Variability Model Research Articles

Improving Variabilty Analysis through Scenario-Based Incompatibility Detection

Software Product Line (SPL) developments include Variability Management (VA) as a core activity aiming at minimizing the inherent complexity in commonality and variability manipulation. Particularly, the (automated) analysis of variability models refers to the activities, methods and techniques involved in the definition, design, and instantiation of variabilities modeled during SPL development. Steps of this analysis are defined as a variability analysis process (VA process), which is focused on assisting variability model designers in avoiding anomalies and/or inconsistencies, and minimizing problems when products are implemented and derived. Previously, we have proposed an approach for analyzing variability models through a well-defined VA process (named SeVaTax). This process includes a comprehensive set of scenarios, which allows a designer to detect (and even correct in some cases) different incompatibilities. In this work, we extend SeVaTax by classifying the scenarios according to their dependencies, and by assessing the use of these scenarios. This assessment introduces two experiments to evaluate accuracy and coverage. The former addresses responses when variability models are analyzed, and the latter the completeness of our process with respect to other proposals. Findings show that a more extensive set of scenarios might improve the possibilities of current practices in variability analysis.

Difference Scheme for Differential-Difference Problems with Small Shifts Arising in Computational Model of Neuronal Variability

The solution of differential-difference equations with small shifts having layer behaviour is the subject of this study. A difference scheme is proposed to solve this equation using a non-uniform grid. With the non-uniform grid, finite - difference estimates are derived for the first and second-order derivatives. Using these approximations, the given equation is discretized. The discretized equation is solved using the tridiagonal system algorithm. Convergence of the scheme is examined. Various numerical simulations are presented to demonstrate the validity of the scheme. In contrast to other techniques, maximum errors in the solution are organized to support the method. The layer behaviour in the solutions of the examples is depicted in graphs.

Impact of Local Atmospheric Intraseasonal Variability on Mean Sea Ice State in the Arctic Ocean

AbstractThe Arctic atmosphere shows significant variability on intraseasonal time scales of 10–90 days. The intraseasonal variability in the Arctic sea ice is clearly related to that in the Arctic atmosphere. It is well known that the Arctic mean sea ice state is governed by the local mean atmospheric state. However, the response of the Arctic mean sea ice state to the local atmospheric intraseasonal variability is unclear. The Arctic atmospheric intraseasonal variability exists in both the thermodynamical and dynamical variables. Based on a sea ice–ocean coupled simulation with a quantitative sea ice budget analysis, this study finds that 1) the intraseasonal atmospheric thermodynamical variability tends to reduce sea ice melting through changing the downward heat flux on the open water area in the marginal sea ice zone, and the intraseasonal atmospheric dynamical variability tends to increase sea ice melting by a combination of modified air–ocean heat fluxes, ice–ocean heat fluxes, and sea ice deformation; 2) the intraseasonal atmospheric dynamical variability increases summertime sea ice concentration in the Beaufort Sea and the Greenland Sea but decreases summertime sea ice concentration along the Eurasian continent in the East Siberia–Laptev–Kara Seas, resulting from the joint effects of the modified air–ocean heat fluxes, ice–ocean heat fluxes, and the sea ice deformation, as well as the mean sea ice advection due to the changes of sea ice drift. The large spread in sea ice in the CMIP models may be partly attributed to the different model performances in representing the observed atmospheric intraseasonal variability. Reliable modeling of atmospheric intraseasonal variability is an essential condition in correctly projecting future sea ice evolution.

A Field Validated Model of Temporal Variability in Oyster Habitat Suitability

This work presents the development and validation of a spatially and temporally variable oyster habitat suitability model for the western Mississippi Sound, northern Gulf of Mexico. In the work, we (1) develop an oyster habitat suitability model based on existing conditions in 1 year, (2) forecast habitat suitability throughout the same location in following years, and (3) validate the model using independent data describing field counts of live and recent dead oysters in those following years. The model uses four environmental factors to determine habitat suitability, namely: maximum annual temperature, maximum annual salinity, minimum annual salinity, and minimum annual dissolved oxygen. Overall, the model does not discriminate well between good and poor habitat when the habitat suitability score is less than 0.2; however, when the habitat suitability score is greater than 0.2, there is a high confidence that are more live than recent dead oysters. The results also show that habitat suitability varies by up to 0.45 in any single location (one standard deviation; on a scale from 0 to 1). This is important for evaluating which areas will be most resilient for oyster habitat under a variety of conditions. This study presents the first validated statistical model of temporally and spatially varying oyster habitat suitability.

