English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools annual

Global: Zendy Free Plan

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Fuel particle size and distribution is the key to energy saving of blast furnace. The fuel particle size will have influence to the highest temperature of the sintered material layer. The paper describe a method of adhesive particle image segmentation using based on particle swarm optimization fuzzy C –means clustering. It can measure particle size of fuel using OPENCV image processing real-time. Some experimental results are given,which show that the algorithm has the effective ability of searching global optimal solution. It has get very good effect on particle statistics. Introduction The fuel is the foundation of high temperature furnaces stable operation, so the need of original fuel quality is high[1]. In order to guarantee the stability of the original fuel quality, strengthening digital management is essential. The important operation of furnaces temperature is to ensure the ventilation of fuel columns in the furnaces. In addition, one of the important factor which can affect the ventilation is the particle size and the distributed composition of the fuel those are sent into the furnaces. When the size of the fuel is too big, the decomposition is slow and consumption rate is high, the fuel will burn out after arriving the combustion zone. When the size of is too small, the ventilation in the furnace will be affected and form over fire. Blast furnace iron making is based on concentrate, the particle size of fuel should be uniform. The size of fuel between 5-15mm should be less than 30% of the total, the size smaller than 5mm should be less than 5% of the total. The ingredient and the quality must be stable, the harmful substances should be few[2]. The main methods for particle size detection currently contains mechanical mesh sieve method, laser diffraction, electrical induction method etc.[3]. With the application of digital image processing technology in the recognition particle size area, particle measuring technology own the advantages of wide using range, accuracy measurement and high precision etc.. Although there are lots of methods for particle size detection nowadays, most of them exists the problems of long period, complicated steps, and low accuracy rate. Particle size detection is the important quality standard, a good measure method and the accuracy measure result play an important role in the fuel using and the processing applications[4]. With the rapidly development of the computer, CPU processing speed is greatly improved, CCD technology is becoming stronger. These factors make the quality of digital images higher, the cost of equipment on the image decrease, and make the image technology using into mineral process possible[5]. As a result, the image processing technology can be applied into varied kinds of processing and manufacturing industry. According the features of the fuel particle image, the text analyze the performance of particle and the applicability of using in the particle image. In the text, the methods include points fuel particle swarm fuzzy Caverage clustering algorithm for image segmentation, colleting the perimeter, area, particle size, shape index and shape feature parameters. The study applies Open CV library and 32 bites C/C++ translator-editor system. Under the International Conference on Automation, Mechanical Control and Computational Engineering (AMCCE 2015) © 2015. The authors Published by Atlantis Press 655 VC++6.0 circumstance, the automatic segmentation, edge tracking, testing and statistical analysis of the particle have been realized. In the paper, statistics of computing and distribution of particle size real time have also been completed, and according the statistics, correct and suitable method for quantitative analysis of particle size has been offered[6]. Particle size analyze system The analyze system of particle size consists of hardware and software two groups. The original fuel size on-line automatic detection device includes camera and protection devices, external light source, hood, industrial computers, and fiber optic cable as major components[7], as the figure 1 shown. It mainly accomplish the work of particle image acquisition, storage and analysis. Fig.1 Automatic detection device of fuel particle on-line Industrial cameras used a fixed focal length of high-speed industrial camera, the installed camera was perpendicular to shoot material, as the production environment of blast furnaces was filled with dust, the protector for camera was essential, in addition, the purge with compressed air could be used in order to keep the camera lens clear. As a result, the image of the material on the conveyor could be shot clear, and the images would be sent to the industry computer. And then the computer computed the size of the original fuel out, the computer got a picture every 10s, the result was saved in the memory. Particle detection algorithm principle The algorithm principle is shown in figure 2. The camera which was used in the experiment was camera DS-2CD4032FWD-SDI. It is 300 million pixels CMOS ICR Day & Night Gun Digital Camera, with H.264 High Profile code, the highest resolution can be up to 300 million pixels @20fps and exports 1080p@30fps real time image. The method for dealing with the images was median filtering method. The median filtering is one of non-linear signal processing techniques which can effectively suppress noise and it is based on sort statistical theory. The basic principle of median filtering is to replace a point in the digital image or digital sequence with the median of the each point in the neighborhood. The aim is to make the pixel value be closest to the real value and remove the single noise point. The method adopted dimensional slide template of a certain structure, and pixel in the template would be sorted by size according the pixel value. Dimensional data sequences those were monotonically rising or falling came into being. They were usually 3*3, 5*5 area or different shapes. The study adopter the 3*3 square templates. When the median filtering eliminates noise, the mainly two factors which are related but entirely separated the median filtering dependents on are the spatial extent of the neighborhood and number of pixels involved. Figure 3 is the result of comparing before and after treatment with median filtering. After the treatment, the edge of collected particle is clear and inner grayscale difference becomes smaller.

Analysis of fuel particle size based on particle swarm fuzzy C- average clustering