Search NASA⌕ Search

SEARCH · Search NASA

Results for “Event Detection”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 649 records · Page 36

Assessment of SRS ambient air monitoring network

Three methodologies have been used to assess the effectiveness of the existing ambient air monitoring system in place at the Savannah River Site in Aiken, SC. Effectiveness was measured using two metrics that have been utilized in previous quantification of air-monitoring network performance; frequency of detection (a measurement of how frequently a minimum number of samplers within the network detect an event), and network intensity (a measurement of how consistent each sampler within the network is at detecting events). In addition to determining the effectiveness of the current system, the objective of performing this assessment was to determine what, if any, changes could make the system more effective. Methodologies included 1) the Waite method of determining sampler distribution, 2) the CAP88- PC annual dose model, and 3) a puff/plume transport model used to predict air concentrations at sampler locations. Data collected from air samplers at SRS in 2015 compared with predicted data resulting from the methodologies determined that the frequency of detection for the current system is 79.2% with sampler efficiencies ranging from 5% to 45%, and a mean network intensity of 21.5%. One of the air monitoring stations had an efficiency of less than 10%, and detected releases during just one sampling period of the entire year, adding little to the overall network intensity. By moving or removing this sampler, the mean network intensity increased to about 23%. Further work in increasing the network intensity and simulating accident scenarios to further test the ambient air system at SRS is planned

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗

A comparison of earthquake backprojection imaging methods for dense local arrays

Backprojection imaging has recently become a practical method for local earthquake detection and location due to the deployment of densely sampled, continuously recorded, local seismograph arrays. While backprojection sometimes utilizes the full seismic waveform, the waveforms are often pre-processed and simplified to overcome imaging challenges. Real data issues include aliased station spacing, inadequate array aperture, inaccurate velocity model, low signal-to-noise ratio, large noise bursts and varying waveform polarity. We compare the performance of backprojection with four previously used data pre-processing methods: raw waveform, envelope, short-term averaging/long-term averaging and kurtosis. Our primary goal is to detect and locate events smaller than noise by stacking prior to detection to improve the signal-to-noise ratio. The objective is to identify an optimized strategy for automated imaging that is robust in the presence of real-data issues, has the lowest signal-to-noise thresholds for detection and for location, has the best spatial resolution of the source images, preserves magnitude, and considers computational cost. Imaging method performance is assessed using a real aftershock data set recorded by the dense AIDA array following the 2011 Virginia earthquake. Our comparisons show that raw-waveform backprojection provides the best spatial resolution, preserves magnitude and boosts signal to detect events smaller than noise, but is most sensitive to velocity error, polarity error and noise bursts. On the other hand, the other methods avoid polarity error and reduce sensitivity to velocity error, but sacrifice spatial resolution and cannot effectively reduce noise by stacking. Of these, only kurtosis is insensitive to large noise bursts while being as efficient as the raw-waveform method to lower the detection threshold; however, it does not preserve the magnitude information. For automatic detection and location of events in a large data set, we therefore recommend backprojecting kurtosis waveforms, followed by a second pass on the detected events using noise-filtered raw waveforms to achieve the best of all criteria.

58 GEOSCIENCES↗

Real-Time Detection and Localization of Line Trip Event Via Relative Phase Angles

Line trip events widely exist in power systems. They can result in power outages and a huge economic loss if not promptly detected and localized. To provide a fast and precise solution, this paper presents a Complete Coverage of Voltage Measurement (CCVM)-based line trip event detection algorithm and a Relative Phase Angle (RPA)-based line trip event localization algorithm. First, frequency and relative phase angle features during a line trip event are calculated. Then, the CCVM-based algorithm is proposed from both frequency and rate of change of frequency estimation algorithm aspects. Additionally, the RPA-based algorithm is presented, and two cases are studied to demonstrate the uniqueness of the proposed algorithm. Various experiments are conducted, where the simulation results demonstrate that the proposed CCVM-based algorithm can detect a line trip event as short as 2.07 ms. In addition, the RPA-based algorithm has 1.26 times higher localization accuracy compared with the frequency magnitude-based and phase angle-based algorithms. In conclusion, the experiment results on the examples from two interconnected power systems in the U.S. verified the performance of the proposed algorithms in wide-area power systems.

