur due to insufficient heat. This uneven temperature distribution phenomenon may be caused by the offset of the arc position, abnormal fluctuations in the arc length, or uneven distribution of the furnace charge, further exacerbating the instability of the arc. In addition, the appearance of thermal spots may also cause local overheating of the furnace lining, accelerate the damage of the furnace lining, increase the maintenance cost and the risk of furnace shutdown. Therefore, monitoring the uniformity of the temperature distribution around the arc and the situation of thermal spots can be used as one of the important features to identify arc drift and provide early warning for the intelligent control system.

[0052] Within the monitoring window, the temperature distribution around the arc is converted into an arc thermal spot index through feature processing technology. The specific steps are as follows:

[0053] Within the monitoring window, use a high-resolution thermal imaging sensor to capture the temperature field data around the arc to form a spatial temperature distribution matrix , where represents the pixel coordinates in the thermal image. In order to accurately identify the arc thermal spot area, a temperature gradient aggregation function is used to calculate the spatial change rate of the temperature field around the arc. The calculation expression is: , where, represents the global gradient aggregation value of the temperature field, and Calculate the second-order derivatives of temperature in the horizontal and vertical directions to reflect the regions with drastic changes in the temperature field. is the weighting function, which gives a higher weight to the central region of the arc to enhance the signal response within the influence range of the arc.

[0054] The function of this step is to identify the regions where the abnormal temperature gradients are concentrated, so as to detect the temperature change trend around the arc and provide a basis for the calculation of the subsequent arc hot spot index.

[0055] Use the global gradient aggregation value of the temperature field to calculate the arc hot spot index. The calculation formula is: , where: is the arc hot spot index, which is used to measure the arc drift and its severity. is the adjustment coefficient, which is used to control the influence degree of the hot spot density on the final arc hot spot index. is the arc hot spot density index, which is defined as: , where is the temperature gradient. is the area of the region around the arc, which is used for normalization. The arc hot spot density index is used to measure the local hot spot density of the temperature field.

[0056] The function of this step is to accurately quantify the severity of the arc hot spot. Among them, the global gradient aggregation value of the temperature field reflects the drastic change degree of the temperature field around the arc and can capture the inhomogeneity of the temperature distribution. While the arc hot spot density index measures the local density of the temperature gradient, ensuring that not only the temperature anomalies are identified, but also the concentration degree of the anomalies can be quantified.

[0057] As can be seen from the arc hot spot index, within the monitoring window, the temperature distribution around the arc is converted into an arc hot spot index through feature processing technology. The larger the value of the arc hot spot index, the greater the risk of arc drift in the electric arc furnace. Conversely, it indicates that the arc state of the electric arc furnace is more stable and the drift risk is smaller. This arc hot spot index quantifies the non-uniformity of the temperature field and the density of local hot spots by performing feature processing on the temperature distribution around the arc. When arc drift occurs, the heat transfer path shifts, resulting in abnormally high temperatures (hot spots) in some areas and insufficient temperatures in other areas. This phenomenon is manifested as a higher value in the arc hot spot index. On the other hand, when the arc is stable, the temperature distribution is more uniform, there are fewer hot spot areas, and the index value is lower, indicating that the heat transfer path of the arc remains reasonable. Therefore, the arc hot spot index can not only reflect the existence of arc drift, but also quantify the severity of the drift by its numerical value, providing real-time feedback for the intelligent control system to dynamically adjust the electrode height and response speed, ensuring that the arc is always in the best working state, improving the energy consumption utilization rate and reducing equipment losses.

[0058] Taking the key quantitative indicators as feature vectors and using deep learning algorithms (such as recurrent neural networks, RNNs) to train the feature vectors to build an arc drift prediction model. After the arc drift prediction model is built, based on the current input feature vectors, the arc state prediction results are output to identify arc drift;

[0059] Taking the arc offset index and the arc hot spot index as feature vectors and using deep learning methods to train the feature vectors to build an arc drift prediction model. After the arc drift prediction model is built, based on the current input feature vectors, the drift index is output, and the drift index is used as the arc state prediction result to identify arc drift.

