An intelligent management system and method for smelting equipment

Abstract

The invention discloses an intelligent management system and method for smelting equipment, which relates to the field of smelting technology. Infrared imaging detection is performed on a flash furnace reaction tower in real time to obtain a real-time temperature distribution image in the furnace and divide the area. According to the real-time temperature conditions of the area, the area with abnormally high temperature is determined. Hot spot analysis and slag shedding analysis are performed on the area with abnormally high temperature to identify the cause of the abnormally high temperature phenomenon. If the cause of the abnormally high temperature phenomenon is slag shedding, the change of raw material composition is controlled while the heat load is controlled to accelerate the formation of slag. Active intervention adjustment is adopted, which has the advantages of fast response and high efficiency, and the slag process can be quickly intervened by adjusting the batching composition, the relative position of the concentrate nozzle and the air regulating cone, the cooling water volume of the copper water jacket, etc., which can greatly shorten the slag cycle, thereby reducing the scouring and erosion of the reaction tower body by high-intensity reaction, ensuring the safety of refractory bricks and water jackets, and thus improving the intrinsic safety of metallurgical furnaces.

Description

Technical Field

[0001] The present invention relates to the field of smelting technology, and in particular to an intelligent management system and method applied to smelting equipment. Background Art

[0002] The flash furnace reaction tower consists of a central concentrate nozzle, a gas mixing chamber, and a reaction tower body. The main body is a vertical cylinder, which is composed of smelting slag, refractory bricks, a copper water jacket, and a steel outer shell from the inside to the outside. The reaction tower mainly relies on the slag on the tower wall to resist the erosion of the melt and flue gas and dust produced by the high-intensity smelting reaction in the tower. When the slag falls off or is too thin, it will cause irreversible burning damage to the tower wall. In severe cases, the water jacket will leak and the tower wall will burn through, resulting in shutdown. The conventional slag process has the disadvantages of low human intervention, low efficiency, thin slag that is easy to fall off, and large damage to the refractory brick copper water jacket, which makes the flash furnace unable to operate at full load. Due to the above problems, conventional natural slag cannot be normally applied to the efficient production of flash furnaces. Summary of the invention

[0003] The purpose of the present invention is to provide an intelligent management system and method for smelting equipment to solve the problems raised in the prior art.

[0004] To achieve the above-mentioned purpose, the present invention provides the following technical solutions: an intelligent management system applied to smelting equipment, comprising a data acquisition module, a slag control module, a data storage module and a data analysis module; the output end of the data acquisition module is connected to the input end of the data storage module, and is used to obtain real-time infrared imaging detection of the flash furnace reaction tower; the output end of the data storage module is connected to the input end of the data analysis module, and is used to store the infrared imaging data acquired by the data acquisition module, and to mark the infrared imaging data with slag shedding phenomenon and hot spot phenomenon; the data analysis module is used to detect abnormal temperature in the furnace, and analyze the cause of abnormally high temperature phenomenon; the slag control module is used to accelerate slag formation when slag shedding phenomenon occurs.

[0005] The slag control module also includes a raw material adjustment unit, a load adjustment unit and a cooling adjustment unit; the raw material adjustment unit is used to adjust the raw material composition, reduce the erosion of the reaction tower slag by the melt after the reaction, and improve the density of the formed slag layer; the load adjustment unit is used to reduce the thermal load of the slag shedding and thinning area, and reduce the intensity of the reaction; the cooling adjustment unit is used to enhance the cooling intensity of the slag shedding area. The data analysis module also includes a region division unit, a data integration unit, an estimation unit and a slag shedding judgment unit; the region division unit is used to divide the temperature distribution image into regions; the data integration unit is used to integrate and generate kernel density estimation and unsupervised classification data sets; the estimation unit is used to estimate the probability density function between the probability of occurrence of hot spot phenomenon or slag shedding phenomenon and temperature; the slag shedding judgment unit is used to identify the cause of abnormally high temperature phenomenon. The slag falling judgment unit obtains the change rate Δt1 of the average temperature of the target area before time t-dt, the change rate Δt2 of the average temperature of the target area after time t+dt, and the change rate Δt of the average temperature of the target area between time [t-dt, t+dt], and identifies the cause of the abnormally high temperature phenomenon according to Δbt and Δt, where Δbt is the difference between Δt1 and Δt2, t is the time when the area is identified as an area with abnormally high temperature, and dt is the time window.