24 POWER TRANSMISSION AND DISTRIBUTION↗

A Single-Feature Machine Learning Method for Detecting Multiple Types of Events from PMU Data

This paper describes simple and efficient machine learning (ML) methods for efficiently detecting multiple types of power system events captured by PMUs scarcely placed in a large power grid. It uses a single feature from each PMU based on a rectangle area enclosing the event in a given data window. This single feature is sufficient to enable commonly used ML models to detect different types of events quickly and accurately. The feature is used by five ML models on four different data-window sizes. The results indicated a tradeoff between the execution speed and detection accuracy in variety of data-window size choices. The proposed method is insensitive to most data quality issues typical for data from field PMUs, and thus it does not require major data cleansing efforts prior to feature extraction.

Dokic, Tatjana↗

Automated detection of photovoltaic cleaning events: A performance comparison of techniques as applied to a broad set of labeled photovoltaic data sets

Extracting accurate soiling loss information from photovoltaic (PV) production data first requires segmenting the time series data per natural or manually occurring cleaning events. Maintenance logs are often incomplete, rain data are often unavailable, and the debate on rain thresholds for cleaning and dew or wind cleanings is still ongoing. The present work aims to overtake these issues by improving automated methods to detect these cleaning events and therefore improve extraction of soiling loss information. Time series power production data from 22 PV inverters were labeled for natural or manually occurring cleaning events. The data sets were carefully selected to include varying degrees of soiling, cleaning events, and noise. Several algorithms, including filtering logic and change point detection, were examined for efficacy at detecting the labeled cleanings. All the methods introduced except for changepoint detection showed significant improvement at detecting the labeled cleaning events per the mean F 1 score. Furthermore, the highest performing cleaning detection algorithm achieved an absolute increase in the mean F 1 score of 43% over the default version of the RdTools stochastic rate and recovery (SRR) algorithm. The highest performing algorithm included irradiance filtering and a cleaning detection threshold, adjusted based on the 40-day centered rolling median of the absolute day-to-day deviations in the daily performance index (PI). Furthermore, these improvements are promising as cleaning detection is an essential step in the automated analysis of PV soiling.

14 SOLAR ENERGY↗

ON THE DETECTABILITY OF A PREDICTED MESOLENSING EVENT ASSOCIATED WITH THE HIGH PROPER MOTION STAR VB 10

Extrapolation of the astrometric motion of the nearby low-mass star VB 10 indicates that sometime in late 2011 December or during the first 2-3 months of 2012, the star will make a close approach to a background point source. Based on astrometric uncertainties, we estimate a 1 in 2 chance that the distance of closest approach {rho}{sub min} will be less than 100 mas, a 1 in 5 chance that {rho}{sub min} < 50 mas, and a 1 in 10 chance that {rho}{sub min} < 20 mas. The last would result in a microlensing event with a 6% magnification in the light from the background source and an astrometric shift of 3.3 mas. The lensing signal will however be significantly diluted by the light from VB 10, which is 1.5 mag brighter than the background source in B band, 5 mag brighter in I band, and 10 mag brighter in K band, making the event undetectable in all but the bluer optical bands. However, we show that if VB 10 happens to harbor a {approx}1 M{sub J} planet on a moderately wide ( Almost-Equal-To 0.18 AU-0.84 AU) orbit, there is a chance (1% to more than 10%, depending on the distance of closest approach and orbital period and inclination) that a passage of the planet closer to the background source will result in a secondary event of higher magnification. The detection of secondary events could be made possible with a several-times-per-night multi-site monitoring campaign.

