Dongil Shin

A web-based, interactive virtual laboratory system for unit operations and process systems engineering education: issues, design and implementation

The development of real educational content and customized virtual education systems satisfying the needs of a specific engineering education domain is getting more and more research attention in this era of ubiquitous Web and virtual technologies. By analyzing the characteristics of computer-based educational methods and adopting the rapidly changing Internet and object component technologies, we have developed a Web-based, interactive virtual laboratory system for unit operations and process systems engineering education, and validated its effectiveness by surveying student users of the implemented prototype system. In this paper, we also discuss many technical issues in building Web-based interactive virtual education systems and share the experience obtained. The proposed system is expected to overcome many obstacles in performing the unit operations laboratory in the current setting. More teachware and experiment suites are being added for more extensive, quantitative evaluation of educational efficiency of the proposed methodology and the system.

Web-based interactive virtual laboratory system for unit operations and process systems engineering education

Abstract The development of the real educational contents and customized virtual education satisfying the needs of specific engineering education domain is getting more and more research attention in this area of Web and virtual technologies. By analyzing the characteristics of computer-based educational methods and using the rapidly changing Internet and object technologies, we developed a Web-based, interactive virtual laboratory system of unit operations and process systems engineering education. The proposed system is expected to overcome many obstacles in performing the unit operations laboratory at the current setting. More teachwares and experiments suites are being added for the extensive, quantitative evaluation on educational efficiency of the proposed system, and the result will be reported at the conference.

Cooperative problem solving in diagnostic agents for chemical processes

Abstract In this paper we introduce an agent-based framework for the diagnosis of chemical processes, based on spatially distributed diagnosis architecture. Although the suggested multi-agent diagnostic system uses only the information about process topology and control structure, making the most of local information produces global diagnostic solution. A function-behavior model is also suggested to make possible the collaborative problem solving in chemical process fault diagnosis. To demonstrate the benefits of the suggested collaborative problem solving approach, we solve the diagnostic problem of a stirred-tank reactor system with recycle.

A study on the evaluation of structural controllability in chemical processes

Abstract Controllability should be considered at the design stage before control system is designed, as it is an inherent property of process. Structural information makes controllability assessment possible by giving insights into the pathways of disturbances in the process. In this study, a simple procedure to evaluate controllability using relative order analysis and structural decomposition is suggested to select design alternatives. The effectiveness of the proposed method is validated by comparing the results with the analyses that were performed by using dynamical simulation.

Risk Analysis Using Automatically Synthesized Robust Accident Scenarios and Consequence Assessment for Chemical Processes: Process Partition and Consequence Analysis Approach

Consequence analysis and risk assessment are very important in chemical process industries because of the potential risk of hazardous materials. In this paper, we introduce a new system for consequence analysis and risk management (CARM) and propose a new strategy for producing robust accident scenarios in quantitative risk assessment. The suggested synthesis method analyzes process elements and selects and generates robust accident scenarios that simulate the most possible worst-case accident that should be foremost considered. The scenario-reasoning scheme consists of three types of knowledge base (equipment property, material property, and process unit knowledge) and four reasoning algorithms (macro decomposition, equipment screening, equipment behavior analysis, and accident scenarios reasoning). The synthesized result of the analysis enhances the reliability of the generated accident scenario and prevents the risks from being overestimated. The obtained result, as easily confirmed by using CARM, should be more helpful in proper process design and emergency planning.

Function-behavior modeling and multi-agent approach for fault diagnosis of chemical processes

Publisher Summary This chapter discusses a function–behavior modeling and distributed multi-agent system approach for the diagnosis of chemical processes. In the proposed system, diagnostic agents (DA) of individual units communicate by exchanging messages and try to solve the global fault diagnosis problem using collaborative diagnosis method. The benefits of the suggested collaborative problem solving approach are demonstrated by solving the diagnostic problem of a continuous stirred-tank reactor system with recycle and a first order irreversible reaction. The suggested approach was implemented on the expert system development tool, and a prototype application built showed its effectiveness. Function–behavior model and activities and role of DA are explained using a simple process. This system consists of three process units where input is fed into a tank unit through a pipe unit and output comes out through another pipe unit. DA infers causal relationship through communication with neighboring DAs and fundamental self reasoning. The detecting module of DA is always active, and DAs perform diagnosis by sending and receiving messages when events leading to faults occur.

A study on the improvement of controllability of chemical processes based on structural analysis

Abstract Controllability should be considered at the design stage before control system is designed, as it is an inherent property of process. Structural information makes controllability assessment possible by giving insights on the pathways of disturbances in the process. In this study, a simple procedure to evaluate controllability is suggested to select design alternatives. This procedure is useful in screening out the design alternatives before using detailed controllability evaluation methods such as dynamic simulation.

Qualitative interpretation and compression of process data using clustering method

Abstract This paper presents a new qualitative data interpretation and data compression method, which is based on modified adaptive k-means clustering algorithm. Conventional qualitative data interpretation methods that are based on control charts, such as Shewart, CUSUM, and EWMA control charts, are focused upon detection of changes from steady state values, so they are not suitable for describing transient or dynamic behavior. But the proposed method continuously updates its detection limit, or center of cluster, so it can handle transient or dynamic behavior. Thus it can be applied to fault diagnosis of chemical process by combining with cause-effect digraph model, RCED(Reduced Cause Effect Digraph). The usefulness of the proposed data interpretation method and cause-effect digraph model are illustrated using their application to the water supply unit of a utility boiler plant. The proposed data interpretation method can be used for not only changed detection but also data compression. As the proposed method can store the value that minimizes the SSE between the retrieved data and the original data, it shows better data compression result compared to other conventional data compression methods such as Box Car, Backward Slope, Combined Box Car and Backward Slope algorithm and Swinging Door Trending.

A strategy for the generation of robust accident scenarios in quantitative risk assessment using multi-component analysis

A strategy for producing robust accident scenarios in quantitative risk assessment, which is performed in the process design or operation steps, is proposed. Most governments over the world require industrial companies to submit proper emergency plans through the off-site risk assessment. However, there have been no systematic approaches or criteria for generating reasonable virtual accident scenarios, and it is very difficult to get the unified or coherent assessment result. To get over these shortcomings, this study proposes a strategy for analyzing process elements and then selecting and generating robust accident scenarios that simulate the worst accidents most likely to happen and should be foremost considered. The result through the proposed analysis enhances the reliability of the produced risk scenario and prevents the risks from being overestimated. The obtained result should be more helpful in the proper process design and emergency planning.

Analysis of Abnormality Detection Properties of Nonlinear PCA Methods

Abstract In this work we review PCA and various non-linear PCA methods from the autoassociator point of view. Autoassociator is used to identify and remove correlations among problem variables and can be used to detect abnormality condition of various processes where an early warning of an abnormal condition is required. Feature extraction methods such as PCA and neural network can be an excellent tool of building autoassociator. Several autoassociators based on statistics and neural network have been reviewed and their autoassociative reconstruction properties and abnormality detection performances have been analyzed for a nonlinear 3-dimensionai example. Results show that principal curves & neural network, principal curves & splines, and self-supervised MLP successfully reduces dimensionaiity and produces a feature space map resembling the actual distribution of the underlying system. Also these methods can be reliable solutions for novelty detection and their characteristics are discussed.