[0006] To achieve the above object, the present invention provides the following technical solution: an intelligent management method for smelting equipment, comprising the following steps:

[0007] S11, performing infrared imaging detection on the flash furnace reaction tower in real time, obtaining a real-time temperature distribution image in the furnace, and performing regional division based on the real-time temperature distribution image in the furnace;

[0008] S12, determining an area with abnormally high temperature according to the real-time temperature conditions of the area;

[0009] S13, perform hot spot analysis and slag shedding analysis on the area with abnormally high temperature to identify the cause of the abnormally high temperature phenomenon; if the cause of the abnormally high temperature phenomenon is slag shedding, control the change of raw material composition and control the heat load at the same time to accelerate slag formation; if the cause of the abnormally high temperature phenomenon is not slag shedding, continue to monitor the temperature in the furnace.

[0010] Specifically, in step S11, the region division based on the real-time temperature distribution image in the furnace further includes the following steps:

[0011] S21, obtaining temperature data of each pixel in the real-time temperature distribution image in the furnace;

[0012] S22, randomly select K initial centroids to form classification clusters, K is a constant;

[0013] S23, calculating the Euclidean distance between each pixel and the centroid, and assigning the pixel to the classification cluster of the centroid with the smallest Euclidean distance; the Euclidean distance is calculated by the following formula: ; Where D represents the Euclidean distance, a, b and c are the horizontal coordinates, vertical coordinates and temperature data of the pixel point in the temperature distribution image, a0, b0 and c0 are the horizontal coordinates, vertical coordinates and temperature data of the centroid in the temperature distribution image, k1 is the position weight, and k2 is the temperature weight;

[0014] S24, calculating the average value of all pixels of each classification cluster, and taking the average value as the information centroid;

[0015] k1 and k2 can be adjusted, where k1 is used to make the area continuous, that is, pixels with adjacent positions are divided together; k2 is used to divide pixels with similar temperatures together.

[0016] S25, repeat steps S23 and S24 until the maximum number of iterations is reached.

[0017] In step S12, determining the area with abnormally high temperature according to the real-time temperature distribution of the area also includes the following steps:

[0018] The real-time temperature of all areas in the furnace is obtained, and the real-time temperature of the area is used as input to the probability density function g(x) between the probability of occurrence of hot spot phenomenon or slag shedding phenomenon and temperature to obtain the probability of occurrence of hot spot phenomenon or slag shedding phenomenon; if the probability of occurrence of hot spot phenomenon or slag shedding phenomenon is not less than the threshold, the area is judged to be an area with abnormally high temperature.

[0019] Areas where slag falls off or becomes thin usually have abnormal temperatures. The temperatures in these areas are much higher than those in other areas. This is due to the frequent temperature fluctuations in the edge areas caused by the unreasonable distribution of gas flow in the furnace, while the temperature in normal areas is relatively stable. In addition to slag falling off, the combustion conditions in the furnace can also lead to areas with abnormally high temperatures. Uneven combustion may cause local hot spots in certain areas of the furnace. In these hot spots, even if there is no slag falling off, the temperatures in these hot spots are much higher than those in other areas. For this reason, in addition to high-temperature detection, further analysis is required to determine whether slag falling off occurs.

[0020] Temperature changes in the slag shedding area may be sudden, as the protective effect of the slag layer is lost, causing the furnace wall temperature to rise rapidly. In this case, a sharp rise in temperature is an obvious sign of slag shedding; if the temperature of a certain area continues to rise gradually, this may indicate a hot spot area that is gradually formed due to uneven combustion; and if the temperature record shows a sudden peak or a sharp rise, this may indicate that the protective effect of the slag layer is suddenly lost, causing the furnace wall temperature to rise rapidly, which is usually a sign of slag shedding.