79 ASTRONOMY AND ASTROPHYSICS↗

Evaluation of an event-driven 3FI ASIC for spectroscopic X-ray detection with synchrotron radiation

The novel design and evaluation on the NSLS-II beamline of the 3FI application-specific integrated circuit (ASIC) bump-bonded to a simple, planar, 2D segmented silicon sensor are presented. The ASIC was developed for full-field fluorescence spectral X-ray imaging (3FI). It is a small-scale prototype that features a square array of 32 × 32 pixels, and the size of the pixels is 100 µm × 100 µm. The ASIC was implemented in a 65 nm CMOS integrated circuit fabrication process. Each pixel incorporates a charge-sensitive amplifier, a shaping filter, a discriminator, a peak detector and a sample-and-hold circuit, allowing detection of events and storage of signal amplitudes. The system operates in a frameless event-driven readout mode, outputting analog values for threshold-triggered events, allowing high-speed multi-element X-ray fluorescence data acquisition. The 3FI ASIC achieves per-channel spectrometric performance at a power consumption of only 200 µW per pixel, with nearly all dissipation confined to the analog front-end. An energy resolution is measured at the level of 308 eV full width at half-maximum (FWHM) at 8.04 keV (Cu Kα), and 138 eV FWHM at 3.69 keV (Ca Kα). This per-pixel capability makes the prototype suitable for in situ trace element microanalysis in biological and environmental studies. Moreover, the frameless architecture of the detector is designed to address limitations of conventional X-ray fluorescence microscopy, which typically requires mechanical scanning, by enabling continuous high-throughput data acquisition in future full-field implementations.

47 OTHER INSTRUMENTATION↗

A validation of satellite derived cyanobacteria detections with state reported events and recreation advisories across U.S. lakes

Cyanobacteria harmful algal blooms (cyanoHABs) negatively affect ecological, human, and animal health. Traditional methods of validating satellite algorithms with data from water samples are often inhibited by the expense of quantifying cyanobacteria indicators in the field and the lack of public data. However, state recreation advisories and other recorded events of cyanoHAB occurrence reported by local authorities can serve as an independent and publicly available dataset for validation. State recreation advisories were defined as a period delimited by a start and end date where a warning was issued due to detections of cyanoHABs over a state’s risk threshold. State reported events were defined as any event that was documented with a single date related to cyanoHABs. This study examined the presence-absence agreement between 160 state reported cyanoHAB advisories and 1,343 events and cyanobacteria biomass estimated by a satellite algorithm called the Cyanobacteria Index (CI(cyano)). The true positive rate of agreement with state recreation advisories was 69% and 60% with state reported events. CI(cyano) detected a reduction or absence in cyanobacteria after 76% of the recreation advisories ended. CI(cyano) was used to quantify the magnitude, spatial extent, and temporal frequency of cyanoHABs; each of these three metrics were greater (r > 0.2) during state recreation advisories compared to non-advisory times with effect sizes ranging from small to large. This is the first study to quantitatively evaluate satellite algorithm performance for detecting cyanoHABs with state reported events and advisories and supports informed management decisions with satellite technologies that complement traditional field observations.

remote sensing↗

Mining for Emerging Technologies within Text Streams and Documents

Text streams, collections of documents or messages that are generated and observed over time, are ubiquitous. Our research and development is targeted at developing algorithms to find and characterize changes in topic within text streams. To date, this research has demonstrated the ability to detect and describe 1) short duration, atypical events and 2) the emergence of longer term shifts in topical content. This technology has been applied to pre-defined temporally ordered document collections but is also suitable for application to near real-time textual data streams. The underlying event and emergence detection algorithms have been interfaced to an event detection software user interface named SURPRISE. This software provides an interactive graphical user interface and tools for manipulating and correlating the terms and scores identified by the algorithms. Additionally, SURPRISE has been interfaced with the IN-SPIRE text analytics tool to enable an analyst to evaluate the surprising or emerging terms via a visualization of the entire document collection. IN-SPIRE assists in the exploration of related topics, events and views currently based on single term events. The focus of this research is to contribute to detecting, and preventing, strategic surprise.