Designing and testing systems of systems: From variability models to test cases passing through desirability assessment

AbstractIn the early stages of a system of systems (SoS) conception, several constituent systems could be available that provide similar functionalities. An SoS design methodology should provide adequate means to model variability in order to support the opportunistic selection of the most desirable SoS configuration. We propose the VANTESS approach that (i) supports SoS modeling taking into account the variation points implied by the considered constituent systems; (ii) includes a heuristics to weight benefits and costs of potential architectural choices (called as SoS variants) for the selection of the constituent systems; and finally (iii) also helps test planning for the selected SoS variant by deriving a simulation model on which test objectives and scenarios can be devised. We illustrate an application example of VANTESS to the “educational” SoS and discuss its pros and cons within a focus group.

Predicting response time variability from task and resting-state functional connectivity in the aging brain

Aging is associated with declines in a host of cognitive functions, including attentional control, inhibitory control, episodic memory, processing speed, and executive functioning. Theoretical models attribute the age-related decline in cognitive functioning to deficits in goal maintenance and attentional inhibition. Despite these well-documented declines in executive control resources, older adults endorse fewer episodes of mind-wandering when assessed using task-embedded thought probes. Furthermore, previous work on the neural basis of mind-wandering has mostly focused on young adults with studies predominantly focusing on the activity and connectivity of a select few canonical networks. However, whole-brain functional networks associated with mind-wandering in aging have not yet been characterized. In this study, using response time variability—the trial-to-trial fluctuations in behavioral responses—as an indirect marker of mind-wandering or an “out-of-the-zone” attentional state representing suboptimal behavioral performance, we show that brain-based predictive models of response time variability can be derived from whole-brain task functional connectivity. In contrast, models derived from resting-state functional connectivity alone did not predict individual response time variability. Finally, we show that despite successful within-sample prediction of response time variability, our models did not generalize to predict response time variability in independent cohorts of older adults with resting-state connectivity. Overall, our findings provide evidence for the utility of task-based functional connectivity in predicting individual response time variability in aging. Future research is needed to derive more robust and generalizable models.

Numerical investigation and modeling of residual stress field variability impacting the machining deformations of forged part

Aluminum alloys are widely used for structural parts in the aerospace industry. Those parts are usually machined from rolled plates or forged blanks, and heat treatments are carried out to reach the necessary mechanical properties. Forming and heat treatments induce residual stresses that generate deformations during and after machining because material removal modifies their balance. One major issue in industry is the important variability of the residual stress field in forged blanks. A better knowledge of the residual stress field will enable the definition of better machining strategy to minimize the deformations. In order to understand and manage this variability, a sensibility analysis was performed to investigate the impact of their different potential sources. It relies on a coupling of 2D-FEM and beam model simulations. Then, this work introduces a model reduction technique (by POD and SVD) to build a model of the variability of a stress field and shows how it could be used on the production line.

RuVa: A Runtime Software Variability Algorithm

Context-aware and smart systems that require runtime reconfiguration to cope with changes in the environment increasingly demand variability management mechanisms that can address runtime concerns. In recent years, we have witnessed new dynamic variability solutions using dynamic software product line (DSPL) approaches. However, while few solutions proposed so far have addressed the need to add, change and remove variants dynamically, none of them provide a way to check the constraints between features at runtime. Because all SAT solvers perform variability constraint checking in off-line mode, we suggest in this ongoing research paper the integration of RuVa, a runtime variability algorithm, with the FaMa tool suite to check feature constraints dynamically before a new feature is added or an existing feature is removed. This research suggests a novel approach to modifying the variability model of context-aware systems dynamically and check the feature constraints on the fly.We integrate our solution with a SAT solver that can be invoked at runtime by a cyber-physical system.We validate the effectiveness and performance of the proposed algorithm using simulations.We also provide a proof-of-concept for updating the configuration of a robot’s variability model based on contextual changes.