[0021] The probability density function g(x) between the probability of occurrence of hot spot phenomenon or slag shedding phenomenon and temperature is obtained by the following steps:

[0022] S41, obtaining historical smelting data of a flash furnace reaction tower and obtaining a temperature distribution image, dividing the temperature distribution image into regions, obtaining the average temperature of the regions on the temperature distribution image of the historical smelting data, and marking the regions with hot spots or slag shedding;

[0023] S42, generate a data set based on historical smelting data: set temperature nodes a1, a2, ..., an, where n is the number of nodes; let T represent the temperature random variable, and obtain the probability P{T≥a1} of the occurrence of the labeled data when the temperature random variable T is greater than or equal to a1, P{T≥a1}=n a1 / N a1 , where n a1 Indicates the number of areas in the historical smelting data where the temperature is greater than or equal to a1 and where hot spots or slag shedding occur, N a1 Indicates the number of regions with a temperature greater than or equal to a1 in the historical smelting data; in the same way as the temperature node a1, the probability of the occurrence of the labeled data when the temperature random variable T is greater than or equal to a2, ..., an is obtained P{T≥a2}, ..., P{T≥an}; n data points are formed by a1 and P{T≥a1}, a2 and P{T≥a2}, ..., an and P{T≥an}, and the data points are recorded as x1, x2, ..., xn to obtain a data set;

[0024] S43, obtain the probability density function g(x) between the probability of occurrence of hot spot phenomenon or slag shedding phenomenon and temperature through kernel density estimation; select a kernel function with non-negativity and symmetry and whose integral in the real number domain R is 1; set a constant greater than zero as the bandwidth h of the kernel function K; scale the kernel function according to the bandwidth to obtain Kh, where Kh(u)=1 / h×K(u / h), u is the input of the kernel function; obtain the contribution rate Kh(x-xi) of the data point xi in the data set to the estimated point x; add the contribution rates of all data points in the data set to the estimated point x to obtain the kernel density estimate g(x) at the estimated point x; g(x)=1 / n∑Kh(x-xi); change the position of the estimated point x to obtain the kernel density estimate of the access feedback data on the entire data set; the data set is the interval between the minimum and maximum values ​​of the historical access feedback data; verify the effect of the kernel function, and select the bandwidth h with the best verification effect through cross-validation.

[0025] In step S13, the process of identifying the cause of abnormally high temperature further includes the following steps:

[0026] S51, obtaining the average temperature time series of the area with abnormally high temperature, assuming that the area with abnormally high temperature is the target area, recording the time when the target area is identified as the area with abnormally high temperature as t, obtaining the change rate Δt1 of the average temperature of the target area before time t-dt, the change rate Δt2 of the average temperature of the target area after time t+dt, and the change rate Δt of the average temperature of the target area between time [t-dt, t+dt]; where dt is the time window;

[0027] S52, calculate the difference between Δt1 and Δt2 to obtain the change rate difference Δbt, and perform unsupervised classification on [Δbt Δt]: obtain historical smelting data with slag shedding phenomenon, determine the slag shedding area, calculate the change rate of the average temperature of the area before and after the slag shedding, and obtain the difference v1 of the change rate of the average temperature before and after the slag shedding; obtain the change rate v2 of the average temperature of the area when the slag shedding occurs, and add [v1 v2] to the unsupervised classification data set; obtain historical smelting data with hot spot phenomenon, determine the hot spot area, calculate the change rate of the average temperature of the hot spot area before time rt-dt and after time rt+dt, and obtain the difference v3 of the temperature change rate; obtain the change rate v4 of the average temperature of the hot spot area within the time [t-dt, t+dt], and add [v3 v4] to the unsupervised classification data set; use the classified data set to perform unsupervised classification on [Δbt Δt], according to [Δbt The proportion p of the data belonging to the slag shedding phenomenon in the classification cluster to which Δt] belongs is obtained, and the probability p that the abnormally high temperature phenomenon is caused by the slag shedding is obtained. If p is not less than the set threshold, it is judged that the cause of the abnormally high temperature phenomenon is the slag shedding, otherwise the cause of the slag shedding is the hot spot phenomenon.

[0028] Before the slag falls off, if the slag skin suddenly falls off, the temperature in the furnace may suddenly rise. This is because the protective effect of the slag skin is suddenly lost, causing the furnace wall to be directly exposed to the high temperature environment. In this case, the sharp rise in temperature is an obvious sign of slag falling off, so the temperature change rate when the slag falls off is analyzed. The temperature difference inside the slag falling area before and after the slag falls off may be relatively stable, because the heat conduction effect caused by the presence or absence of the slag layer is relatively stable, but there is a large difference between the two, so the temperature change rate of the slag falling area before and after the slag falls off is analyzed.