99 GENERAL AND MISCELLANEOUS↗

Machine Learning for Synchrophasor Analysis

The report presents results from the development of a cloud-based, Big Data analysis framework for power systems. The computational pipeline uses the Apache Spark framework running in an OpenStack cloud infrastructure. A real-world phasor measurement unit (PMU) dataset has been used to carry out the analysis. Several Machine Learning (ML) methods have been developed and implemented for event and anomaly detection and classification. Actual examples of power system events detection and analysis using synchrophasor data are presented. It has been shown that applications of the cloud-based computing environment and the Apache Spark framework enable a significant increase in the computational efficiency of large-scale PMU data analysis.

20 FOSSIL-FUELED POWER PLANTS↗

Prompt fission neutron spectra from fission induced by 1 to 8 MeV neutrons on {sup 235}U and {sup 239}Pu using the double time-of-flight technique

Prompt fission neutron spectra from {sup 235}U and {sup 239}Pu were measured for incident neutron energies from 1 to 200 MeV at the Weapons Neutron Research facility (WNR) of the Los Alamos Neutron Science Center, and the experimental data were analyzed with the Los Alamos model for the incident neutron energies of 1-8 MeV. A CEA multiple-foil fission chamber containing deposits of 100 mg {sup 235}U and 90 mg {sup 239}Pu detected fission events. Outgoing neutrons were detected by the Fast Neutron-Induced {gamma}-Ray Observer array of 20 liquid organic scintillators. A double time-of-flight technique was used to deduce the neutron incident energies from the spallation target and the outgoing energies from the fission chamber. These data were used for testing the Los Alamos model, and the total kinetic energy parameters were optimized to obtain a best fit to the data. The prompt fission neutron spectra were also compared with the Evaluated Nuclear Data File (ENDF/B-VII.0). We calculate average energies from both experimental and calculated fission neutron spectra.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

Process Anomaly Detection for Sparsely Labeled Events in Nuclear Power Plants

An essential aspect of online monitoring, subtle anomaly detection increases the detection lead time for equipment failure and enables a nuclear power plant (NPP) to mitigate unexpected partial or full outages, resulting in significant cost saving to the plant. Once an anomaly is detected by plant staff, its cause and severity are investigated. Because the vast majority of anomalies require some level of investigation, including some that require time-consuming examination, before they are passed over to the engineering organization for further analysis, plants are often equipped with tools to assist the staff in performing anomaly detection. Those tools operate as a black box and are often based on statistical methods that establish sensor correlations using preconfigured mathematical models and flag correlation deviations as anomalies. Due to the number of anomalies detected at a given NPP on a daily basis, a significant number of flagged anomalies usually await examination for days or weeks. A primary cause of this backlog is that the methods used by the tools generate many false positives. Though this is usually attributed to oversensitive model settings due to very narrow normal operation bands, it can also be associated with the model development being inadequate for the process being monitored, or with missing model inputs that could have explained misclassified positives. The performance of anomaly detection tools impacts their plant acceptance and utilization, especially when the effort to address false positives generated by the tool depletes the value or cost saved by using that tool. Thus, means to advance anomaly detection performance have been investigated by the Department of Energy’s Light Water Reactor Sustainability program. Previous and ongoing efforts have targeted unsupervised machine-learning (ML) methods, which do not require the labeling of any data fed into the ML model. By contrast, in supervised anomaly detection methods, every data point is labeled as either a normal or abnormal process condition, and the model is trained to replicate the classification process. Supervised methods usually outperform unsupervised methods, due to the added value in differentiating normal from anomalous states of the monitored process. An NPP’s corrective action program requires it to track and document, via a dedicated report, the resolution of any issues that occur within the plant. Once created, each report is reviewed by a plant screening committee, and several classifications and decisions are made. Recently, a collaborating NPP developed an artificial intelligence and ML-based classifier to categorize a condition report (CR) into classes that can serve to label the data as normal or anomalous. Applying CRs as labels represents a semi-supervised use case. Semi-supervised ML assumes that labels exist for some data points (i.e., labeled anomalies, in this case) but not for the rest. In this effort, semi-supervised ML methods were used to fuse data from CRs with anomaly detection methods in order to test the hypothesis that partially labeled anomalies would improve the accuracy of the anomaly detection methods. Specifically, two methods were used. The first is the deep Semi-supervised Anomaly Detection (deep SAD) method, which can handle labels ranging from fully unsupervised to fully supervised cases. The second is a newly designed ML method developed specifically for this effort and referred to as the high-order feature (HOF)-based method. To evaluate these two methods in controlled environments, synthetic data generators were developed and used. The first datasets used a spring-mass-damper (SMD) system simulator commonly found in mechanical engineering references. This was used to create two use cases: a one- and a three-mass system. Anomalies were introduced by changing the spring and damper coefficients while the system was actuated by random forces. The second datasets used the commercial Dymola-Modelica software to build a simplified nuclear reactor model. Anomalies were added in the form of corrupted sensor readings and/or control commands. The deep SAD method was tested using the SMD system, while the HOF method was tested using both datasets. Application of the deep SAD semi-supervised ML method demonstrated that labels can generate increased confidence in detecting true anomalies. This helped increase the number of true positives and decrease the number of false negatives—something that would aid in addressing the backlog of possible anomalies. Application of the HOF method demonstrated that labels can aid in down selecting from a candidate set of features to a more optimal subset in order to better differentiate between normal and anomalous conditions.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Tracing the impacts of Mount Pinatubo eruption on regional climate using spatially-varying changepoint detection