[0060] Taking the arc offset index and the arc hot spot index as feature vectors and using deep learning algorithms to train the feature vectors to build an arc drift prediction model. Specifically:

[0061] First, systematic data acquisition and processing of the arc offset index and the arc hot spot index are required to use them as feature vectors for the deep learning model. Specifically, the arc offset index is mainly quantified by comparing the ideal position and the actual position of the arc in the molten pool area. When the arc drifts, the arc offset index will increase significantly with the increase of the drift degree; while the arc hot spot index measures the concentration and intensity of the local high-temperature area by extracting the characteristics of the temperature distribution around the arc. If the arc deviates significantly from the normal melting area, it often forms overheating or uneven temperature in a specific area, resulting in a sharp rise in the arc hot spot index. In actual deployment, first, internal data of the electric arc furnace is collected in real time through high-resolution sensors (such as thermal imaging, optical measurement, electrical signal acquisition, etc.), and these raw data are preprocessed, such as denoising, filtering, and normalization, to ensure the accuracy and consistency of the data. Subsequently, based on a carefully designed algorithm, multi-dimensional analysis of the temperature field distribution and the arc position information is carried out to obtain numerical indicators that can directly reflect the arc offset and the hot spot degree, and these two numerical indicators are recorded in the form of "arc offset index" and "arc hot spot index" respectively. After combining them into a two-dimensional or multi-dimensional feature vector, auxiliary parameters that are helpful for judging the arc state (such as arc voltage fluctuation, molten pool surface temperature difference, etc.) can be added according to needs to jointly form richer input features. Through such data preparation and feature engineering, a solid foundation can be laid for the subsequent deep learning model training, making the model more targeted to capture abnormal changes in the electric arc furnace.

[0062] After the construction of the feature vector is completed, deep learning methods (such as convolutional neural network, recurrent neural network, or a hybrid architecture integrating multi-layer perceptrons) can be used to train these feature vectors to establish an arc drift prediction model. During the training process, the historical arc operation data and the corresponding arc drift labels (or discrimination criteria) collected will be divided into a training set and a validation set. Through repeated iterations of forward propagation and backward propagation, the model gradually learns the non-linear mapping relationship between the arc offset index, the arc hot spot index, and the arc drift risk. The multi-layer network structure of deep learning can extract higher-level feature combinations in the hidden layer, enabling the model to not only identify simple numerical increases and decreases, but also mine the additional information brought by the interaction or superposition of the two, so as to make a more accurate prediction of the arc drift. When the training converges and obtains an ideal effect in the validation set, the model can receive the arc offset index and the arc hot spot index calculated in real time during the online monitoring stage, and output the prediction results of the occurrence probability and severity of the arc drift after network calculation.

[0063] The arc drift prediction model is not specifically limited here, and it can realize comprehensive analysis of the arc offset index and the arc hot spot index to generate a drift index Any arc drift prediction model can be used. To implement the technical solution of the present invention, the present invention provides a specific implementation; the drift index The generated expression is: , where , are respectively the preset proportionality coefficients of the arc offset index and the arc hot spot index , and , are both greater than 0.

[0064] It can be seen from the drift index that within the monitoring window, through the feature processing technology, the deviation between the actual arc position and the ideal position is converted into the arc offset index. The larger the value of the arc offset index, the temperature distribution around the arc is converted into the arc hot spot index through the feature processing technology. The larger the value of the arc hot spot index, that is, the larger the value of the drift index generated when predicting the arc state through the arc drift prediction model, it indicates that the risk of arc drift in the electric arc furnace is greater. On the contrary, it indicates that the arc state of the electric arc furnace is more stable and the drift risk is smaller.

