Search NASA⌕ Search

SEARCH · Search NASA

Results for “training effectiveness”

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 127 records · Page 7

NiO/Fe(001): Magnetic anisotropy, exchange bias, and interface structure

The magnetic and structural properties of NiO/Fe epitaxial bilayers grown on MgO(001) were studied using magnetooptic Kerr effect (MOKE) and conversion electron Moessbauer spectroscopy (CEMS). The bilayers were prepared under ultra high vacuum conditions using molecular beam epitaxy with oblique deposition. Two systems were compared: one showing the exchange bias (100ML-NiO/24ML-Fe), ML stands for a monolayer, and another where the exchange bias was not observed (50ML-NiO/50ML-Fe). For both, the magnetic anisotropy was found to be complex, yet dominated by the growth-induced uniaxial anisotropy. The training effect was observed for the 100ML-NiO/24ML-Fe system and quantitatively described using the spin glass model. The composition and magnetic state of the interfacial Fe layers were studied using {sup 57}Fe-CEMS. An iron oxide phase (Fe{sup 3+}{sub 4}Fe{sup 2+}{sub 1}O{sub 7}), as thick as 31 A, was identified at the NiO/Fe interface in the as-deposited samples. The ferrimagnetic nature of the interfacial iron oxide film explains the complex magnetic anisotropy observed in the samples.

36 MATERIALS SCIENCE↗

Efficient Reinforcement Learning for Real-Time Hardware-Based Energy System Experiments: Preprint

In the context of urgent climate challenges and the pressing need for rapid technology development, Reinforcement Learning (RL) stands as a compelling data-driven method for controlling real-world physical systems. However, RL implementation often entails time-consuming and computationally intensive data collection and training processes, rendering them inefficient for real-time applications that lack non-real-time models. To address these limitations, real-time emulation techniques have emerged as valuable tools for the lab-scale rapid prototyping of intricate energy systems. While emulated systems offer a bridge between simulation and reality, they too face constraints, hindering comprehensive characterization, testing, and development. In this research, we construct a surrogate model using limited data from simulated systems, enabling an efficient and effective training process for a Double Deep Q-Network (DDQN) agent for future deployment. Our approach is illustrated through a hydropower application, demonstrating the practical impact of our approach on climate-related technology development.

deep Q-learning↗

Virtual Reality for Shoot/No-Shoot Decision Training in Law Enforcement: A Literature Review and Research Agenda

Virtual reality (VR) can materially improve “shoot / no-shoot” (SNS) training by giving officers realistic, repeatable practice making high-stakes decisions under pressure. Traditional tools—live-fire ranges and video simulators—build basics, but they cannot adapt to each officer in real time or fully mirror the complexity of the field. VR closes that gap by creating immersive scenarios that are safer, more flexible, easier to scale across units, and able to capture objective performance data. SNS decisions are not just about marksmanship; they rely on perception, judgment, memory, and the ability to hold fire when a threat is uncertain. Effective training therefore needs realism, decision complexity, and branching outcomes that reflect the true consequences of choices. These elements strengthen recognition of hostile intent while reducing false positives and building the self-control required in ambiguous situations. VR brings specific advantages: dynamic environments, full-body interaction, and the ability to measure performance with precision—enabling targeted feedback and better transfer of learning to the street. At the same time, responsible deployment must address scenario quality (credible environments and behaviors), lawful decision models, and user wellbeing (appropriate stress levels, comfort, and safety). Sandia’s VIPER Lab is positioned to lead this work. The team combines human-performance science, AI/ML, and VR/AR development with a deep equipment bench (e.g., omnidirectional treadmill, eye-tracking, haptics, multiple HMDs). This ecosystem supports building and validating next-generation SNS training that is immersive, measurable, and trustworthy. Bottom line: Investment in VR-enabled SNS training that blends evidence-based design with careful validation and legal safeguards is expected to pay off in safer, more consistent decision-making and improved community trust, delivered through training that is practical to deploy at scale.