[0029] In step S13, controlling the raw material composition change and the heat load to accelerate the slag formation also includes the following steps:

[0030] Use high iron content copper concentrate and simultaneously reduce the quartz powder flux ratio to control the increase of slag iron silicon ratio during the reaction process;

[0031] Add special material adjustment fins at the feed end of the concentrate nozzle to increase the amount of ore in the area, reduce the oxygen content per ton of ore, achieve local reduction of reaction heat load, reduce the temperature of the melt and flue gas and dust, make the amount of melt adhesion greater than the amount of erosion, and improve the slagging progress;

[0032] Adjust the relative position of the air regulating cone and the concentrate nozzle to the air chamber to reduce the ventilation cross-sectional area of ​​the area, reduce the oxygen content of the area, thereby reducing the heat load of the area, increasing the melt viscosity, and accelerating the slagging process;

[0033] Increasing the cooling water volume of the regional copper water jacket and adding external forced cooling air can enhance the cooling intensity and promote the slagging process.

[0034] Compared with the prior art, the beneficial effects of the present invention are: the active intervention adjustment is adopted, which has fast response and high efficiency, and can directly intervene in the slagging process by adjusting the ingredients, the relative position of the concentrate nozzle and the air regulating cone, the cooling water volume of the copper water jacket, etc., which can greatly shorten the slagging cycle, thereby reducing the erosion of the reaction tower body by high-intensity reaction, ensuring the safety of refractory bricks and water jackets, and thus improving the intrinsic safety of metallurgical furnaces. The slagging management of the flash furnace reaction tower can be accelerated, and can be gradually adjusted according to the daily furnace condition control, and used for the dynamic maintenance process of the daily slagging of the reaction tower. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 The present invention is a schematic diagram of the structure of an intelligent management system applied to smelting equipment. DETAILED DESCRIPTION

[0036] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions 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 are within the scope of protection of the present invention.

[0037] Example: Figure 1 As shown, the present invention provides a technical solution, an intelligent management system for smelting equipment, comprising a data acquisition module, a slag control module, a data storage module and a data analysis module; the output end of the data acquisition module is connected to the input end of the data storage module, and is used to obtain real-time infrared imaging detection of the flash furnace reaction tower; the output end of the data storage module is connected to the input end of the data analysis module, and is used to store the infrared imaging data acquired by the data acquisition module, and to mark the infrared imaging data with slag falling off and hot spot phenomena; the data analysis module is used to detect abnormal temperature in the furnace, and analyze the cause of the abnormally high temperature phenomenon; the slag control module is used to accelerate slag formation when slag falling off occurs.

[0038] The slag control module also includes a raw material adjustment unit, a load adjustment unit and a cooling adjustment unit; the raw material adjustment unit is used to adjust the raw material composition, reduce the erosion of the reaction tower slag by the melt after the reaction, and improve the density of the formed slag layer; the load adjustment unit is used to reduce the thermal load of the slag shedding and thinning area, and reduce the intensity of the reaction; the cooling adjustment unit is used to enhance the cooling intensity of the slag shedding area. The data analysis module also includes a region division unit, a data integration unit, an estimation unit and a slag shedding judgment unit; the region division unit is used to divide the temperature distribution image into regions; the data integration unit is used to integrate and generate kernel density estimation and unsupervised classification data sets; the estimation unit is used to estimate the probability density function between the probability of occurrence of hot spot phenomenon or slag shedding phenomenon and temperature; the slag shedding judgment unit is used to identify the cause of abnormally high temperature phenomenon. The slag falling judgment unit obtains the change rate Δt1 of the average temperature of the target area before time t-dt, the change rate Δt2 of the average temperature of the target area after time t+dt, and the change rate Δt of the average temperature of the target area between time [t-dt, t+dt], and identifies the cause of the abnormally high temperature phenomenon according to Δbt and Δt, where Δbt is the difference between Δt1 and Δt2, t is the time when the area is identified as an area with abnormally high temperature, and dt is the time window.