[0065] Compare and analyze the drift index generated when predicting the arc state through the arc drift prediction model with the preset drift index reference threshold to identify arc drift. The specific steps are as follows:

[0066] If the drift index is greater than the drift index reference threshold, it indicates that there is a risk of arc drift in the electric arc furnace. If the drift index is less than or equal to the drift index reference threshold, it indicates that the arc of the electric arc furnace operates stably and there is no risk of arc drift.

[0067] When the arc drift prediction model detects the risk of arc drift, adjust the electrode height according to the preset adjustment strategy, dynamically change the height of the electrode, so that the arc quickly returns to the ideal position, correct the arc position in real time, prevent the arc from being too long or extinguished, improve the heat transfer efficiency, reduce energy consumption waste, and at the same time protect the furnace lining and electrodes from overheating damage. At the same time, intelligently adjust the electrode response speed for different severities of arc drift. When detecting a relatively slight drift, maintain the normal response speed; in the case of severe drift, accelerate the response speed of the electrode lifting to ensure more rapid and accurate adjustment;

[0068] When the arc drift prediction model detects the risk of arc drift, adjust the electrode height according to the preset adjustment strategy, and intelligently adjust the electrode response speed for different severities of arc drift. The specific steps are as follows:

[0069] When the arc drift prediction model detects that the arc drift index exceeds the reference threshold, immediately adjust the electrode height to quickly return the arc to the ideal position. The goal of adjusting the electrode height is to maintain the arc length within the optimal range, avoiding an overly long arc (energy dissipation) or an overly short arc (extinction). The calculation formula for adjusting the electrode height is as follows: , where: is the adjusted electrode height, is the current electrode height, is the electrode height adjustment coefficient, which determines the sensitivity of the electrode adjustment after the drift index exceeds the drift index reference threshold, is the drift index predicted in real time by the arc drift prediction model, used to quantify the current arc offset degree, is the drift index reference threshold, used to judge the safety limit of the drift risk, is the arc length change trend. If the arc is too long, the electrode height should be reduced; if the arc is too short, the electrode height should be increased, is the sign function, used to determine the adjustment direction (a positive value indicates lowering the electrode, and a negative value indicates raising the electrode);

[0070] The arc length change trend refers to the direction and amplitude of the arc length change over time and operating conditions during the operation of the electric arc furnace. It can reflect whether the arc is in a stable state or shows unstable conditions such as elongation or shortening. The arc length is mainly affected by factors such as electrode height, arc voltage, current intensity, and charge distribution. When the arc is offset or disturbed, the arc may be abnormally elongated (energy dissipation, reduced heat transfer efficiency) or suddenly shortened (local overheating, increased risk of arc extinction). To quantify this trend, the arc length change trend can be calculated, and its mathematical expression can be defined as: , where: is the arc length of the current monitoring window, is the arc length of the previous monitoring window;

[0071] If the arc length change trend , it indicates that the arc is elongating and the electrode may need to be adjusted downward to shorten the arc; if the arc length change trend , it indicates that the arc is shortening and the electrode may need to be adjusted upward to lengthen the arc. By real-time monitoring the arc length change trend, the electrode position can be adjusted in time before or just after the arc drift occurs to ensure arc stability, optimize the energy transfer efficiency, and reduce energy consumption losses and equipment damage during the smelting process.

[0072] The function of this step is to real-time correct the arc position, ensure that the arc heat is transferred to the correct area of the molten bath, improve the smelting efficiency, reduce energy consumption losses, and at the same time reduce overheating damage to the electrodes and furnace lining.

[0073] On the basis of completing the electrode height adjustment, it is also necessary to dynamically adjust the electrode response speed to adapt to arc drifts of different severities. When the drift degree is relatively light, the system maintains a normal response speed; while when the drift is severe, it is necessary to accelerate the lifting and lowering of the electrode to ensure faster and more accurate adjustment. The calculation formula for dynamically adjusting the electrode response speed is as follows: , where: is the electrode adjustment speed after dynamic adjustment, is the basic electrode adjustment speed, which is used for the lifting and lowering rate of the electrode under normal conditions, is the speed gain coefficient, which is used to adjust the increase amplitude of the speed with the change of the drift index, is the degree to which the drift index exceeds the drift index reference threshold, which is used to dynamically calculate the additional acceleration ratio;

[0074] The function of this step is to improve the sensitivity of the adjustment, ensure that the greater the arc drift degree, the faster the adjustment speed of the electrode, so as to avoid energy consumption waste, extended smelting time or equipment damage caused by slow response. This method ensures the dual control of the electrode height and response speed, enabling the electric arc furnace to maintain the best smelting effect and energy utilization rate under different operating conditions.

