In-Koo Hwang

Development of advanced I&C in nuclear power plants: ADIOS and ASICS

In this paper Automatic Startup Intelligent Control System (ASICS) that automatically controls the PWR plant from cold shutdown to 5% of reactor power and Alarm and Diagnosis-Integrated Operator Support System (ADIOS) that is integrated with alarms, process values, and diagnostic information to an expert system focused on alarm processing are described. Nuclear Power Plant is manually controlled from cold shutdown to 5% according to the general operation procedures for startup operation of nuclear power plant. Alarm information is the primary sources to detect abnormalities in nuclear power plants or other process plants. The conventional hardwired alarm systems, characterized by one sensor-one indicator may lead the control room operators to be confused with avalanching alarms during plant transients. ASICS and ADIOS are designed to reduce the operator burden. The advances in computer software and hardware technology and also in information processing provide a good opportunity to improve the control systems and the annunciator systems of nuclear power plants or other similar process plants. It is very important to test and evaluate the performance and the function of the computer- or software-based systems like ASICS and ADIOS. The performance and the function of ASICS and ADIOS are evaluated with the real-time functional test facility and the results have shown that the developed systems are efficient and useful for operation and operator support.

A Study on Loose Part Monitoring System in Nuclear Power Plant Based on Neural Network

The Loose Part Monitoring System(LPMS) has been designed to detect. locate and evaluate detached or loosened parts and foreign objects in the reactor coolant system. In this paper, at first, we presents an application of the back propagation neural network. At the preprocessing step, the moving window average filter is adopted to reject the reject the low frequency background noise components. And then, extracting the acoustic signature such as Starting point of impact signal. Rising time. Half period. and Global time, they are used as the inputs to neural network . Secondly, we applied the neural network algorithm to LPMS in order to estimate the mass of loose parts. We trained the impact test data of YGN3 using the backpropagation method. The input parameter for training is Rising clime. Half Period amplitude. The result shored that the neural network would be applied to LPMS. Also, applying the neural network to thin practical false alarm data during startup and impact test signal at nuclear power plant, the false alarms are reduced effectively.

Development of integrated automatic diagnosis method for loose parts monitoring system

Abstract Generally, it is known that loose parts in the reactor coolant systems (RCS) bring serious damage into the system components and impede the normal function of the system. So, it is necessary to rapidly respond when the impact event has occurred. This paper presents a realization of automatic diagnosis algorithm for LPMS (Loose Parts Monitoring System) and application results to the impact test data at YongGwang Nuclear Power Plant Unit 3 (YGN3), Kori Nuclear Power Plant Unit 4 (KNU4) and the real data at YongGwang Nuclear Power Plant Unit 1 (YGN1). The integrated diagnosis algorithm is composed of three parts; prefiltering, impact location and mass estimation. The prefiltering is needed to reject low frequency background noises. To estimate the impact location, the starting points of impact are detected from the filtered signals and compared to produce the time differences, and then the modified triangular method is applied. To estimate the mass and energy of a loose part, we automatically compute the maximum amplitude and the initial half period. Additionally, a modified impact theory considering amplitude and energy attenuation effects is applied. To show the effectiveness of the proposed diagnostic method, the real impact test data at YGN3, KNU4 and the real impact data at YGN 1 is used. The analysis results show that the location estimation error is on average below 7.5%, and the average mass estimation is within 40%.

Practical Alarm Suppression Rules and their Implementation for Nuclear Power Plants

It is necessary to adopt some logical techniques and methods of alarm processing for a large complex plant such as nuclear power plants in order to present the occurred alarm messages properly and concisely. Among such alarm processing techniques, the alarm suppressing function is a strong tool to avoid alarm flooding during the sudden transients of plant output power such as turbine trips, reactor trips and other incidents. Unless any suppression or representation technologies are used in an alarm message listing system, it cannot provide quick assistance to plant operators or supervisors during plant upsets because too many alarm messages are presented in an alarm list window. This paper presents the key suppression methods and analysis processes developed for implementing a suppressed alarm message listing function of an integrated alarm system called LogACTs which has been applied to a CANDU nuclear power plant. A simulation testing of the suppressing function conducted with the real plant alarm message list data has demonstrated an effective performance of the developed logics with the high suppression rate.

A continuous solution of the heat equation based on a fuzzy system

A continuous solution of the Dirichlet boundary value problem for the heat equation ＝ using a fuzzy system is described. We first apply the Crank-Nicolson method to obtain a discrete solution at the grid points for the heat equation. Then we find a continuous function to represent approximately the discrete values at the grid points in the form of a bicubic spline function (equation omitted) that can in turn be represented exactly by a fuzzy system. We show that the computed values at non-grid points using the bicubic spline function is much smaller than the ones obtained by linear interpolations of the values at the grid points. We also show that the fuzzy rule table in the fuzzy system representation of the bicubic spline function can be viewed as a gray scale image. Hence, the fuzzy rules provide a visual representation of the functions of two variables where the contours of different levels for the function are shown in different gray scale levels

Expert System-Based Implementation of Failure Detection

In light of the need to improve MMIS of NPPs, the advanced I&C research team of KAERI has embarked on developing an Alarm and Diagnosis-Integrated Operator Support System, called ADIOS. ADIOS was composed of two modules such as alarm process module and fault diagnostic module. The alarm process module can offer key alarms to operator using dynamic alarm filtering methods and the fault diagnostic module was able to diagnose the component failure caused by a sensor fault and a hardware fault. ADIOS has been built in an object-oriented AI environment of G2 expert system software tool. In the G2 environment, every alarm and diagnostic component have been treated as an object of class. The attributes of each alarm object include activation status, alarm message, process value, activated time, priority, acknowledgment state, interlocked equipment and caused alarm. If an alarm is activated, its icon color on an overview process mimic diagram changes corresponding to its priority, which can be set initially or determined dynamically by reasoning rules and procedures. The process conditions, such as plant or equipment status and correlated alarms’ states have determined the priority of the activated alarm and the fault diagnostic condition. The knowledge base and the inference engines of ADIOS have been constructed by process analysis of the plant and discussion with operators and nuclear plant experts.