[0039] Embodiment: The present invention provides a technical solution, an intelligent management method applied to smelting equipment, comprising the following steps:

[0040] S11, perform infrared imaging detection on the flash furnace reaction tower in real time, obtain a real-time temperature distribution image in the furnace, and divide the area based on the real-time temperature distribution image in the furnace:

[0041] Obtain the temperature data of each pixel in the real-time temperature distribution image in the furnace;

[0042] Randomly select K initial centroids to form classification clusters, where K is a constant;

[0043] S100, calculating the Euclidean distance between each pixel and the centroid, and assigning the pixel to the classification cluster of the centroid with the smallest Euclidean distance; the Euclidean distance is calculated by the following formula: ; Where D represents the Euclidean distance, a, b and c are the horizontal coordinates, vertical coordinates and temperature data of the pixel point in the temperature distribution image, a0, b0 and c0 are the horizontal coordinates, vertical coordinates and temperature data of the centroid in the temperature distribution image, k1 is the position weight, and k2 is the temperature weight;

[0044] S200 calculates the average value of all pixels in each classification cluster and takes the average value as the information centroid;

[0045] Steps S100 and S200 are repeated until the maximum number of iterations is reached.

[0046] S12, according to the real-time temperature conditions of the area, determine the area with abnormally high temperature:

[0047] Obtain historical smelting data of the flash furnace reaction tower and obtain a temperature distribution image, divide the temperature distribution image into regions, obtain the average temperature of the region on the temperature distribution image of the historical smelting data, and mark the region with hot spots or slag shedding;

[0048] Generate a data set based on historical smelting data: set temperature nodes a1, a2, …, an, where n is the number of nodes; let T represent the temperature random variable, and obtain the probability of the labeled data appearing when the temperature random variable T is greater than or equal to a1, P{T≥a1}, P{T≥a1}=n a1 / N a1 , where n a1 Indicates the number of areas in the historical smelting data where the temperature is greater than or equal to a1 and where hot spots or slag shedding occur, N a1 Indicates the number of regions with a temperature greater than or equal to a1 in the historical smelting data; in the same way as the temperature node a1, the probability of the occurrence of the labeled data when the temperature random variable T is greater than or equal to a2, ..., an is obtained P{T≥a2}, ..., P{T≥an}; n data points are formed by a1 and P{T≥a1}, a2 and P{T≥a2}, ..., an and P{T≥an}, and the data points are recorded as x1, x2, ..., xn to obtain a data set;

[0049] Obtain the probability density function g(x) between the probability of occurrence of hot spot phenomenon or slag shedding phenomenon and temperature through kernel density estimation; select a kernel function with non-negativity and symmetry and whose integral in the real number domain R is 1; set a constant greater than zero as the bandwidth h of the kernel function K; scale the kernel function according to the bandwidth to obtain Kh, where Kh(u)=1 / h×K(u / h), and u is the input of the kernel function; obtain the contribution rate Kh(x-xi) of the data point xi in the data set to the estimated point x; add the contribution rates of all data points in the data set to the estimated point x to obtain the kernel density estimate g(x) at the estimated point x; g(x)=1 / n∑Kh(x-xi); change the position of the estimated point x to obtain the kernel density estimate of the access feedback data on the entire data set; the data set is the interval between the minimum and maximum values ​​of the historical access feedback data; verify the effect of the kernel function, and select the bandwidth h with the best verification effect through cross-validation;

[0050] The real-time temperature of all areas in the furnace is obtained, and the real-time temperature of the area is used as input to the probability density function g(x) between the probability of occurrence of hot spot phenomenon or slag shedding phenomenon and temperature to obtain the probability of occurrence of hot spot phenomenon or slag shedding phenomenon; if the probability of occurrence of hot spot phenomenon or slag shedding phenomenon is not less than the threshold, the area is judged to be an area with abnormally high temperature.

[0051] S13, performing hot spot analysis and slag shedding analysis on the area with abnormally high temperature to identify the cause of the abnormally high temperature phenomenon; if the cause of the abnormally high temperature phenomenon is slag shedding, controlling the change of raw material composition and the heat load at the same time to accelerate the formation of slag; if the cause of the abnormally high temperature phenomenon is not slag shedding, continuing to monitor the temperature in the furnace;

[0052] The method of identifying the cause of abnormally high temperature phenomenon further comprises the following steps:

[0053] Get the average temperature time series of the area with abnormally high temperature, let the area with abnormally high temperature be the target area, record the time when the target area is identified as the area with abnormally high temperature as t, get the change rate Δt1 of the average temperature of the target area before time t-dt, the change rate Δt2 of the average temperature of the target area after time t+dt, and the change rate Δt of the average temperature of the target area between time [t-dt, t+dt]; where dt is the time window;