45 MILITARY TECHNOLOGY, WEAPONRY, AND NATIONAL DEF↗

“Just Do It” Mission Operations Training in a COVID World

“No”- “not”- “can’t do it” – these words don’t fly in M2020 Mission Operations. The M2020 Surface Mission Operations Team trained for landing the Perseverance Rover under COVID-19 remote work/mandatory stay-at-home conditions. Training activities included presenting Flight Schools to the team via video conferencing, training COVID personal safety requirements to on-premises staff, and constantly updating and communicating COVID restrictions to the team as safety requirements changed. This paper explores the impact of COVID on the Mission System Training for M2020 Mission Operations. The layers of COVID, aptly named the “COVID Tax” by team management, affected project roles, communications, personnel interactions, operations facility usage and training exercises practiced by the team. Flight Schools and Operational Readiness Tests (ORTs) are driving forces behind the surface mission operations training. These activities work hand-in-hand to prepare the team for landing day, surface operations that transition from cruise to nominal operations, and nominal operations. Under normal training conditions, Flight Schools and ORTs are only concerned with tactical operations for the Uplink (Command) Downlink (Analysis) and Campaign Implementation (Planning) Teams of scientists and system engineers. Due to COVID, and the necessity to maintain physical distance between people, training for the landing team needed to include the new category of COVID personal safety. COVID also necessitated remote teams and video conferencing of the entirety of Flight Schools. This reliance on distance learning had not been done on previous missions. We will explore the advantages and disadvantages of video conferencing as a training platform; training effectiveness in communicating and practicing the multiple changes to the COVID safety protocols; and the timing in which we received and responded to COVID directives. Additionally, this paper will review the practical measures taken by the on-premises team and how the team adapted during the readiness tests, landing, and early mission operations, as well as how the team responded to the post-vaccine ramping down of COVID Protocols. The M2020 Surface Mission Operations Team responded very well to the challenges of COVID. All pre-landing operational and COVID-related training was completed. Post-landing COVID Training was provided as needed to new on-premises personnel. Training presented on COVID for the first operational readiness test (ORT) consisted of 2.5 hours of training. Each ORT had COVID Training. A person who had participated in all of the ORTs from September 2020 – February 2021 would have received over 7 hours of COVID Training. By ORT-12 (approximately 8 weeks after the first ORT), the Training team consolidated the original COVID training to a one-hour COVID Basic Training course, with additional recommended training. The Basic training course was updated after landing, as vaccines became available. COVID Training had multiple Flight Schools in the self-directed Blackboard Learning system, as well as a Quick Reference Wiki for Onboarding of new on-premises personnel and keeping on-premises personnel up-to-date. COVID Training covered many topics including practical methods of 6-foot distancing, how to interact with an IT professional when help was needed at a workstation, and challenging indoor meal-eating protocols. COVID Training continued to be updated, as the lab and the project respond to the virus variants and federal, state and local ordinances.

Rosette, Theresa↗

Demonstration of Decentralized Physics-Driven Learning

In typical artificial neural networks, neurons adjust according to global calculations of a central processor, but in the brain, neurons and synapses self-adjust based on local information. Contrastive learning algorithms have recently been proposed to train physical systems, such as fluidic, mechanical, or electrical networks, to perform machine-learning tasks from local evolution rules. However, to date, such systems have only been implemented in silico due to the engineering challenge of creating elements that autonomously evolve based on their own response to two sets of global boundary conditions. Here, we introduce and implement a physics-driven contrastive learning scheme for a network of variable resistors, using circuitry to locally compare the response of two identical networks subjected to the two different sets of boundary conditions. Using this method, our system effectively trains itself, optimizing its resistance values without the use of a central processor or external information storage. Once the system is trained for a specified allostery, regression, or classification task, the task is subsequently performed rapidly and automatically by the physical imperative to minimize power dissipation in response to the given voltage inputs. Here, we demonstrate that, unlike typical computers, such learning systems are robust to extreme damage (and thus manufacturing defects) due to their decentralized learning. Our twin-network approach is therefore readily scalable to extremely large or nonlinear networks, where its distributed nature will be an enormous advantage; a laboratory network of only 500 edges will already outpace its in silico counterpart.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

Particle Size-Dependent Zero-Field Exchange Bias in LaFeO {sub 3} Nanoparticles

The present communication deals with particle size-dependent zero-field cooled exchange effect in LaFeO{sub 3} (LFO). LFO nanoparticles were prepared by PVA-based sol-gel method. The sample was calcined at different temperatures (700, 750, and 800 {sup ∘}C) to tune the average particle size. The detailed crystal structural and their geometrical parameters were examined by means of Rietveld refinement using x-ray diffraction data. The different techniques such as Scherrer’s method, Williamson Hall method, FESEM, and TEM were applied to accurately determine the average particle size. Zero-field cooled exchange bias and weak ferromagnetism in antiferromagnetic nanoparticles was observed for all samples using hysteresis loop measurement. The training effect was studied to confirm that the loop shift originates due to exchange bias. Exchange bias as well as coercivity are strongly related to the particle size at low and room temperature. The results clearly demonstrate that the process of calcination tailor the overall properties of material which makes it a promising candidate for device applications.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