Probabilistic Generative Model for Hyperspectral Unmixing Accounting for Endmember Variability

The complex nature of hyperspectral images makes the analysis of spectral signatures a challenging task in remote sensing. For quantitative analysis, spectral unmixing is a well-established and effective tool to analyze the spectra and spatial distribution of substances in the scene. The classical unmixing algorithms usually fail to tackle spectral variability caused by variations in environmental conditions. Many variants based on the linear mixing process have been proposed to tackle this problem; however, the spectral variability modeling capacity of these algorithms is usually insufficient. In this article, we present a probabilistic generative model to address endmember variability and provide more accurate abundance and endmember estimates. The proposed model simultaneously extracts the endmembers and estimates abundances in an unsupervised manner. In particular, it allows fitting arbitrary endmember distributions through the nonlinear modeling capability of neural networks compared to other methods that use parametric endmember variability models. The performance of the proposed approach is evaluated on both synthetic and real datasets. Experimental results show its superiority in comparison with other state-of-the-art methods. The code of this work is available at <uri xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">https://github.com/shuaikaishi/PGMSU</uri> for the sake of reproducibility.

Neuromorphic In-Memory RRAM NAND/NOR Circuit Performance Analysis in a CNN Training Framework on the Edge for Low Power IoT

Training a CNN involves computationally intense optimization algorithms to fit the network using a training dataset, to update the network weight for inferencing and then pattern classification. Hence, the application of in-memory computation would enable a highly power-efficient low latency on-the-edge CNN training technique by avoiding the memory wall created during the external memory read/write operation (for off chip instruction and data transfer). A memory write-verify, and re-program technique can control the RRAM variability. Still, memory verification and re-program is a complex process with additional resources needed for practical implementation of verification circuit. In this study, we have demonstrated a practical (First-in Max-Out) FIMO-based cache memory called Maximum Count Binary Comparator Layer (MCBC), using 1T3R, 1T5R, and 1T7R RRAM structures by using a probability-based accuracy improvement architecture, without the conventional verification process. We constructed 10 layered modified MobileNET with filter size ranging from 32 - 512 and trained with Traffic Sign Recognition Database (TSRD) using a three-tier abstraction simulation learning framework - (1) High level, 10 layered CNN implementation with Python+TensorFlow; (2) Verilog HDL based FP32MUL and FP32ADD (32-bits Floating Point adder and multiplier) circuits constructed with RRAM NAND gates using 1T2R structures; and (3) Digital Look-Up-Table (LUT) model for RRAM variability. An edge learning framework (for the forward pass) is demonstrated using digital RRAM-NAND/NOR universal gates integrated with the Maximum Count Binary Comparator Layer (MCBC) to partially circumvent the impact of RRAM variability and to quantify the RRAM variability on the CNN training prediction accuracy for 65nm CMOS OxRAM (TiN/HfO <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub> /Hf/TiN) with varying device current compliance of 5, 10, and 50μA for low power IoT applications. The MCBC layer was simulated using a SPICE model, for which the estimated chip layout is 1150 × 1230 nm <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> per logical gate input, which resulted in an overall prediction accuracy improvement from 10% to 60% by repeating the logical operations of the NOR gate for {1, 3, 5, and 7} cycles respectively.

