Results 1 - 5 of 5
Results 1 - 5 of 5. Search took: 0.016 seconds
|Sort by: date | relevance|
[en] Artificial neural networks (ANNs) have been applied to various fields due to their fault and noise tolerance and generalization characteristics. As an application to nuclear engineering, we apply neural networks to the early recognition of nuclear power plant operational transients. If a transient or accident occurs, the network will advise the plant operators in a timely manner. More importantly, we investigate the ability of the network to provide a measure of the confidence level in its diagnosis. In this research an ANN is trained to diagnose the status of the San Onofre Nuclear Generation Station using data obtained from the plant's training simulator. Stacked generalization is then applied to predict the error in the ANN diagnosis. The data used consisted of 10 scenarios that include typical design basis accidents as well as less severe transients. The results show that the trained network is capable of diagnosing all 10 instabilities as well as providing a measure of the level of confidence in its diagnoses
[en] Resolving the uncertainties associated with solutions obtained from artificial neural networks (ANNs) is a major concern for ANN researchers. Error bounds on the solutions are important because they are in integral part of verification and validation. In this research, stacked generalization (SG) is applied to provide error bounds for novel solutions obtained from ANNs. This work shows that SG can provide error bounds on ANN results. We have applied SG to nuclear power plant fault detection for verification of diagnoses provided by ANNs
[en] Eruption mass and mass flow rate are critical parameters for determining the aerial extent and hazard of volcanic emissions. Infrasound waveform inversion is a promising technique to quantify volcanic emissions. Although topography may substantially alter the infrasound waveform as it propagates, advances in wave propagation modeling and station coverage permit robust inversion of infrasound data from volcanic explosions. The inversion can estimate eruption mass flow rate and total eruption mass if the flow density is known. However, infrasound-based eruption flow rates and mass estimates have yet to be validated against independent measurements, and numerical modeling has only recently been applied to the inversion technique. Furthermore we present a robust full-waveform acoustic inversion method, and use it to calculate eruption flow rates and masses from 49 explosions from Sakurajima Volcano, Japan.
[en] Monochromatic infrasound waves are scarcely reported volcanic infrasound signals. These waves have the potential to provide constraints on the conduit geometry of a volcano. However, to further investigate the waves scientifically, such as how the conduit shape modulates the waveforms, we still need to examine many more examples. In this paper, we provide the most detailed descriptions of these monochromatic infrasound waves observed at Aso volcano in Japan. At Aso volcano, a 160-day-long magmatic eruption occurred in 2014–2015 after a 20-year quiescent period. This eruption was the first event that we could monitor well using our infrasound network deployed around the crater. Throughout the entire eruption period, when both ash venting and Strombolian explosions occurred, monochromatic infrasound waves were observed nearly every day. Although the peak frequency of the signals (0.4–0.7 Hz) changed over time, the frequency exhibited no reasonable correlation with the eruption style. The source location of the signals estimated by considering topographic effects and atmospheric conditions was highly stable at the active vent. Based on the findings, we speculate that these signals were related to the resonant frequencies of an open space in the conduit: the uppermost part inside the vent. Based on finite-difference time-domain modeling using 3-D topographic data of the crater during the eruption (March 2015), we calculated the propagation of infrasound waves from the conduit. Assuming that the shape of the conduit was a simple pipe, the peak frequency of the observed waveforms was well reproduced by the calculation. The length of the pipe markedly defined the peak frequency. By replicating the observed waveform, we concluded that the gas exhalation with a gas velocity of 18 m/s occurred at 120 m depth in the conduit. However, further analysis from a different perspective, such as an analysis of the time difference between the arrivals of infrasound and seismic waves, is required to more accurately determine the conduit parameters based on observational data. .