고등기술연구원

논문현황

Institute for Advanced Engineering

논문현황

INSTITUTE FOR ADVANCED ENGINEERING

그린시스템/인텔리전스센터 Object Detection for P&ID Images

· 저자

차재민,이태경,Shelly Salim, 류보현


Piping and instrumentation diagram(P&ID) is an essential design document which is continuously modified and managed from design phase to O&M phase. For the ease of data transfer, P&IDs are generally converted into PDF which is hard to be modified. Therefore, engineering companies who need to manage P&IDs should manually re-convert their P&ID images into the P&IDs in CAD formats. To reduce the inefficiency of the P&ID re-conversion, various symbols and texts in P&ID images should be automatically detected before the re-conversion. As the first step to the auto P&ID conversion, in this study, we propose methods detecting symbols and texts in P&ID image using geometrical and deep learning-based approaches.


그린시스템/인텔리전스센터 Features Recognition from Piping and Instrumentation Diagrams in Image Format Using a Deep Learning Network

· 저자

유은섭,차재민,이태경,김진일,문두환


A piping and instrumentation diagram (P&ID) is a key drawing widely used in the energy industry. In a digital P&ID, all included objects are classified and made amenable to computerized data management. However, despite being widespread, a large number of P&IDs in the image format still in use throughout the process (plant design, procurement, construction, and commissioning) are hampered by difficulties associated with contractual relationships and software systems. In this study, we propose a method that uses deep learning techniques to recognize and extract important information from the objects in the image-format P&IDs. We define the training data structure required for developing a deep learning model for the P&ID recognition. The proposed method consists of preprocessing and recognition stages. In the preprocessing stage, diagram alignment, outer border removal, and title box removal are performed. In the recognition stage, symbols, characters, lines, and tables are detected. The objects for recognition are symbols, characters, lines, and tables in P&ID drawings. A new deep learning model for symbol detection is defined using AlexNet. We also employ the connectionist text proposal network (CTPN) for character detection, and traditional image processing techniques for P&ID line and table detection. In the experiments where two test P&IDs were recognized according to the proposed method, recognition accuracies for symbol, characters, and lines were found to be 91.6%, 83.1%, and 90.6% on average, respectively.


그린시스템/인텔리전스센터 Sensor Drift Detection in SNG Plant using Auto-Associative Kernel Regression

· 저자

차재민,이태경,김준영,신중욱,김진일,염충섭


With the rapid development of ICTs, condition monitoring has been used as a key technology in the plant industries. For reliable condition monitoring, sensors should output same values under same conditions regardless of time, but the sensitivity of sensors is gradually changed due to several factors such as temperature, humidity, contamination, aging, and etc. This type of situation is called as sensor drift problem. To solve this, several methods such as auto-associative neural network, auto-associative support vector regression, and etc. have been developed to detect sensor drifts earlier by estimating new input based on historical data. This study applied the auto-associative kernel regression model into a synthetic natural gas plant which produces synthetic natural gas from coals to detect sensor drifts during operation phase. To validate the auto-associative kernel regression model in the synthetic natural gas plant, a real data collected from an experimental operation are used. Based on the experimental results, the auto-associative kernel regression model can rapidly detect the sensor drift in the synthetic natural gas plant.


그린시스템/인텔리전스센터 Steel Plate Faults Diagnosis with S-MTS

· 저자

김준영,차재민,신중욱,염충섭,


Steel plate faults is one of important factors to affect the quality and price of the steel plates. So far many steelmakers generally have used visual inspection method that could be based on an inspector's intuition or experience. Specifically, the inspector checks the steel plate faults by looking the surface of the steel plates. However, the accuracy of this method is critically low that it can cause errors above 30% in judgment. Therefore, accurate steel plate faults diagnosis system has been continuously required in the industry. In order to meet the needs, this study proposed a new steel plate faults diagnosis system using Simultaneous MTS (S-MTS), which is an advanced Mahalanobis Taguchi System (MTS) algorithm, to classify various surface defects of the steel plates. MTS has generally been used to solve binary classification problems in various fields, but MTS was not used for multiclass classification due to its low accuracy. The reason is that only one mahalanobis space is established in the MTS. In contrast, S-MTS is suitable for multi-class classification. That is, S-MTS establishes individual mahalanobis space for each class. 'Simultaneous' implies comparing mahalanobis distances at the same time. The proposed steel plate faults diagnosis system was developed in four main stages. In the first stage, after various reference groups and related variables are defined, data of the steel plate faults is collected and used to establish the individual mahalanobis space per the reference groups and construct the full measurement scale. In the second stage, the mahalanobis distances of test groups is calculated based on the established mahalanobis spaces of the reference groups. Then, appropriateness of the spaces is verified by examining the separability of the mahalanobis diatances. In the third stage, orthogonal arrays and Signal-to-Noise (SN) ratio of dynamic type are applied for variable optimization. Also, Overall SN ratio gain is derived from…