Lessons Learned from NASA Goddard Space Flight Center’s Product Development Lead Training Schedule and Cost Development Workshop: Continuous Improvement

This presentation provides a status of the Goddard Space Flight Center (GSFC) effort to increase foundational knowledge of Product Development Leads (PDLs) in schedule and cost management including earned value management (EVM). In 2012, GSFC’s Engineering and Technology Directorate (ETD) implemented an in-house training program to prepare PDLs for managing the technical, cost, schedule, and risk aspects of spaceflight systems to meet their subsystem commitments. Developed in-house, the PDL training program provides an integrated approach to requirements development, risk, schedule and cost management, EVM, performance tracking, and other areas. The program has been held twice yearly since its inception with 531 participating and 451 completing the curriculum. In 2017, the program won the Robert H. Goddard award for Quality and Process Improvement. Program development and evolution were presented in the 2018 NASA Schedule and Cost Symposium. The presentation was so well received that this year we focus on one workshop within the program: Schedule and Cost Development, including EVM. We examine the on-going logic modeling process and how participant and stakeholder data influence workshop content and design, and how the disciplines of schedule and cost contribute to mission success. In this presentation we refresh you on how the approach integrates lecture, small group discussion, estimating, case study exercises, and problem solving. We update you on the data collected from participants and stakeholders, and we discuss how we use these data to measure training effectiveness. Specific topics include: • How the logic model is used as the backbone for continuous program improvement, • How feedback influences implementation and curriculum updates, • How data collection and analysis inform workshop content and development, including participant discoveries of EVM data, • How including the resource analyst and planner in the product development team supports project success.

Lessons Learned↗

Magnetization reversal in TmCrO{sub 3}

Highlights: ► We observed two magnetization reversals in TmCrO{sub 3}. ► The reversal at 28 K is attributed to antiparallel coupling between Cr{sup 3+} and Tm{sup 3+}. ► The other reversal originates from spin reorientation. ► Magnetocaloric effect is observed at the spin reorientation temperature. ► Characteristic magnetization switching is demonstrated. -- Abstract: The perovskite chromite TmCrO{sub 3} shows magnetization reversal at two temperatures. The reversal at ∼28 K is attributed to the antiparallel coupling between Tm{sup 3+} and Cr{sup 3+} moments, while that at the lower temperature (∼6–7 K) is rooted in a rotation of the magnetic moments. Magnetocaloric measurements offer a relatively large entropy change (∼4–5 J kg{sup −1} K{sup −1}) at the lower temperature. The reversal at ∼28 K is accompanied by a sign change of an exchange-bias-like field. The absence of the training effect suggests that this behavior is rooted in unidirectional magnetic anisotropy. The existence of the two magnetization reversals offers the characteristic switching of magnetization. For example, the magnetization is flipped without changing the direction of the applied magnetic field.

36 MATERIALS SCIENCE↗

Adaptation to vection-induced symptoms of motion sickness

The effects of repeated exposures to a rotating circular vection drum on the symptoms of motion sickness and tachygastria in humans were investigated. Subjects were sitting in a drum and were exposed to 15 min baseline (no rotation), followed by 15 min drum rotation at 60 deg/s, and, then, by 15 min recovery. Gastric myoelectric activity was continuously recorded with the electrogastrogram. Subjects who were exposed to the drum three times with intervals of 4-24 days all showed symptoms of tachygastria and failed to show an amelioration of motion sickness symptoms. On the other hand subjects who had only 48 h between the three sessions of drum exposure, experienced a reduction in motion-sickness symptoms and in tachygastsria upon repeated exposure to the drum, indicating that training effected a symptomatic and physiological adaptation. It is suggested that preflight adaptation to visual-vestibular sensory mismatch may reduce motion sickness in astronauts.