[0054] Calculate the difference between Δt1 and Δt2 to get the change rate difference Δbt, and perform unsupervised classification on [Δbt Δt]: obtain historical smelting data with slag shedding phenomenon, determine the slag shedding area, calculate the change rate of the average temperature of the area before and after slag shedding, and obtain the difference v1 of the change rate of the average temperature before and after slag shedding; obtain the change rate v2 of the average temperature of the area when slag shedding, and add [v1 v2] to the unsupervised classification data set; obtain historical smelting data with hot spot phenomenon, determine the hot spot area, calculate the change rate of the average temperature of the hot spot area before time rt-dt and after time rt+dt, and obtain the difference v3 of the temperature change rate; obtain the change rate v4 of the average temperature of the hot spot area within the time [t-dt,t+dt], and add [v3 v4] to the unsupervised classification data set; use the classified data set to perform unsupervised classification on [Δbt Δt], according to [Δbt The proportion p of the data belonging to the slag shedding phenomenon in the classification cluster to which Δt] belongs is obtained, and the probability p that the abnormally high temperature phenomenon is caused by the slag shedding is obtained. If p is not less than the set threshold, it is judged that the cause of the abnormally high temperature phenomenon is the slag shedding, otherwise the cause of the slag shedding is the hot spot phenomenon.

[0055] Controlling the raw material composition changes while controlling the heat load to accelerate slag formation also includes the following steps:

[0056] By adding special material adjustment fins at the feed end of the concentrate nozzle, the regional ore discharge volume is increased and the oxygen content per ton of ore is reduced, thereby achieving a local reduction in the reaction heat load, reducing the temperature of the melt and flue gas and dust, making the melt adhesion amount greater than the erosion amount, and improving the slagging progress;

[0057] Adjust the relative position of the air regulating cone and the concentrate nozzle to the air chamber to reduce the ventilation cross-sectional area of ​​the area, thereby reducing the oxygen content in the area, reducing the regional heat load, increasing the melt viscosity, and accelerating the slagging process;

[0058] As the slag falls off and becomes thinner, the high-temperature melt gas in the area directly contacts the refractory bricks and the copper water jacket, causing damage to the equipment. Therefore, it is necessary to immediately increase the cooling intensity in this area to form a melt protection layer on the surface to facilitate the adhesion of slag. The cooling intensity is increased by increasing the cooling water volume of the copper water jacket in this area and adding forced cooling air externally to effectively promote the slag process.

[0059] The above method takes about 12-24 hours to implement, which greatly shortens the natural drossing process. The degree of dross recovery can be confirmed through infrared imaging detection. When there is no obvious deviation between the data and the surrounding area, it can be confirmed that the drossing is completed and a record is formed.

[0060] It will be apparent to those skilled in the art that the invention is not limited to the details of the exemplary embodiments described above and that the invention can be implemented in other specific forms without departing from the spirit or essential features of the invention. Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description, and it is intended that all variations falling within the meaning and scope of the equivalent elements of the claims be included in the invention. Any reference numeral in a claim should not be considered as limiting the claim to which it relates.

