CATNet: A Small Sample Transfer Learning Method for Key Point Detection of Blast Furnace Burden Surface

Real-time and accurate measurement of blast furnace (BF) stockline is important for charging control and BF safety. However, traditional non-contact methods for stockline detection face many challenges such as low accuracy, poor anti-interference ability, and stability in harsh BF environments. The high-temperature novel industrial endoscope (NIE), a visible light detection equipment for BF burden surface (BBS), has captured real-time and continuous video streams of BBS, which provides a brand-new means and ideas for non-contact stockline detection. Based on this, a novel non-contact stockline detection method based on real-time burden surface video streams is proposed. First, an edge detection network based on bootstrap sampling and eigenvector coding clustering (ECC) is established to extract accurate edge distribution of burden surface video streams. Then, an actual imaging region model of the NIE is developed by the latitude circle method to quantify the relationship between the actual imaging region and the BF stockline. Finally, the edge of burden surface video streams is correlated with the corresponding actual imaging region through the NIE imaging model, and the non-contact stockline detection model is derived to realize the BF stockline detection based on video streams. Experiments and industrial applications show that the proposed method is superior to the traditional non-contact stockline detection methods in terms of detection accuracy, stability, and real-time performance.

At present, by establishing complex mathematical model to simulate the ironmaking process in Blast Furnace (BF), the real-time and applicability are not strong in actual production. To solve this problem, an image reconstruction algorithm for BF Burden surface based on Compressive Sensing is proposed. Sparse matrix is constructed by Discrete Cosine Transform (DCT) and random Gaussian matrix is constructed to obtain part of the image information of BF Burden surface directly. Then the original image information is reconstructed by orthogonal matching pursuit algorithm. The final simulation results show that the image reconstruction algorithm proposed in this study works well.

Obtaining a high-precision 3-D burden surface (BS) shape of a blast furnace (BF) is of significance for improving energy efficiency in the ironmaking process, and for stabilizing the operation of the BF. However, existing methods used for detecting the BS cannot provide precise data to guide the burden distribution operation of the BF effectively, because they cannot overcome the harsh environment inside the BF. Therefore, to obtain a reliable BS shape under a high-temperature, high-pressure, high-dust, and low-light environment, a 3-D high-temperature industrial endoscope with a large depth of field and wide field of view is proposed in this article. First, image-forming optical system indices are obtained by analyzing the environmental factors and geometry of the BF. Based on these indices, an optical system with a large depth of field, wide field of view, and low-light imaging capacity is designed. Next, a 3-D depth calculation method based on a virtual and real twin binocular optical system is used to obtain 3-D depth of feature points on the burden distribution ring. Furthermore, a functional relationship model between the image defocusing degree and depth of the BS is established, based on the 3-D depth. Finally, a monocular vision 3-D reconstruction method based on the depth from defocus parameter is used to realize the 3-D reconstruction of the BS. Experimental industrial applications of the proposed device demonstrate that it can provide data support for the BF burden distribution operation and has a significant industrial application value.

At blast furnace B at Salzgitter Flachstahl a series of innovative measuring techniques are installed to monitor the processes at the blast furnace top, making this furnace one of the best equipped furnaces in Europe. These techniques comprise full 2D measurement of the temperature profile of the top gas shortly above the burden surface, 3D radar scan of the whole burden surface and online measurement of the dust concentration in the top gas. After more than 5 years’ experience with most of these techniques, they enable to better understand the complex chemical and physical interrelations occurring in the BF stack between the ascending process gas and the descending solid burden. A couple of examples of incidents that were monitored are presented in this article, including influences of charging programmes on top gas temperature profiles and influences of disturbed gas solids interaction on the BF working state. The new measuring techniques with tailor-made data processing enable the operators to gain a better picture of the processes currently occurring in the blast furnace, consequently supporting them in keeping the blast furnace operation as stable and efficient as possible.

The top gas temperature is an apparent indicator of the distribution state of gas flow, and it is an important parameter to judge the furnace condition. It is related to the top equipment in addition to the charge characteristics and bottom-up generation and distribution. Therefore, the fluid-solid heat transfer model from burden surface -chute-rising ducts is established in this paper. The finite element method is used to calculate the flow and heat transfer of the gas at the top of the blast furnace, and the influence of the chute on the top gas temperature is analyzed. The accuracy of the model is verified by the actual blast furnace top temperature. The results show that the chute has a guiding effect on the gas flow. As the chute rotates, the top gas temperature fluctuates. This research is helpful for actual blast furnace operators to understand the top temperature and make the condition judgment more accurate.

Numerical investigation of burden distribution in ironmaking blast furnace

As indicated above, we, at the Chiba Works, classify and blend various materials in a rationalized system, so as to produce the most suitable blast furnace burden. Table VIII shows an outline of the equipment and Table IX the performance result. With special reference to No. 1 blast furnace, it has been giving as high a yield as 4.5 net tons per sq ft of the hearth area, without using any special techniques such as high top pressure operation. As for No. 5 blast furnace, which has the world’s largest inner volume, it was blown-in in March this year and has been in an uprating operation with an excellent performance. During the month of July, it marked a yield of 3470 net tons per day. These achievements, we believe, are largely due to our intensified pretreatment system, under which economically obtained iron ore of all kinds of size are so treated as to be used most efficiently at the blast furnace, sintering plant, and pelletizing plant. No less important to this success is the use of bedding system which makes it possible that iron-bearing waste or residue materials generated in small lots from various stages of iron and steelmaking operations at the Chiba Works are turned into large uniformed lots for complete recovery.