그린시스템/인텔리전스센터 Plant Modeling Based on SysML Domain Specific Language

· 저자

이태경,차재민,김준영,신정욱,김진일, 염충섭


Successful implementation of Model-based Systems Engineering(MBSE) obviously needs a model supporting efficient communication among engineers of various domains. The system modeling language standard, SysML is designed to create MBSE supporting models. However, SysML itself is not practical enough to be used for real-world engineering projects. As SysML is designed for generic systems and requires specialized knowledge, a model written in SysML has a limited capability to support communication between a systems engineer and a subsystem engineer. Our research's main goal is to develop a SysML based plant model integrating most outputs from plant design phases. As mentioned, a standard SysML based plant model is not specific enough to be understood by plant engineers. To make the SysML model more practical, a customized SysML for the plant engineering domain is required. Unfortunately, current researches on SysML Domain Specific Language(DSL) for the plant engineering industry are still on the early stage. So, as a pilot, we have developed our own SysML-based Piping &Instrumentation Diagram (P&ID) creation environment and P&ID itself for a specific plant system, via widely known SysML modeling tool called MagicDraw. P&ID is one of the most important output during the plant design phase, which contains all information for the plant construction phase. So a SysML based P&ID has a great potential to bridge gaps between plant engineers.


그린시스템/인텔리전스센터 Smart Chemical Plant Architecture Development Based on a Systems Engineering

· 저자

김준영


Chemical plant is a factory producing chemical goods through a continuous chemical reactions flow process in a set of equipment and apparatus. In this factory, equipment is the most crucial components in the whole life cycle of the plant encompassing planning, design, engineering, construction, commissioning, operation &maintenance (O&M) and disposal. According to the industrial statistics, cost of equipment takes over 52% of the total construction cost, and maintenance fee due to equipment failure during operation is reported over 10M,andaveragedailylossof150K in some plants. The main reason for this is due to lack of data and information flow during the planning and operation stage. In other words, transparent and live information for the whole stakeholders of the chemical plant is utmost important for the chemical plant industry. In this thesis, a way of achieving such a purpose is investigated based on the technology called IoT (Internet of Things) and Smart Factory. Specifically, architecture of smart plant for chemical plant was developed based on the Systems Engineering (SE), a powerful methodology for successfully developing the complex system, involving multi-disciplinary domain and a variety of stakeholders. The developed architecture was illustrated and verified by developing a prototype for a natural gas based hydrogen generation plant. Through the prototype, the impact of the cost saving as well as information infrastructure of the developed architecture is significant. Even so, it should be verified through more industrial cases; i.e., tailoring process in the future. Considering that there is almost no work in the open literature dealing with the smart factory for continuous process plant, such as the chemical plant dealt with in the thesis, the research results will contribute in this area.


그린시스템/인텔리전스센터 정상 운전조건 중 반잠수식 플랫폼의 평형 자세에 따른 부유식 해상풍력 시스템의 하중 완화 연구

· 저자

김정태, 김관수, 고혁준, 김현종, 박종포


Rotor thrust causes the tilt of floating offshore wind systems. Due to wind-induced inclination, not only does the power performance of the wind turbine decrease, but also the load of the floating offshore wind system increases. In this study, to investigate the effects of load mitigation on a floating offshore wind system, the inclination of the floating offshore wind system is straightened by adjusting the amount of ballast water in the floater. The coupled aero-hydro-servo-elastic simulation code FAST is used to perform the load analysis. Then the short-term fatigue damage is calculated using MCrunch based on rainflow counting and Miner’s rule. As a result of dynamic analysis, it was found that the power performance is improved and the bending moment of the tower base is mitigated by ballasting. However, in the case of fatigue load, it was shown that load mitigation is insignificant from the results of DEL.


그린시스템/인텔리전스센터 Resonance Avoidance Control Algorithm for Semi-submersible Floating Offshore Wind Turbine

· 저자

김관수 김현종 김현규 손재훈 김정태 박종포


In this study, a resonance avoidance control algorithm was designed to address the tower resonance problem of a semi-submersible floating offshore wind turbine (FOWT) and the dynamic performance of the wind turbine, floater platform, and mooring line at various exclusion zone ranges were evaluated. Simulations were performed using Bladed, a commercial software for wind turbine analysis. The length of simulation for the analysis of the dynamic response of the six degrees of freedom (DoF) motion of the floater platform under a specific load case was 3600 s. The simulation results are presented in terms of the time domain, frequency domain, and using statistical analysis. As a result of applying the resonance avoidance control algorithm, when the exclusion zone range was ± 0.5 rpm from the resonance rpm, the overall performance of wind turbine was negatively affected and when the range was sufficiently wide at ± 1 rpm, the mean power was reduced by 0.05%, and the damage equivalent load of the tower base side-side bending moment was reduced by approximately 14%. The tower resonance problem of the FOWT caused by practical limitations in design and cost issues can be resolved by changing the torque control algorithm.


그린시스템/인텔리전스센터 반잠수식 풍력터빈 공진 회피 제어 알고리즘 설계

· 저자

김관수, 김정태, 김현종, 박종포


In this paper, a resonance avoidance control algorithm was designed to suppress the resonance phenomenon acting on the tower of a semi-submersible wind power system. Floating wind systems have some parts that need to be redesigned due to interface problems between the floater and turbine. In particular, the resonance of the tower is a serious issue, and a tower redesign process is generally conducted. Since the target wind power system has operating point in which resonance easily occurs due to the frequency of wind, it is difficult to satisfy design life requirements. Therefore, in this study, DNV GL Bladed wind turbine design software was used to analyze the performance of the control algorithm in wind conditions that frequently produce resonance. The results showed that the maximum load and fatigue load of the bending moment in the side-side direction of the tower base were reduced by 21.84 % and 48.05 %, respectively.