Claims

1. An intelligent management method for smelting equipment, characterized in that: The following steps are involved: S11, performing infrared imaging detection on the flash furnace reaction tower in real time, obtaining a real-time temperature distribution image in the furnace, and performing area division based on the real-time temperature distribution image in the furnace, including the following steps: S21, obtaining temperature data of each pixel in the real-time temperature distribution image in the furnace; S22, randomly select K initial centroids to form classification clusters, K is a constant; S23, calculating the Euclidean distance between each pixel and the centroid, and assigning the pixel to the classification cluster of the centroid with the smallest Euclidean distance; the Euclidean distance is calculated by the following formula: ; Where D represents the Euclidean distance, a, b and c are the horizontal coordinates, vertical coordinates and temperature data of the pixel point in the temperature distribution image, a0, b0 and c0 are the horizontal coordinates, vertical coordinates and temperature data of the centroid in the temperature distribution image, k1 is the position weight, and k2 is the temperature weight; S24, calculating the average value of all pixels of each classification cluster, and taking the average value as the information centroid; S25, repeating steps S23 and S24 until the maximum number of iterations is reached; S12, determining an area with abnormally high temperature according to the real-time temperature conditions of the area; S13, performing hot spot analysis and slag shedding analysis on the area with abnormally high temperature to identify the cause of the abnormally high temperature phenomenon; if the cause of the abnormally high temperature phenomenon is slag shedding, controlling the change of raw material composition and the heat load at the same time to accelerate the formation of slag; if the cause of the abnormally high temperature phenomenon is not slag shedding, continuing to monitor the temperature in the furnace; The method of identifying the cause of abnormally high temperature phenomenon further comprises the following steps: S51, obtaining the average temperature time series of the area with abnormally high temperature, assuming that the area with abnormally high temperature is the target area, recording the time when the target area is identified as the area with abnormally high temperature as t, obtaining the change rate Δt1 of the average temperature of the target area before time t-dt, the change rate Δt2 of the average temperature of the target area after time t+dt, and the change rate Δt of the average temperature of the target area between time [t-dt, t+dt]; where dt is the time window; S52, calculate the difference between Δt1 and Δt2 to obtain the change rate difference Δbt, and perform unsupervised classification on [Δbt Δt]: obtain historical smelting data with slag shedding phenomenon, determine the slag shedding area, calculate the change rate of the average temperature of the area before and after the slag shedding, and obtain the difference v1 of the change rate of the average temperature before and after the slag shedding; obtain the change rate v2 of the average temperature of the area when the slag shedding occurs, and add [v1 v2] to the unsupervised classification data set; obtain historical smelting data with hot spot phenomenon, determine the hot spot area, calculate the change rate of the average temperature of the hot spot area before time rt-dt and after time rt+dt, and obtain the difference v3 of the temperature change rate; obtain the change rate v4 of the average temperature of the hot spot area within the time [t-dt, t+dt], and add [v3 v4] to the unsupervised classification data set; use the classified data set to perform unsupervised classification on [Δbt Δt], according to [Δbt The proportion p of the data belonging to the slag shedding phenomenon in the classification cluster to which Δt] belongs is obtained, and the probability p that the abnormally high temperature phenomenon is caused by the slag shedding is obtained. If p is not less than the set threshold, it is judged that the cause of the abnormally high temperature phenomenon is the slag shedding, otherwise the cause of the slag shedding is the hot spot phenomenon.

2. The intelligent management method for smelting equipment according to claim 1, characterized in that: In step S12, determining the area with abnormally high temperature according to the real-time temperature distribution of the area also includes the following steps: The real-time temperature of all areas in the furnace is obtained, and the real-time temperature of the area is used as input to the probability density function g(x) between the probability of occurrence of hot spot phenomenon or slag shedding phenomenon and temperature to obtain the probability of occurrence of hot spot phenomenon or slag shedding phenomenon; if the probability of occurrence of hot spot phenomenon or slag shedding phenomenon is not less than the threshold, the area is judged to be an area with abnormally high temperature.

3. The intelligent management method for smelting equipment according to claim 2 is characterized in that: The probability density function g(x) between the probability of occurrence of hot spot phenomenon or slag shedding phenomenon and temperature is obtained by the following steps: S41, obtaining historical smelting data of a flash furnace reaction tower and obtaining a temperature distribution image, dividing the temperature distribution image into regions, obtaining the average temperature of the regions on the temperature distribution image of the historical smelting data, and marking the regions with hot spots or slag shedding; S42, generating a data set according to historical smelting data: setting temperature nodes a1, a2, ..., an, where n is the number of nodes; Let T represent the temperature random variable, and obtain the probability P{T≥a1} of the labeled data appearing when the temperature random variable T is greater than or equal to a1, P{T≥a1}=n a1 / N a1 , where n a1 Indicates the number of areas in the historical smelting data where the temperature is greater than or equal to a1 and where hot spots or slag shedding occur, N a1 Indicates the number of regions with a temperature greater than or equal to a1 in the historical smelting data; in the same way as the temperature node a1, the probability of the occurrence of the labeled data when the temperature random variable T is greater than or equal to a2, ..., an is obtained P{T≥a2}, ..., P{T≥an}; n data points are formed by a1 and P{T≥a1}, a2 and P{T≥a2}, ..., an and P{T≥an}, and the data points are recorded as x1, x2, ..., xn to obtain a data set; S43, obtain the probability density function g(x) between the probability of occurrence of hot spot phenomenon or slag shedding phenomenon and temperature through kernel density estimation; select a kernel function with non-negativity and symmetry and whose integral in the real number domain R is 1; set a constant greater than zero as the bandwidth h of the kernel function K; scale the kernel function according to the bandwidth to obtain Kh, where Kh(u)=1 / h×K(u / h), u is the input of the kernel function; obtain the contribution rate Kh(x-xi) of the data point xi in the data set to the estimated point x; add the contribution rates of all data points in the data set to the estimated point x to obtain the kernel density estimate g(x) at the estimated point x; g(x)=1 / n∑Kh(x-xi); change the position of the estimated point x to obtain the kernel density estimate of the access feedback data on the entire data set; the data set is the interval between the minimum and maximum values ​​of the historical access feedback data; verify the effect of the kernel function, and select the bandwidth h with the best verification effect through cross-validation.