[0075] Through the above solution, it is possible to effectively improve the arc stability during the operation of the electric arc furnace, reduce the energy consumption waste caused by arc drift, improve the smelting efficiency, and extend the service life of the equipment. Specifically, this method uses multi-sensor fusion technology to achieve precise monitoring and data collection of the arc operating state. Combining with the deep learning prediction model, it can identify and predict the severity of the arc drift before or just after it occurs, ensuring the timeliness and accuracy of the adjustment response. At the same time, this solution introduces intelligent electrode height adjustment and adaptive response speed optimization, which can dynamically adjust the lifting and lowering rate of the electrode according to the arc drift degree, quickly restore the arc to the ideal state, prevent the arc from being too long (resulting in energy diffusion) or too short (causing arc extinction), thereby improving the electric energy utilization rate (reducing the unit energy consumption kWh / t), and reducing problems such as extended smelting time, furnace lining damage, and abnormal electrode consumption caused by arc instability. In addition, through the real-time adjustment of the intelligent control system, this method reduces the dependence on manual intervention, improves the automation and intelligence level of the smelting process, enables metallurgical enterprises to reduce production costs while improving the overall production efficiency and equipment safety, and promotes the development of the metallurgical industry towards energy conservation, intelligence and greenness.

[0076] The present invention provides an energy consumption control system for a metallurgical device as shown in Figure 2 , including an arc state monitoring module, an arc drift feature extraction module, an arc drift prediction and identification module, and an electrode intelligent adjustment and dynamic response module;

[0077] The arc state monitoring module monitors the real-time operation state of the arc by installing various sensors around the electrode during the actual operation of the electric arc furnace;

[0078] The arc drift feature extraction module extracts the key features reflecting the arc drift state, and converts the complex original data into quantitative indicators reflecting the arc drift and its severity;

[0079] The arc drift prediction and identification module uses the key quantitative indicators as feature vectors and adopts deep learning methods to train the feature vectors to construct an arc drift prediction model. After the arc drift prediction model is constructed, based on the current input feature vectors, it outputs the arc state prediction results to identify the arc drift;

[0080] The electrode intelligent adjustment and dynamic response module, when the arc drift prediction model detects the risk of arc drift, adjusts the electrode height according to the preset adjustment strategy, dynamically changes the height of the electrode, so that the arc quickly returns to the ideal position, corrects the arc position in real time, and at the same time, intelligently adjusts the electrode response speed for different severities of arc drift.

[0081] The energy consumption control method of a metallurgical device provided by an embodiment of the present invention is realized through the above-mentioned energy consumption control system of a metallurgical device. The specific methods and processes of the energy consumption control system of a metallurgical device are detailed in the embodiments of the above-mentioned energy consumption control method of a metallurgical device, and will not be elaborated here.

[0082] The above formulas are all dimensionless and take their numerical calculations. The formula is obtained by collecting a large amount of data for software simulation to obtain a formula closest to the real situation. The preset parameters in the formula are set by those skilled in the art according to the actual situation.

[0083] Only some exemplary embodiments of the present invention have been described by way of illustration above. Undoubtedly, for those of ordinary skill in the art, without departing from the spirit and scope of the present invention, the described embodiments can be modified in various different ways. Therefore, the above drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

[0084] Those of ordinary skill in the art can realize that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are executed in hardware or software depends on the specific application and design constraints of the technical solution. Professional technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of this application.