Ensemble Kalman filter based data assimilation for tropical waves in the MJO skeleton model

AbstractThe Madden–Julian oscillation (MJO) is the dominant component of tropical intraseasonal variability, with wide‐reaching impacts even on extratropical weather and climate patterns. However, predicting the MJO is challenging. One reason is the suboptimal state estimates obtained with standard data assimilation (DA) approaches. These are typically based on filtering methods with Gaussian approximations and do not take into account physical properties that are important specifically for the MJO. In this article, a constrained ensemble DA method is applied to study the impact of different physical constraints on the state estimation and prediction of the MJO. The quadratic programming ensemble (QPEns) algorithm utilized extends the standard stochastic ensemble Kalman filter (EnKF) with specifiable constraints on the updates of all ensemble members. This allows us to recover physically more consistent states and to respect possible associated non‐Gaussian statistics. The study is based on identical twin experiments with an adopted nonlinear model for tropical intraseasonal variability. This so‐called skeleton model succeeds in reproducing the main large‐scale features of the MJO and closely related tropical waves, while keeping adequate simplicity for fast experiments on intraseasonal time‐scales. Conservation laws and other crucial physical properties from the model are examined as constraints in the QPEns. Our results demonstrate an overall improvement in the filtering and forecast skill when the model's total energy is conserved in the initial conditions. The degree of benefit is found to be dependent on the observational setup and the strength of the model's nonlinear dynamics. It is also shown that, even in cases where the statistical error in some waves remains comparable with the stochastic EnKF during the DA stage, their prediction is improved remarkably when using the initial state resulting from the QPEns.

Towards a Proactive Variability Modeling Process for Cyber-Physical Systems Supported by Modern Product Line Engineering

Towards a Proactive Variability Modeling Process for Cyber-Physical Systems Supported by Modern Product Line Engineering

An Accurate Process-Induced Variability-Aware Compact Model-Based Circuit Performance Estimation for Design-Technology Co-Optimization

In sub-10nm FinFETs, Line-edge-roughness (LER) and metal-gate granularity (MGG) are the two most dominant sources of variability and are mostly modeled semi-empirically. In this work, compact models of LER and MGG are used. We show an accurate process-induced variability (PIV) aware compact model-based circuit performance estimation for Design-Technology Co-optimization (DTCO). This work is carried out using an experimentally validated BSIM-CMG model on a 7nm FinFET node. First, we have shown performance bench-marking of LER and MGG models with the state-of-the-art and shown {\textbackslash}4x({\textbackslash}2.3x) accuracy improvement for NMOS(PMOS) in the estimation of device figure of merits (DFoMs). Second, RO and SRAM circuits performance estimation is carried out for LER and MGG variability. Further, {\textbackslash}22\% more optimistic estimate of ({\sigma}/{\mu})\textsubscript{SHM} (Static Hold Margin) compared to the state-of-the-art model with V\textsubscript{DD} variation is shown. Finally, we demonstrate our improved DFoMs accuracy translated to more accurate circuits figure of merits (CFoMs) performance estimation. For worst-case SHM (3({\sigma}/{\mu})\textsubscript{SHM}@VDD=0.75 V) compared to state-of-the-art, dynamic(standby) power reduction by {\textbackslash}73\%({\textbackslash}61\%) is shown. Thus, our enhanced variability model accuracy enables more credible DTCO with significantly better performance estimates.

Power-based modelling of renewable variability in dispatch models with clustered time periods

The planning of future power systems with high shares of renewable generation requires modelling large and complex systems over long time periods, resulting in models which are computationally heavy to solve. For this reason methods that can be used to decrease the size of power system dispatch models are needed. A common method in large scale planning models is to decrease the model size by increasing the size of the time steps. However, using larger time steps makes the representation of variability of renewable generation and load less accurate, which can affect the results from the model. In this paper, we investigate the possibility to use a power-based version of an economic dispatch model to decrease the model time resolution while getting results which are close to the original high-resolution model. We implement both power-based and the conventional, energy-based, versions of a dispatch model with different time resolutions, and show that the power-based model has better agreement with the high-resolution model, especially as the model time step increases. For example, using the power-based model gives more accurate results for wind power curtailment in a high-renewable scenario.