Significant events, such as volcanic eruptions, can have global and long-lasting impacts on climate. These global impacts, however, are not uniform across space and time. Understanding how the Mt. Pinatubo eruption affects global and regional climate is of great interest for predicting the impact on climate due to similar events as well as understanding the possible effect of the stratospheric aerosol injections proposed to combat climate change. While many studies illustrated the impact of the Pinatubo eruption on a global scale, studies at a fine regional scale are scarce. Here, we propose a Bayesian spatially-varying changepoint detection and estimation method to trace the impact of Mt. Pinatubo eruption on regional climate. Our approach takes into account the diffusing nature and spatial correlation of the climate changes attributed to the volcanic eruption. We illustrate our method and demonstrate its advantages over an existing changepoint detection method through simulations. Finally, we apply our method to monthly stratospheric aerosol optical depth and surface temperature data from 1985 to 1995 to detect and estimate changepoints following the 1991 Mt. Pinatubo eruption. Our results quantitatively characterize the spatial pattern of the eruption’s impact on regional climate, complementing the previous studies on the global impact of the Pinatubo eruption.

Aerosol optical depth↗

New detector for use in fast neutron radiography

We have developed and tested a new type detector for use in the fast neutron (FN) imaging radiography applications. FN radiography is generally used for nondestructive material testing, medical and biology applications, border patrol, transportation and cargo screening tasks. It is complementary to other types of radiography, providing additional information on light element content of the material samples. Distinct from other FN imagers presently known, our device implements a neutron-sensitive scintillator attached to a position-sensitive photomultiplier tube (PSPMT), and operates in an event-by-event readout mode, acquiring energy, timing, and pulse shape information for all detected radiation events. The information is used to help separate events of FN interactions in the scintillator from the background events caused by the electronics noise and by the other types of background radiation. Selection of pure fast neutron events in the final image allows us to achieve ultimate image contrast and resolution, as compared with other types of FN imaging devices operating most commonly in an integration mode, in which the detector?s dark noise and radiation background dilute the images. The detector performance for FN imaging application was tested using D-D neutron generator, designed and manufactured by Adelphi Technology, Inc. This essentially point-like neutron source operates in continuous mode producing up to 109 of 2.5 MeV neutrons per second. Samples made of metals plastic and other material were used to measure the detector resolution, efficiency and uniformity. Results of these tests are presented and discussed. Fig. 1 shows one of the test FN radiographic images obtained using the sample made of 11 styrene plastic strips. All strips are squares 4.8 x 4.8 mm2 with six different lengths 10 to 60 mm with 10 mm increment. [A] [B] [C] Fig. 1. [A]-layout of the test sample; [B]-raw FN shadow image of the sample; [C]-map of the plastic strips as they appear on the image, with measured FN absorption factor shown in percent. Both X and Y position resolution of the FN imaging detector are estimated to be less than 0.5 mm (sigma) based on the analysis of the edges of the strip shadows in the plot. Because this detector shows the fraction-of-a-millimeter resolution desirable for most of FN applications, and is built using radiation hard materials, we believe it could be a good alternative to other FN imaging systems based on CCD or solid state detectors. In addition, having a sub-nanosecond timing resolution, it is suitable for the time-of-flight energy-resolved FN imaging.