[0085] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the systems, devices, and units described above can refer to the corresponding processes in the foregoing method embodiments and will not be elaborated herein.

[0086] In several embodiments provided in the present application, it should be understood that the disclosed systems and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the units is only a logical function division, and there can be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed couplings or direct couplings or communication connections to each other can be through some interfaces, and the indirect couplings or communication connections of the devices or units can be in electrical, mechanical, or other forms.

[0087] The units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they can be located in one place or distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0088] In addition, in each embodiment of the present application, the functional units can be integrated into one processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit.

[0089] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, in essence, or the part that contributes to the prior art or a part of this technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in each embodiment of the present application. The foregoing storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical discs that can store program codes.

[0090] As described above, it is only the specific implementation manner of this application, but the protection scope of this application is not limited thereto. Any person skilled in the art within the technical scope disclosed by this application can easily think of changes or substitutions, which should all be covered within the protection scope of this application. Therefore, the protection scope of this application shall be subject to the protection scope of the claims described.

Claims

1. A method for controlling the energy consumption of a metallurgical device, characterized in that, It includes the following steps: During the actual operation of the electric arc furnace, by installing various sensors around the electrode, the operation state of the electric arc is monitored in real time; Extract the key features reflecting the arc drift state, and convert the complex original data into quantitative indicators reflecting the arc drift and its severity; Use the key quantitative indicators as feature vectors, and adopt deep learning methods to train the feature vectors to construct an arc drift prediction model. After the arc drift prediction model is constructed, based on the current input feature vector, output the arc state prediction result to identify the arc drift; When the arc drift prediction model detects the arc drift risk, adjust the electrode height according to the preset adjustment strategy, and dynamically change the height of the electrode to quickly restore the arc to the ideal position, correct the arc position in real time. At the same time, intelligently adjust the electrode response speed for different severity levels of arc drift; Extract the key features reflecting the arc drift state. Among them, the extracted features include the deviation between the actual arc position and the ideal position and the temperature distribution around the arc. In the monitoring window, through feature processing technology, the deviation between the actual arc position and the ideal position and the temperature distribution around the arc are converted into an arc offset index and an arc hot spot index, and the arc drift and its severity are reflected through the arc offset index and the arc hot spot index.

2. The energy consumption control method of a metallurgical device according to claim 1, characterized in that, Use the arc offset index and the arc hot spot index as feature vectors, and adopt deep learning methods to train the feature vectors to construct an arc drift prediction model. After the arc drift prediction model is constructed, output the drift index based on the current input feature vector, and use the drift index as the arc state prediction result to identify the arc drift.

3. The energy consumption control method of a metallurgical device according to claim 2, characterized in that, Compare and analyze the drift index generated when predicting the arc state through the arc drift prediction model with the preset drift index reference threshold to identify the arc drift. The specific steps are as follows: If the drift index is greater than the drift index reference threshold, it indicates that there is an arc drift risk in the electric arc furnace. If the drift index is less than or equal to the drift index reference threshold, it indicates that the arc in the electric arc furnace operates stably and there is no arc drift risk.

4. The energy consumption control method of a metallurgical device according to claim 1, characterized in that, In the monitoring window, convert the deviation between the actual arc position and the ideal position into an arc offset index through feature processing technology. The specific steps are as follows: To quantify the degree of arc deviation, first establish a spatial deviation model between the actual arc position and the ideal arc position. Let the arc position be represented in three-dimensional space as , and the ideal arc position be . Then the spatial deviation vector D of the arc is defined as: ; The nonlinear normalization method is used to map the spatial offset to the exponential scale. The specific expression is: ,in, represents the normalized spatial offset vector; After obtaining the normalized space offset vector Afterwards, the arc offset index is constructed by the weighted geometric mean method, and the constructed expression is: ,in: is the arc offset indicator, 、 、 Represents the weight coefficients along the x, y, and z axes, respectively, which are used to adjust the offset contribution in different directions.