Stern, Robert M.↗

Enhanced exchange bias effect in size modulated Sm{sub 0.5}Ca{sub 0.5}MnO{sub 3} phase separated manganite

The effect of grain size modulation on exchange bias effect in CE-type antiferromagnetic Sm{sub 0.5}Ca{sub 0.5}MnO{sub 3} phase separated manganite is reported here. With the reduction of particle size, ferromagnetic clusters are found to form in the charge ordered antiferromagnetic matrix and gradually become larger. The horizontal and vertical shifts of the magnetic hysteresis loops in the field cooled magnetization process clearly indicate the size dependent exchange bias effect and it can be tuned with the reduction of particle sizes. The values of exchange bias parameter, i.e., exchange bias field (H{sub E}), coercivity (H{sub C}), remanence asymmetry (M{sub E}), and magnetic coercivity (M{sub C}) are found to depend strongly on the particle size. The variations of H{sub E} follow non-monotonic dependencies with reduction in particle size and show maximum (1205 Oe) at particle size of 150 nm at T = 5 K, which can be ascribed due to the changes in uncompensated surface spins. The values of H{sub E} and M{sub E} are found to decrease exponentially with increasing temperature below the spin- or cluster-glass like freezing temperature. The spin relaxation model has been employed for analysis of large magnetic training effect. The linear relationship between H{sub E} and M{sub E} further confirms the role of uncompensated surface spins. In view of spintronics application of manganites, the present observation of large exchange bias shift in this half-doped manganite may have great technological importance.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

Quantum Loop Topography for Machine Learning

Despite rapidly growing interest in harnessing machine learning in the study of quantum many-body systems, training neural networks to identify quantum phases is a nontrivial challenge. The key challenge is in efficiently extracting essential information from the many-body Hamiltonian or wave function and turning the information into an image that can be fed into a neural network. When targeting topological phases, this task becomes particularly challenging as topological phases are defined in terms of nonlocal properties. Here, we introduce quantum loop topography (QLT): a procedure of constructing a multidimensional image from the “sample” Hamiltonian or wave function by evaluating two-point operators that form loops at independent Monte Carlo steps. The loop configuration is guided by the characteristic response for defining the phase, which is Hall conductivity for the cases at hand. Feeding QLT to a fully connected neural network with a single hidden layer, we demonstrate that the architecture can be effectively trained to distinguish the Chern insulator and the fractional Chern insulator from trivial insulators with high fidelity. In addition to establishing the first case of obtaining a phase diagram with a topological quantum phase transition with machine learning, the perspective of bridging traditional condensed matter theory with machine learning will be broadly valuable.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Kinematics and Kinetics of Squat and Deadlift Exercises with Varying Stance Widths

The primary motion of squat and deadlift exercise involves flexion and extension of the hips, knees, and ankles, but each exercise can be performed with variations in stance width. These variations may result in differing kinematics and ground reaction forces (GRF), which may in turn affect joint loading. PURPOSE: The purpose of this investigation was to compare ankle, knee, and hip kinematics and kinetics of normal squat (NS), wide-stance squat (WS), normal deadlift (ND), and sumo deadlift (SD). We hypothesized that hip joint kinematics and work at each joint would differ between exercise variations. METHODS: Six subjects (3 m/3 f; 70.0 plus or minus 13.7 kg; 168 plus or minus 9.9 cm) performed each lift in normal gravity on the ground-based version of the Advanced Resistive Exercise Device (ARED) used on the International Space Station. The ARED provided resistance with a combination vacuum tube/flywheel mechanism designed to replicate the gravitational and inertial forces of free weights. Subjects completed each lift with their 10-repetition maximum load. Kinematic data were collected at 250 Hz by a 12-camera motion-capture system (Smart-D, BTS Bioengineering, Milan, Italy), and GRF data were collected at 1000 Hz with independent force platforms for each leg (Model 9261, Kistler Instruments AG, Winterhur, Switzerland). All data were captured simultaneously on a single workstation. The right leg of a single lift for each motion was analyzed. Modeling software (OpenSim 2.2.0, Simbios, Palo Alto, CA) determined joint kinematics and net positive and negative work at each lower extremity joint. Total work was found as the sum of work across all joints and was normalized by system mass. Effect sizes and their 95% confidence intervals were computed between conditions. RESULTS: Peak GRF were similar for each lift. There were no differences between conditions in hip flexion range of motion (ROM). For hip adduction ROM, there were no differences between the NS, WS, and SD. However, hip adduction ROM was greater during the NS and SD than during the ND. Hip rotation ROM was greater during the WS than during the NS and SD, and was greater during the SD than during the ND. For knee and ankle flexion ROM, the ND, WS, and SD were not different, but ROM was greater during the NS than the ND and greater during the WS than the SD. Total eccentric work was greater during the WS than the SD. Otherwise, there were no differences in eccentric or concentric work between conditions. CONCLUSIONS: Although squat and deadlift exercises consist of similar motions, there are kinematic differences between them that depend on stance width. Total eccentric and concentric work are similar for different lifts, but differing kinematics may require activation of different musculature for each variation. With respect to each condition, in the ND the ROM of each joint tended to be less, and the WS tended to trade knee motion for hip motion. PRACTICAL APPLICATIONS: Knowledge of differences in kinematics and kinetics between different squat and deadlift variations is important for coaches and rehabilitation personnel to understand when prescribing exercise. Our results suggest that each variation of the squat and deadlift should be considered a separate exercise that may induce different long-term training effects.

