Prior Fault Research Articles

Traditional signal pattern recognition versus artificial neural networks for nuclear plant diagnostics

This paper introduces a new tool based on a traditional noise analysis techniques for monitoring reactor components' signal condition. It also presents the performance of artificial neural networks for pattern recognition to the same set of reactor signals and provides a comparison of these two techniques. Reactor pump signals from the Experimental Breeder Reactor (EBR-II) are utilized here. Collected signals such as pump power, pump speed, and pump pressure are obtained from already installed sensors in the reactor. The signals utilized are collected signals as well as generated signals simulating the pump shaft degradation progress. From the study of time series analysis and regression modeling of these signals, a parameter related to degradation and material buildup in the shaft is identified and used in the development of a monitoring tool. The results are then used as a benchmark against which to test the performance of artificial neural networks as a tool for reactor diagnostics. Several neural networks are examined in this study, including Restricted Coulomb Energy (RCE), Cascade Correlation, and Backpropagation paradigms of artificial neural networks. RCE is selected due to its unique design and speed, Backpropagation is selected because it is widely used and well accepted in the neural network research community, and Cascade correlation is selected because it overcomes some of the problems associated with the Backpropagation paradigm. Similar study is performed using the Adaptive Resonance Theory (ART) family of neural network paradigms. The results of this study indicate that artificial neural networks are simpler techniques for pattern recognition than noise analysis techniques such as the one introduced here. Neural networks do not require prior fault related parameter identification; they generate their own rules by learning from being shown examples. On the other hand, noise analysis and regression modeling can provide very sensitive techniques for monitoring of a detected problem in a component.

Automatic feature extraction of waveform signals for in-process diagnostic performance improvement

In this paper, a new methodology is presented for developing a diagnostic system using waveform signals with limited or with no prior fault information. The key issues studied in this paper are automatic fault detection, optimal feature extraction, optimal feature subset selection, and diagnostic performance assessment. By using this methodology, a diagnostic system can be developed and its performance is continuously improved as the knowledge of process faults is automatically accumulated during production. As a real example, the tonnage signal analysis for stamping process monitoring is provided to demonstrate the implementation of this methodology.

A Functional Decomposition Method for Redundancy Identification and Test Generation

We present a new combinational circuit automatic test-pattern generation (ATPG) acceleration method called EST that detects equivalent search states, which are saved for later use. The search space is learned and characterized using E-frontiers, which are circuit cut-sets induced by the implication stack contents. The search space is reduced by matching the current search state against previously-encountered search states (possibly from prior faults), and this reduces the length of the search. A second contribution is a calculus of redundant faults, which enables EST to make many more mandatory assignments before search than is possible by prior algorithms, by effectively using its knowledge of prior faults proven to be redundant. This accelerates ATPG for subsequent faults. These methods accelerate the TOPS algorithm 33.3 times for the hard-to-test faults in the ISCAS ‘85 benchmarks, and the SOCRATES algorithm 5.6 times for the same hard-to-test faults, with little memory overhead.

Blaming the victim: Belief in control or belief in justice?

The attribution of responsibility to victims of bad fate (“blaming the victim”) is discussed under the perspective of Just World Theory (Lerner, 1980) and the Defensive Attribution Hypothesis (Walster, 1966; Shaver, 1970). Whereas Just World Theory suggests that the belief in a just world is the decisive motive of increased attributions of responsibility, the Defensive Attribution Hypothesis assumes that these attributions are motivated by the need to believe in internal locus of control. Research evidence shows both motives as conceptually linked and empirically correlated. The central question is whether belief in a just world and belief in internal control are facets of the same latent variable or empirically distinguishable constructs, and whether they contribute independently to attributions of responsibility and blame to victims of misfortune. Results of a questionnaire study assessing opinions about cancer and cancer victims are reported. There is evidence from factor analyses that the two motives are indeed distinguishable constructs. The correlation patterns and the results of multiple regression analyses show that both motives are meaningfully related to attributions of responsibility. Moreover, it is suggested that belief in a just world is not a homogeneous construct. Belief in immanent justice according to which present misfortune is seen as a consequence of prior faults and sins is differentiated from belief in ultimate justice according to which one can be sure that present misfortune will be compensated in the long run. Whereas belief in immanent justice is the most important predictor for attributions of responsibility, the suggested emotional consequences of such attributions, like belief in invulnerability or confidence in coping, can be predicted by belief in internal control and belief in ultimate justice. Finally, suggestions are made to extend Just World Theory to clarify the function of justice motives in the person's search for meaning in his or her life.

Electrovacuum Glass Quality Control System Transferring

Abstract In this paper we discuss the problems of the model based quality control of the electrovacuum glass production processes. We focuse on the problem of such systems transfer in order to intensify their implementation at the typical plants. A flexible analytical decision rules are proposed to support a prior fault detection of the typical model estimated using training data set at the new experimental conditions with the different level of addtive noise. Theoretical, as well as practical aspects of these methodology will be considered. Finally, our industrial experience with these is discussed.