Impact of various NBTI distributions on SRAM performance for FinFET technology

NBTI degradation of FinFET devices is stochastic due to the inevitable variations in manufacturing processes. The stochastic behavior of NBTI degradation is modeled with different ΔVTH distributions. Various leading semiconductor industries report different ΔVTH distributions like Gamma, Normal, and Dispersive Skellam (DS) to model the NBTI induced variability. In this work, different NBTI induced variability models from the literature are used to investigate the impact on circuit performance. The combined effect of time-zero and various NBTI induced variability distributions on SRAM read and write metrics is discussed in detail. NBTI degradation in SRAM depends on the stored data pattern. Hence, we also report the impact of stored data patterns on the SRAM metric. To further analyze the impact of ΔVTH distributions on SRAM metrics, the correlation between SRAM metrics is studied.

Conditional Neural Process for nonparametric modeling of active galactic nuclei light curves

AbstractThe consequences of complex disturbed environments in the vicinity of a supermassive black hole are not well represented by standard statistical models of optical variability in active galactic nuclei (AGN). Thus, developing new methodologies for investigating and modeling AGN light curves is crucial. Conditional Neural Processes (CNPs) are nonlinear function models that forecast stochastic time series based on a finite amount of known data without the use of any additional parameters or prior knowledge (kernels). We provide a CNP algorithm that is specifically designed for simulating AGN light curves. It was trained using data from the All‐Sky Automated Survey for Supernovae, which included 153 AGN. We present CNP modeling performance for a subsample of five AGNs with distinctive difficult‐to‐model properties. The performance of CNP in predicting temporal flux fluctuation was assessed using a minimizing loss function, and the results demonstrated the algorithm's usefulness. Our preliminary parallelization experiments show that CNP can efficiently handle large amounts of data. These results imply that CNP can be more effective than standard tools in modeling large volumes of AGN data (as anticipated from time‐domain surveys such as the Vera C. Rubin Observatory's Legacy Survey of Space and Time).

Investigating flood processes in karst catchments by combining concentration-discharge relationship analysis and lateral flow simulation

Concentration-Discharge (C-Q) relationship analyses allow better understanding catchment water origin and hydrological process variability across time (from storm-event to pluriannual scale). Though, those approaches mostly include integrative methods that consider catchments as a whole. The aim of this article is to assess the potential of a combined approach to investigate the spatio-temporal variability of flood processes within catchments (across river reaches and seasons), focusing on karst areas. The methodology consists in combining C-Q relationship analysis in nested catchments and simulation of lateral exchanges at the reach scale. For that, we analyze high-frequency records of discharge and electrical conductivity (EC) in karst catchments, in which EC informs on water residence time. Contributions of pre-event water (PEW) and event water (EW) during storm events are investigated through hysteresis loop analysis at each gauging station. Combined with lateral EC simulation at the reach scale, they make it possible to infer runoff processes during storms. A unary karst catchment in a temperate/mountainous climate (Loue River) and a binary catchment in a Mediterranean climate (Cèze River) were selected in France, and their conceptual models of water origin and hydrological process variability were drawn. In the Loue catchment, summer and fall storm events are characterized by contributions of PEW through piston-type flows, whereas decreasing EC values in winter and spring storm events indicate the major contribution of EW through surface runoff and fast infiltration. Within the catchment, EW contribution increases downstream, due to the canyon river morphology grading into open valleys. Regarding the Cèze catchment, contributions of EW are higher, indicating that fast infiltration and surface runoff are the dominant processes, associated with a PEW signature in summer and fall. Within the catchment, PEW contribution is found higher in karstified areas. These results show that flood process seasonality is mainly related to karst aquifer saturation rate, while intra-site variability is linked to karst area extension and river morphology. Results are encouraging to extend this approach to a variety of sites, notably those affected by significant surface water-groundwater interaction and groundwater flooding. Such approach, by providing a discretized information on flood processes, could help refining lumped hydrological models, or facilitate the use of semi-distributed ones.