DeWitt, John K.↗

Improved pulmonary function in working divers breathing nitrox at shallow depths

INTRODUCTION: There is limited data about the long-term pulmonary effects of nitrox use in divers at shallow depths. This study examined changes in pulmonary function in a cohort of working divers breathing a 46% oxygen enriched mixture while diving at depths less than 12 m. METHODS: A total of 43 working divers from the Neutral Buoyancy Laboratory (NBL), NASA-Johnson Space Center completed a questionnaire providing information on diving history prior to NBL employment, diving history outside the NBL since employment, and smoking history. Cumulative dive hours were obtained from the NBL dive-time database. Medical records were reviewed to obtain the diver's height, weight, and pulmonary function measurements from initial pre-dive, first year and third year annual medical examinations. RESULTS: The initial forced vital capacity (FVC) and forced expiratory volume in 1 s (FEV1) were greater than predicted, 104% and 102%, respectively. After 3 yr of diving at the NBL, both the FVC and FEV1 showed a significant (p < 0.01) increase of 6.3% and 5.5%, respectively. There were no significant changes in peak expiratory flow (PEF), forced mid-expiratory flow rate (FEF(25-75%)), and forced expiratory flow rates at 25%, 50%, and 75% of FVC expired (FEF25%, FEF50%, FEF75%). Cumulative NBL dive hours was the only contributing variable found to be significantly associated with both FVC and FEV1 at 1 and 3 yr. CONCLUSIONS: NBL divers initially belong to a select group with larger than predicted lung volumes. Regular diving with nitrox at shallow depths over a 3-yr period did not impair pulmonary function. Improvements in FVC and FEV1 were primarily due to a training effect.

Non-programmatic↗

The effects of exercise training interventions on depression in hemodialysis patients

Purpose Depression considerably influences the clinical outcomes, treatment compliance, quality of life, and mortality of hemodialysis patients. Exercise plays a beneficial role in depressive patients, but its quantitative effects remain elusive. This study aimed to summarize the effects of exercise training on depression in patients with end-stage renal disease undergoing hemodialysis. Methods The PUBMED, EMBASE, and Cochrane Library databases were systematically searched from inception to April 2023 to identify published articles reporting the effect of exercise training on the depression level of patients with End-Stage Renal Disease undergoing hemodialysis. Data were extracted from the included studies using predefined data fields by two independent researchers. The Cochrane Handbook for Systematic Reviews of Interventions and Joanna Briggs Institute Critical Appraisal Checklist for Quasi-Experimental Studies were employed for quality evaluation. Results A total of 22 studies enrolling 1,059 patients who participated in exercise interventions were included. Hemodialysis patients exhibited superior outcomes with intradialytic exercise (SMD = −0.80, 95% CI: −1.10 to −0.49) and lower levels of depression following aerobic exercise (SMD = −0.93, 95%CI: −1.32 to −0.55) compared to combined exercise (c − 0.85, 95% CI: −1.29 to −0.41) and resistance exercise (SMD = −0.40, 95%CI: −0.96 to 0.17). Regarding exercise duration, patients manifested lower depression levels when engaging in exercise activities for a duration exceeding 6 months (SMD = −0.92, 95% CI: −1.67 to −0.17). Concerning the duration of a single exercise session, the most significant improvement was noted when the exercise duration exceeded 60 min (SMD = −1.47, 95% CI: −1.87 to −1.06). Conclusion Our study determined that exercise can alleviate depression symptoms in hemodialysis patients. This study established the varying impacts of different exercise parameters on the reduction of depression levels in hemodialysis patients and is anticipated to lay a theoretical reference for clinicians and nurses to devise tailored exercise strategies for interventions in patients with depression. Systematic review registration https://www.crd.york.ac.uk/prospero/ , This study was registered in the International Prospective Register of Systematic Reviews (PROSPERO) database, with registration number CRD42023434181.