4. The intelligent management method for smelting equipment according to claim 1 is characterized in that: In step S13, controlling the raw material composition change and the heat load to accelerate the slag formation also includes the following steps: Use high iron content copper concentrate and simultaneously reduce the quartz powder flux ratio to control the increase of slag iron silicon ratio during the reaction process; Add special material adjustment fins at the feed end of the concentrate nozzle to increase the amount of ore in the area, reduce the oxygen content per ton of ore, achieve local reduction of reaction heat load, reduce the temperature of the melt and flue gas and dust, make the amount of melt adhesion greater than the amount of erosion, and improve the slagging progress; Adjust the relative position of the air regulating cone and the concentrate nozzle to the air chamber to reduce the ventilation cross-sectional area of ​​the area, reduce the oxygen content of the area, thereby reducing the heat load of the area, increasing the melt viscosity, and accelerating the slagging process; Increasing the cooling water volume of the regional copper water jacket and adding external forced cooling air can enhance the cooling intensity and promote the slagging process.

5. An intelligent management system for smelting equipment, using an intelligent management method for smelting equipment as claimed in any one of claims 1 to 4, characterized in that: It includes a data acquisition module, a slag control module, a data storage module and a data analysis module; the output end of the data acquisition module is connected to the input end of the data storage module, and is used to obtain real-time infrared imaging detection of the flash furnace reaction tower; the output end of the data storage module is connected to the input end of the data analysis module, and is used to store the infrared imaging data obtained by the data acquisition module, and to mark the infrared imaging data with slag shedding and hot spot phenomena; the data analysis module is used to detect abnormal temperatures in the furnace and analyze the causes of abnormally high temperatures; the slag control module is used to accelerate slag formation when slag shedding occurs.

6. The intelligent management system for smelting equipment according to claim 5 is characterized in that: The slag control module also includes a raw material adjustment unit, a load adjustment unit and a cooling adjustment unit; the raw material adjustment unit is used to adjust the raw material composition, reduce the erosion of the reaction tower slag by the melt after the reaction, and improve the density of the formed slag layer; the load adjustment unit is used to reduce the thermal load in the slag falling and thinning area, and reduce the intensity of the reaction; the cooling adjustment unit is used to enhance the cooling intensity of the slag falling area.

7. The intelligent management system for smelting equipment according to claim 5 is characterized in that: The data analysis module also includes a region division unit, a data integration unit, an estimation unit and a slag shedding judgment unit; the region division unit is used to divide the temperature distribution image into regions; the data integration unit is used to integrate and generate data sets for kernel density estimation and unsupervised classification; the estimation unit is used to estimate the probability density function between the probability of occurrence of hot spot phenomena or slag shedding phenomena and temperature; the slag shedding judgment unit is used to identify the cause of abnormally high temperature phenomena.

8. The intelligent management system for smelting equipment according to claim 7 is characterized in that: The slag falling judgment unit obtains the change rate Δt1 of the average temperature of the target area before time t-dt, the change rate Δt2 of the average temperature of the target area after time t+dt, and the change rate Δt of the average temperature of the target area between time [t-dt, t+dt], and identifies the cause of the abnormally high temperature phenomenon according to Δbt and Δt, where Δbt is the difference between Δt1 and Δt2, t is the time when the area is identified as an area with abnormally high temperature, and dt is the time window.

Images