Popov, V.↗

Object detection with deep learning for rare event search in the GADGET II TPC

In the pursuit of identifying rare two-particle events within the GADGET II Time Projection Chamber (TPC), this paper presents a comprehensive approach for leveraging Convolutional Neural Networks (CNNs) and various data processing methods. To address the inherent complexities of 3D TPC track reconstructions, the data is expressed in 2D projections and 1D quantities. This approach capitalizes on the diverse data modalities of the TPC, allowing for the efficient representation of the distinct features of the 3D events, with no loss in topology uniqueness. Additionally, it leverages the computational efficiency of 2D CNNs and benefits from the extensive availability of pre-trained models. Given the scarcity of real training data for the rare events of interest, simulated events are used to train the models to detect real events. To account for potential distribution shifts when predominantly depending on simulations, significant perturbations are embedded within the simulations. This produces a broad parameter space that works to account for potential physics parameter and detector response variations and uncertainties. These parameter-varied simulations are used to train sensitive 2D CNN object detectors. When combined with 1D histogram peak detection algorithms, this multi-modal detection framework is highly adept at identifying rare, two-particle events in data taken during experiment 21072 at the Facility for Rare Isotope Beams (FRIB), demonstrating a 100% recall for events of interest. Here, we present the methods and outcomes of our investigation and discuss the potential future applications of these techniques.

Convolutional neural network↗

Deep Learning Method for Detecting Precursors to Adverse Events

With the recent advancements in Deep Learning methods, the ability to model large complex heterogeneous data sets are fundamentally changing industry and research. Coupled with hardware improvements, and ease of implementation, a wide variety of deep neural network architectures can quickly be developed to solve a sweeping range of problems such as: object detection in images, automatic healthcare diagnosis using heterogenous data sources, real time language translating and sentence prediction, upscaling low resolution images, and forecasting of multivariate timeseries. Generally, many of these architectures outperform classical machine learning approaches in their respective tasks, however, this typically comes at a cost of interpretability. These black box algorithms generally suffer from lack of transparency in both model complexity as well as the rationale behind the prediction. This lack of comprehension, is driving an emerging area of interest in “Explainable AI”. An algorithm called: “Deep Temporal Multiple Instance Learning”1 was a recently developed to identify precursors to adverse events and has been applied in the aviation domain. The deep learning architecture is designed to capture the evolution of the probability of the outcome over the time preceding the adverse event using a multiple instance learning approach as illustrated in Figure 1. Precursors are defined when the probability of the event has exceeded a threshold at some point in the timeseries, at which point, a sensitivity analysis is performed to determine contributing factors. The contributing factors are used to explain and define the precursor during the periods where the probability score is high. The identified contributing factors are then presented to subject matter experts to provide objective insights into the leading factors associated with the particular adverse event. The algorithm has been tested on flight data from a commercial airline and has the ability to discover precursors to known adverse events that take the form of safety critical operations, such as unstable approach events on final approach. Apart from detecting precursors to adverse events, the converse can also be leveraged to discover corrective actions. These positive actions manifest themselves as periods in the timeseries when the precursor score has been lowered from an elevated state; meaning that if the system had been left uncorrected, it would have eventually reached the adverse event state. Characterizing these state changes can help identify successful interventions that may not have been known before. Policy makers and procedure designers can use this additional knowledge to craft more safety and efficient resilient procedures for future operations and therefore improve the overall performance of the National Airspace.