GRASS: Distinguishing Planet-induced Doppler Signatures from Granulation with a Synthetic Spectra Generator

Abstract Owing to recent advances in radial-velocity instrumentation and observation techniques, the detection of Earth-mass planets around Sun-like stars may soon be primarily limited by intrinsic stellar variability. Several processes contribute to this variability, including starspots, pulsations, and granulation. Although many previous studies have focused on techniques to mitigate signals from pulsations and other types of magnetic activity, granulation noise has to date only been partially addressed by empirically motivated observation strategies and magnetohydrodynamic simulations. To address this deficit, we present the GRanulation And Spectrum Simulator (GRASS), a new tool designed to create time-series synthetic spectra with granulation-driven variability from spatially and temporally resolved observations of solar absorption lines. In this work, we present GRASS, detail its methodology, and validate its model against disk-integrated solar observations. As a first-of-its-kind empirical model for spectral variability due to granulation in a star with perfectly known center-of-mass radial-velocity behavior, GRASS is an important tool for testing new methods of disentangling granular line-shape changes from true Doppler shifts.

Assessment of reclamation-induced consolidation settlement considering stratigraphic uncertainty and spatial variability of soil properties

Consolidation analysis is a key task for reclamation design. Although consolidation is a long-term process, acceleration of consolidation is often preferred for speeding up the reclamations. Before proposing measures to accelerate the consolidation and reclamation process, it is imperative to have an accurate prediction of consolidation settlement for fine-grained materials, which is greatly affected by spatial distribution of subsurface zones with different soil types (i.e., stratigraphic heterogeneities and uncertainty) and spatial variability of soil properties. In current practice, calculation of consolidation settlement often uses simplified stratigraphic boundaries and deterministic consolidation parameters without considering stratigraphic uncertainty or soil property spatial variability. The oversimplified practice might result in unconservative estimations of consolidation settlement and pose threats to safety and serviceability of constructed facilities on reclaimed lands. In this study, a stochastic framework is proposed for consolidation settlement assessment with explicit modeling of stratigraphic uncertainty and spatial variability of soil properties by machine learning and random field simulation from limited site investigation data. Performance of the proposed framework is demonstrated using an illustrative example. Results indicate that the proposed method effectively generates multiple realizations of geological cross section and random field samples of geotechnical properties from limited measurements and offers valuable insights into spatial distribution of the estimated total primary consolidation settlement curves and angular distortion.

Environmental Variability and Threshold Model’s Predictions for Coral Reefs

Current models of the future of coral reefs rely on threshold (TM) and multivariate environmental variability models (VM) that vary in how they account for spatial and temporal environmental heterogeneity. Here, a VM based on General Additive Model (GAM) methods evaluated the empirical relationships between coral cover (n= 905 sites pooled to 318 reef cells of the Western and Central Indian Ocean Provinces) and 15 potentially influential variables. Six environmental and one fisheries management variables were selected as significant including SST shape distributions, dissolved oxygen, calcite, and fisheries management. Common predictive variables, including cumulative degree-heating weeks (DHW), pH, maximum light, SST bimodality and rate of rise, and two multivariate metrics were either weak or not significant predictors of coral cover. A spatially-resolved 2020 baseline for future predictions of coral cover within 11,678 reef ∼6.25 km2cells within 13 ecoregions and 4 fisheries management categories using the 7 top VM variables was established for comparing VM and TM coral cover prediction for the year 2050. We compared the two model’s predictions for high and low Relative Concentration Pathway (CMIP5; RCP8.5 and 2.6) scenarios using the four available future-cast SST variables. The excess heat (DHW)-coral mortality relationship of the TM predicted considerably lower coral cover in 2050 than the VM. For example, for the RCP8.5 and RCP2.6 scenarios, the decline in coral for the TM predicted was 81 and 58% compared to a 29 and 20% for the VM among reef cells with &amp;gt;25% coral cover in 2020, if a proposed optimal fisheries management was achieved. Despite differences, coral cover predictions for the VM and TM overlapped in two environmental regions located in the southern equatorial current region of the Indian Ocean. Historical and future patterns of acute and chronic stresses are expected to be more influential than cumulative heat stress in predicting coral cover, which is better accounted for by the VM than the TM.