Recently, the grade of iron ore deposits has deteriorated and further development of low grade deposits is desired. Presently, the most effective and often followed route taken to utilize such deposits is the provision of beneficiation plants for upgrading iron ore and pelletizing plants for agglomerating. The iron ore fines cannot be directly used in the blast furnace as they severely reduce the permeability of the blast furnace bed and bring down the efficiency of the furnace. To overcome this, agglomeration of high grade iron ore fines is done to serve as blast furnace burden. Both the beneficiation and pelletization techniques are key process in utilization of low grade iron ores. This study deals with the beneficiation and agglomeration of the low grade iron ore; up to the extent such that it becomes an ideal blast furnace feed. Effect of particle size on pellet quality during pellet making has also been aimed and studied along with the development of flowsheet for beneficiation of low grade iron ore.

An ironmaking blast furnace (BF) is a complex high-temperature moving bed reactor involving counter-, co- and cross-current flows of gas, liquid and solid, coupled with heat and mass exchange and chemical reactions. Two-dimensional (2D) models were widely used for understanding its internal state in the past. In this paper, a three-dimensional (3D) CFX-based mathematical model is developed for describing the internal state of a BF in terms of multiphase flow and the related thermochemical behavior, as well as process indicators. This model considers the intense interactions between gas, solid and liquid phases, and also their competition for the space. The model is applied to a BF covering from the burden surface at the top to the liquid surface in the hearth, where the raceway cavity is considered explicitly. The results show that the key in-furnace phenomena such as flow/temperature patterns and component distributions of solid, gas and liquid phases can be described and characterized in different regions inside the BF, including the gas and liquids flow circumferentially over the 3D raceway surface. The in-furnace distributions of key performance indicators such as reduction degree and gas utilization can also be predicted. This model offers a cost-effective tool to understand and control the complex BF flow and performance.

Radar technology is performed to obtain the position and morphology of burden surface in blast furnace, which is conducive to the optimal process of decision-making over burden distribution. Considering the imaging of the of echo spectrum, includes the characteristics of local highlighting and low signal-to-noise ratio, the adaptive gray-scale compensation algorithm combined with multi-scale Retinex image enhancement is adopted for high-quality brightness adjustment and color correction. Then, in order to make the target burden surface prominent and distinguishable, the improved hot metal coding method is applied to conduct pseudo-color processing of the enhanced image as obtained previously. In the meantime, the guided image filtering algorithm is introduced to maintain the target boundary of the burden surface while ensuring the effective removal of noise. Finally, the fractal interpolation algorithm is employed in line with the principle of SAR imaging to reconstruct the radar image of burden surface in blast furnace. Moreover, a qualitative analysis is conducted during production of the major characteristics of burden surface.

Model of burden distribution in operating blast furnaces

The present study deals with the utilization of fines generated from comminution process (crushing, grinding and screening) of the Run of Mines into value added products i.e. fluxed iron ore pellets. The study comprises to understand the physical and mechanical behavior of five distinguished chemical compositions of green and dried iron ore pellets with respect to a typical Mini Blast furnace (MBF) burden data and furnace operating parameter. The maximum basicity of pellets was calculated 2.37 to make slag neutral when blast furnace runs at 100% high ash coke (avg. ash content= 29%). The crushing strength and drop number of various green pellets were measured. Green Crushing Strength was decreased with increasing lime fines. The addition of lime fines as a burnt lime, which has acicular structure creates less plasticity and brittle like fracture occurred. Due to formation of hard CaCO3 layer on the surface, after increasing lime contain crushing strength was increased in the air and oven dry pellets with respect to acid pellet (0% lime fines addition).

Harsh environment inside blast furnace (BF) makes it hard to monitoring burden surface shape with heterogeneous fluidization characteristic, which is an important factor affecting the production efficiency and safety of the iron-making process. In this paper, based on the imaging principle of remote sensing Synthetic Aperture Radar (SAR) radar, we design the industrial scanning radar, multiplying the density of the sample points of the surface. By analyzing the characteristics of spectrum of BF radial echo signal, this paper proposes a new intelligent image algorithm that improves the Kmeans segmentation rectified by the minimum error Bayes decision (MEBD) to extract a strip of surface point cloud. Energy centrobaric correction method is used to estimate the distance frequency of the blast surface and the results of it are used to reconstruct 3D burden surface model. Compared with other typical methods with measured data, the presented method is verified to be more effective and robust.

Mining of expert rules for blast furnace air volume regulation

A system that calculates the particle size and fluid area from a blast furnace burden surface image is developed. Off-line test results of the calculation algorithms show that a spatial frequency calculation using the power spectrum of the image is sufficient for the particle size measurement. In addition, a spatial differential of maximum intensity of the sequential image is suitable for fluid area measurement. The system is installed and operating.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>