Jo Feb 24, 2022
The research group led by Ma Min Gon, vice dean of the Faculty of Information Science and Technology, has developed a Korean braille character reader which converts raised letters read by an image input device into a character code readable by ordinary people.
When books published in raised type for the blind to read by groping are read by an image input device like a scanner, the raised parts are shown in dim dots.
In order to extract these dim dots more precisely, they proposed and realized several new basic methods: binarization by wavelet difference image, tilt estimation by Hough conversion, dot extraction by dot database and Maharanobis distance for precise extraction without being affected by the scanning direction, charline extraction from dots capable of strongly overcoming noises, substitution of braille code for unicode by bidirection scanning, postprocess based on geometrical position of dots, etc.
Thus, they were able to raise the extraction precision up to 96% and the recognition speed up to 1 second per page.
...
Jo Feb 22, 2022
A research group led by Ri Yong Min, a researcher at the Faculty of Materials Science and Technology, has developed a screw rolling mill with individually driven rollers and established a rolling process which can produce bars of more than 20mm in diameter through one pass without a roller change from round steel of 50mm in diameter, consuming less power. They will no longer need an ordinary light section rolling mill in machine factories and repair bases that have a great demand for bars in various diameters.
In this rolling process, round steel of 50mm in diameter is heated to 850~950℃ in a heating furnace and is rolled into bars of 20mm in diameter through one pass in a cantilevered 3-helical rolling mill with individually driven rollers that has a large feed angle and an elongation angle before they are cut into necessary lengths.
Thus, they can produce bars of different diameters from short scrap iron by changing no rolls but adjusting the roll gap, and ensure the metal yield of more than 99% and the roundness of ±0.1.
...
Jo Feb 21, 2022
Kim Won Ok, a researcher at the Faculty of Information Science and Technology, has developed a C/C++ programming evaluation system to raise practical programming skills of students who are learning C/C++ programming including those who are receiving distance education, and to make an accurate evaluation of the programming assignments performed by them.
What she has done includes the combination of web and other applications by TCP socket communication, the implementation of a C/C++ programming environment by C++ compiler and I/O stream pipeline, accurate evaluation of assignment performance, and the security of learning-related data including assignments by user certification, copy paste protection and encryption.
She herself tried making and running a program so as to make a programming tool and a system for evaluating programming accuracy. This helped raise students’ programming skills and enthusiasm for study, and evaluate their assignment performance rapidly and accurately.
...
Jo Feb 19, 2022
Pak Il Chol, a researcher at Faculty of Information Science and Technology, has developed a Java API compound retrieval program on the basis of the collection of code snippets and feature extraction. In this program it is possible to realize ranking of code snippets suitable for user intention from their application degree and brevity degree.
He increased the rate of indexing by 0.3GB per hour and the precision and recall by 6% and 9% respectively, which resulted in the improvement of the development speed and quality of programs. Thus, he ensured the guarantee for reducing the labour and time needed for program development to one tenth.
...
Jo Feb 17, 2022
Ri Sok Chol, a section head at the Faculty of Metal Engineering employed genetic programming (GP) to derive a CaO-SiO2-Al2O3 slag viscosity equation and presented an algorithm for it.
Slag viscosity is a key property that determines the stability and productivity in the metallurgical furnace operation, and it is of great importance in understanding the fluid dynamic of molten slags and slag-metal reaction kinetics during the pyrometallurgy process.
Up to now, there have been many experiments to measure, and models to estimate slag viscosity. Viscosity models can be classified into explicit models, structural models, and miscellaneous models according to the way in which they are connected to the description of slag structure.
A slag viscosity equation described with only slag composition and temperature was successfully derived by genetic programming (GP).
The viscosities of CaO-SiO2-Al2O3 system slag can be simply and expediently calculated in a wide range of compositions and temperatures by using the derived viscosity equation.
The average relative error between the values obtained by the equation and the experimental data used for its derivation is as low as 17.1%. To evaluate the accuracy of the viscosity equation of CaO-SiO2-Al2O3 system slag, experimental data and results calculated by many researchers and not used for the derivation of the viscosity equation were compared, and a reasonably good agreement was demonstrated between the results calculated by viscosity equation and other researchers’ experimental and calculated data. The calculation of viscosity showed higher accuracy than previous models.
His essay “Derivation of CaO-SiO2-Al2O3 System Slag Viscosity Equation by GP” was carried in SCI Journal “Journal of Mining and Metallurgy, Section B: Metallurgy”.
...
Jo Jan 29, 2022
These days study on a learning method for getting a dynamic characteristic model without the knowledge of an object in the field of a hierarchical supervisory control system has been actively being conducted. However, no method of applying it to existing production processes by speeding up Q-learning convergence has been proposed yet.
Kim Song Ho, a section head at Faculty of Automation Engineering, proposed a two-layer intelligent supervisory control system to adjust the online optimum setting point of the control system and set forth a new method to improve Q-learning convergence.
Unlike previous methods, the core of his suggestion is that it helps overcome trial and error Q-learning process, the weakest point of this learning, and achieve the fastest Q-learning convergence by setting optimally the way of automatically extracting experience rules of process operation from historical operation data and the initial phase of Q-learning.
His suggestion is estimated to have great practical significance in making large-scale continuous industrial processes unmanned and intelligent as it is supposed to minimize the effect of environmental changes and to guarantee product quality and stability of process operation.
He presented his essay titled “On-line set-point optimization for Intelligent Supervisory Control and Improvement of Q-learning Convergence” to SCI journals “Control Engineering Practice”.
...
© 2021 Kim Chaek University of Technology