Yu, Huihui↗

A novel implicit hybrid machine learning model and its application for reinforcement learning

A novel methodology to develop implicit hybrid models is presented. PyTorch is used to integrate physics-based equations with machine learning models. Automatic differentiation of the hybrid model is leveraged to solve the implicit equations. Iterative solving enables gradient based updates to the machine learning model. The novel methodology is compared to an explicit hybrid approach on a continuously stirred tank reactor (CSTR). The novel method results in a lower modelling error. Both hybrid models effectively train with noisy data. To test the implicit hybrid model, it is employed as a reinforcement learning (RL) training model. The RL algorithm trained on the hybrid model outperforms real time optimization of the CSTR and performs nearly as well as RL trained directly on the CSTR and a traditional gradient based approach. Training RL directly on the CSTR requires over 60,000 system interactions compared to 6000 historical data points for hybrid model development.

42 ENGINEERING↗

Bayesian sparse learning with preconditioned stochastic gradient MCMC and its applications

Deep neural networks have been successfully employed in an extensive variety of research areas, including solving partial differential equations. Despite its significant success, there are some challenges in effectively training DNN, such as avoiding overfitting in over-parameterized DNNs and accelerating the optimization in DNNs with pathological curvature. Here, we propose a Bayesian type sparse deep learning algorithm. The algorithm utilizes a set of spike-and-slab priors for the parameters in the deep neural network. The hierarchical Bayesian mixture will be trained using an adaptive empirical method. That is, one will alternatively sample from the posterior using preconditioned stochastic gradient Langevin Dynamics (PSGLD), and optimize the latent variables via stochastic approximation. The sparsity of the network is achieved while optimizing the hyperparameters with adaptive searching and penalizing. A popular SG-MCMC approach is Stochastic gradient Langevin dynamics (SGLD). However, considering the complex geometry in the model parameter space in nonconvex learning, updating parameters using a universal step size in each component as in SGLD may cause slow mixing. To address this issue, we apply a computationally manageable preconditioner in the updating rule, which provides a step-size parameter to adapt to local geometric properties. Moreover, by smoothly optimizing the hyperparameter in the preconditioning matrix, our proposed algorithm ensures a decreasing bias, which is introduced by ignoring the correction term in the preconditioned SGLD. According to the existing theoretical framework, we show that the proposed algorithm can asymptotically converge to the correct distribution with a controllable bias under mild conditions. Numerical tests are performed on both synthetic regression problems and learning solutions of elliptic PDE, which demonstrate the accuracy and efficiency of the present work.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Antiferromagnet thickness dependence and rotatable spins in exchange biased CoO/Fe films

The emergence of exchange bias and coercivity enhancement has been investigated in epitaxial CoO/Fe films with varied antiferromagnet (AF) thicknesses, even smaller than the critical value where the frozen CoO spins are detectable. Vector magnetometry and first-order reversal curve (FORC) measurements reveal different CoO thickness dependence of the exchange bias and coercivity enhancement, including the evolution of magnetization reversal from a high coercivity, low bias phase due to rotatable CoO moments to a high bias, low coercivity phase due to frozen CoO moments. Furthermore, the AF domain state is found to be metastable, which can be reoriented by external and exchange fields prior to the appearance of frozen spins, pointing to a generic origin of the training effect. Monte Carlo simulations show that the AF anisotropy energy barrier and the rotatable spins induced by magnetic field and exchange interaction at the interface are responsible for the observed effects.

36 MATERIALS SCIENCE↗

Ambient temperature modelling with soft computing techniques

This paper proposes a hybrid approach based on soft computing techniques in order to estimate monthly and daily ambient temperature. Indeed, we combine the back-propagation (BP) algorithm and the simple Genetic Algorithm (GA) in order to effectively train artificial neural networks (ANN) in such a way that the BP algorithm initialises a few individuals of the GA's population. Experiments concerned monthly temperature estimation of unknown places and daily temperature estimation for thermal load computation. Results have shown remarkable improvements in accuracy compared to traditional methods. (author)

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