5. A method for controlling the energy consumption of a metallurgical device according to claim 1, characterized in that, In the monitoring window, convert the temperature distribution around the arc into an arc hot spot index through feature processing technology. The specific steps are as follows: In the monitoring window, high-resolution thermal imaging sensors are used to capture the temperature field data around the arc to form a spatial temperature distribution matrix. ,in Represents the pixel coordinates in the thermal image. In order to accurately identify the arc hot spot area, the temperature gradient aggregation function is used to calculate the spatial change rate of the temperature field around the arc. The calculation expression is: ,in, Represents the global gradient aggregation value of the temperature field, and Calculate the second-order derivatives of temperature in the horizontal and vertical directions to reflect the areas with drastic changes in the temperature field. is a weighting function that gives a higher weight to the arc center area to enhance the signal response within the arc influence range; Using the global gradient aggregation value of the temperature field to calculate the arc hot spot index, and the calculation expression is: , where: is the arc hot spot index, which is used to measure the arc drift and its severity, is the adjustment coefficient, which is used to control the influence degree of the hot spot density on the final arc hot spot index, is the arc hot spot density index, which is defined as: , where is the temperature gradient, is the area of the area around the arc for normalization processing, and the arc hot spot density index is used to measure the local hot spot density of the temperature field.

6. The energy consumption control method of metallurgical equipment according to claim 3, characterized in that: When the arc drift prediction model detects the arc drift risk, adjust the electrode height according to the preset adjustment strategy, and intelligently adjust the electrode response speed for different severity levels of arc drift. The specific steps are as follows: When the arc drift prediction model detects that the arc drift index exceeds the reference threshold, immediately adjust the electrode height. The adjustment of the electrode height maintains the arc length within the optimal range. The calculation formula for the adjustment is as follows: , where: is the adjusted electrode height, is the current electrode height, is the electrode height adjustment coefficient, which determines the sensitivity of the electrode adjustment after the drift index exceeds the drift index reference threshold, is the drift index predicted in real time by the arc drift prediction model, which is used to quantify the current degree of arc drift, is the drift index reference threshold, which is used to judge the safety limit of the drift risk, is the arc length change trend, is the sign function, which is used to determine the adjustment direction.

7. The energy consumption control method of metallurgical equipment according to claim 3, characterized in that: On the basis of completing the electrode height adjustment, dynamically adjust the electrode response speed to adapt to arc drifts of different severities. The calculation formula for the dynamic adjustment of the electrode response speed is as follows: , where: is the electrode adjustment speed after dynamic adjustment, is the basic electrode adjustment speed, which is used for the electrode lifting rate under normal conditions, is the speed gain coefficient, which is used to adjust the increase in speed with the change of the drift index, is the degree to which the drift index exceeds the drift index reference threshold and is used to dynamically calculate the additional acceleration ratio.

8. An energy consumption control system for a metallurgical device, which is used to implement the energy consumption control method of the metallurgical device described in any one of the above claims 1-7, characterized in that, It includes an arc state monitoring module, an arc drift feature extraction module, an arc drift prediction and identification module, and an electrode intelligent adjustment and dynamic response module; The arc state monitoring module, during the actual operation of the electric arc furnace, monitors the operation state of the electric arc in real time by installing various sensors around the electrode; The arc drift feature extraction module extracts the key features reflecting the arc drift state, and converts the complex original data into quantitative indicators reflecting the arc drift and its severity; The arc drift prediction and identification module uses key quantitative indicators as feature vectors and adopts deep learning methods to train the feature vectors to construct an arc drift prediction model. After the construction of the arc drift prediction model is completed, the arc state prediction result is output based on the current input feature vector to identify arc drift; The electrode intelligent adjustment and dynamic response module, when the arc drift prediction model detects the risk of arc drift, adjusts the electrode height according to the preset adjustment strategy, dynamically changes the height of the electrode, so that the arc quickly returns to the ideal position, corrects the arc position in real time. At the same time, the electrode response speed is intelligently adjusted for different severities of arc drift.

Images