Matthews, Bryan L.↗

Laboratory investigation of hydraulic fracturing in granitic rocks using active and passive seismic monitoring

SUMMARY Knowledge of the fracturing processes can be important for the optimization of pressurized fluid injection operations in the deep underground rock mass. Active and passive seismic monitoring techniques have been used in the field for tracking or mapping the propagating hydraulic fracture. Although both these monitoring techniques provide valuable information about the generated fracture network, it is difficult for either technique to comprehensibly identify the different processes associated with hydraulic fracturing. The combined active and passive monitoring has the potential for better characterization of the complex hydraulic fracturing phenomena. In this study, laboratory hydraulic fracturing experiments with combined active and passive seismic monitoring were conducted on true triaxially loaded Barre granite cubes with different fluid injection rates. The seismic inelastic fracturing was detected by 16 passive acoustic emission sensors, where 3678 and 2370 seismic source events were detected for the high and low injection rate experiments, respectively. For active monitoring, strong variations in the attributes of signals were observed which were transmitted through four source–receiver pairs, placed both perpendicular and parallel to the generated hydraulic fracture. Positive velocity changes were observed for active sensor pairs with ray paths passing through the generated hydraulic fracture indicating fluid permeation, whereas isolated dry deformation was characterized by a slight but permanent velocity decrease. Compared to velocity, the energy of the active signals was 1–2 orders of magnitude more sensitive to different hydraulic fracturing processes. However, the sensitivity and signatures of the active signal attributes were found to be dependent on the frequency range and direction of ray path with respect to the location of the generated fracture network. Using the coupled evaluation of the active and passive signals we were able to systematically identify various hydraulic fracturing processes including: (1) aseismic deformation, (2) fracture initiation and fluid permeation, (3) pressure build-up, (4) fracture propagation and (5) pressure release and leak-off. The results of this study showed that combining the respective advantages of active and passive seismic techniques and using both of them to monitor the failure processes can facilitate a more comprehensive understanding and better control of the hydraulic stimulations in subsurface operations.

Geochemistry & Geophysics↗

Explainable AI for Multivariate Time Series Pattern Exploration: Latent Space Visual Analytics With Temporal Fusion Transformer and Variational Autoencoders in Power Grid Event Diagnosis

Detecting and analyzing complex patterns in multivariate time-series data is crucial for decision-making in urban and environmental system operations. However, challenges arise from the high dimensionality, intricate complexity, and interconnected nature of complex patterns, which hinder the understanding of their underlying physical processes. Existing AI methods often face limitations in interpretability, computational efficiency, and scalability, reducing their applicability in real-world scenarios. This paper proposes a novel visual analytics framework that integrates two generative AI models, Temporal Fusion Transformer (TFT) and Variational Autoencoders (VAEs), to reduce complex patterns into lower-dimensional latent spaces and visualize them in 2D using dimensionality reduction techniques such as PCA, t-SNE, and UMAP with DBSCAN. These visualizations, presented through coordinated and interactive views and tailored glyphs, enable intuitive exploration of complex multivariate temporal patterns, identifying patterns’ similarities and uncover their potential correlations for a better interpretability of the AI outputs. The framework is demonstrated through a case study on power grid signal data, where it identifies multi-label grid event signatures, including faults and anomalies with diverse root causes. Additionally, novel metrics and visualizations are introduced to validate the models and assess the performance, efficiency, and consistency of latent maps generated by VAE, which have been utilized in prior studies for latent space cartography and used as a benchmark in this study, and the emerging TFT architecture under various configurations. These analyses provide actionable insights for model parameter tuning and reliability improvements. Comparative results highlight that TFT achieves shorter run times and superior scalability to diverse time-series data shapes compared to VAE. This work advances fault diagnosis in multivariate time series, fostering explainable AI to support critical system operations.

Explainable AI↗