Step Size Research Articles

A simplified reliability-based design method for multiple design variables considering different limit states

A generalized reliability model comprising the objective, constraint, and judgment functions is established for the reliability index approach (RIA), taking parameters’ properties of engineering practice and negative reliability index into consideration. Based on this, the reliability-based design (RBD) problem with multiple design variables is translated into the solution to the nonlinear equations, and a simplified method consisting of a simple variant of the Newton iteration method and the finite difference method (FDM) is proposed. Numerical examples are presented to verify the efficiency of the proposed reliability approach and to determine the incremental step size for FDM. RBD of a simply supported beam is illustrated and the variabilities of design variables are investigated considering the uncertainties in the manufacturing process and practical operations. Results reveal that the variations of the design variables should not be ignored. Moreover, analysis results show that the design value might not intuitively increase with the increase of its coefficient of variation (CoV), and it might not increase with the increasing reliability requirement for problems involving multiple variables. The reasons for this phenomenon are very complicated, and it is a systematic problem. One should be aware of this phenomenon, and specific analysis is required for specific problems.

Stochastic modified equations for symplectic methods applied to rough Hamiltonian systems

Abstract We investigate stochastic modified equations to explain the mathematical mechanism of symplectic methods applied to rough Hamiltonian systems. The contribution of this paper is threefold. First, we construct a new type of stochastic modified equation. For symplectic methods applied to rough Hamiltonian systems, the associated stochastic modified equations are proved to have Hamiltonian formulations. Secondly, the pathwise convergence order of the truncated modified equation to the numerical method is obtained by techniques in rough path theory. Thirdly, if increments of noises are simulated by truncated random variables, we show that the error can be made exponentially small with respect to the time step size.

Using the Buckingham π Theorem for Multi-System Transfer Learning: A Case-Study with 3 Vehicles Sharing a Database

Many advanced driver assistance schemes or autonomous vehicle controllers are based on a motion model of the vehicle behavior, i.e., a function predicting how the vehicle will react to a given control input. Data-driven models, based on experimental or simulated data, are very useful, especially for vehicles difficult to model analytically, for instance, ground vehicles for which the ground-tire interaction is hard to model from first principles. However, learning schemes are limited by the difficulty of collecting large amounts of experimental data or having to rely on high-fidelity simulations. This paper explores the potential of an approach that uses dimensionless numbers based on Buckingham’s π theorem to improve the efficiency of data for learning models, with the goal of facilitating knowledge sharing between similar systems. A case study using car-like vehicles compares traditional and dimensionless models on simulated and experimental data to validate the benefits of the new dimensionless learning approach. Preliminary results from the case study presented show that this new dimensionless approach could accelerate the learning rate and improve the accuracy of the model prediction when transferring the learned model between various similar vehicles. Prediction accuracy improvements with the dimensionless scheme when using a shared database, that is, predicting the motion of a vehicle based on data from various different vehicles was found to be 480% more accurate for predicting a simple no-slip maneuver based on simulated data and 11% more accurate to predict a highly dynamic braking maneuver based on experimental data. A modified physics-informed learning scheme with hand-crafted dimensionless features was also shown to increase the improvement to precision gains of 917% and 28% respectively. A comparative study also shows that using Buckingham’s π theorem is a much more effective preprocessing step for this task than principal component analysis (PCA) or simply normalizing the data. These results show that the use of dimensionless variables is a promising tool to help in the task of learning a more generalizable motion model for vehicles, and hence potentially taking advantage of the data generated by fleets of vehicles on the road even though they are not identical.

Gait Characteristics over the Six-minute Walk Test in Older Women of a Specific Portuguese Community: A Cross-Sectional Study.

There is limited research on the changes that can occur in the gait biomechanical parameters of older adults over long-distance walking. Thus, this study aimed to evaluate the gait characteristics of older women of a specific Portuguese community over the six-minute walk test (6MWT) along with gait spatial and temporal parameters and angular kinematics of the lower limb joints. Twenty-six older women voluntarily participated. Each woman performed the 6MWT, and during this, their spatial and temporal parameters were collected through plantar pressure data (100Hz) and angular kinematics through an inertial sensor system (100Hz). The 6MWT was divided into four intervals, and the Friedman test was used to compare them. The median age of the women was calculated, and the Mann-Whitney test was used to compare women above and below the median age value. An increase in gait speed, stance phase, and double stance phase along the intervals was observed, as well as larger angular displacements of the hip, knee, and ankle (p<0.05). Women below or above the median age value also yielded these behaviours. Gait speed, cadence, stride length, and step length were higher in women below the median age value, while stride and step duration were lower (p<0.05). Older women yielded changes in gait spatial and temporal parameters along the 6MWT, as well as larger angular displacements of the lower limb joints. Older women (compared to younger ones) yielded lower gait speed, cadence, stride length, and step length, higher stance and double stance phases, and shorter angular displacements of the lower limb joints.

Shaping the energy curves of a servomotor-based hexapod robot.

The advantageous versatility of hexapod robots is often accompanied by high power consumption, while animals have evolved an energy efficient locomotion. However, there are a lack of methods able to compare and apply animals' energetic optimizations to robots. In this study, we applied our method to a full servomotor-based hexapod robot to evaluate its energetic performance. Using an existing framework based on the laws of thermodynamics, we estimated four metrics using a dedicated test bench and a simulated robotic leg. We analyzed the characteristics of a single leg to shape the energetic profile of the full robot to a given task. Energy saving is improved by 10% through continuous duty factor adjustment with a 192% increase in power maximization. Moreover, adjusting the robot's velocity by the step length and associating this with gait switching, reduces the power loss by a further 10% at low-speed locomotion. However, unlike in animals, only one unique optimal operating point has been revealed, which is a disadvantage caused by the low energetic efficiency of servomotor-based hexapods. Thus, these legged robots are severely limited in their capacity to optimally adjust their locomotion to various tasks-a counter-intuitive conclusion for a supposedly versatile robot.

Importance of microphysical settings for climate forcing by stratospheric SO2 injections as modeled by SOCOL-AERv2

Abstract. Solar radiation modification by a sustained deliberate source of SO2 into the stratosphere (strat-SRM) has been proposed as an option for climate intervention. Global interactive aerosol–chemistry–climate models are often used to investigate the potential cooling efficiencies and associated side effects of hypothesized strat-SRM scenarios. A recent model intercomparison study for composition–climate models with interactive stratospheric aerosol suggests that the modeled climate response to a particular assumed injection strategy depends on the type of aerosol microphysical scheme used (e.g., modal or sectional representation) alongside host model resolution and transport. Compared to short-duration volcanic SO2 emissions, the continuous SO2 injections in strat-SRM scenarios may pose a greater challenge to the numerical implementation of microphysical processes such as nucleation, condensation, and coagulation. This study explores how changing the time steps and sequencing of microphysical processes in the sectional aerosol–chemistry–climate model SOCOL-AERv2 (40 mass bins) affects model-predicted climate and ozone layer impacts considering strat-SRM by SO2 injections of 5 and 25 Tg(S) yr−1 at 20 km altitude between 30° S and 30° N. The model experiments consider the year 2040 to be the boundary conditions for ozone-depleting substances and greenhouse gases (GHGs). We focus on the length of the microphysical time step and the call sequence of nucleation and condensation, the two competing sink processes for gaseous H2SO4. Under stratospheric background conditions, we find no effect of the microphysical setup on the simulated aerosol properties. However, at the high sulfur loadings reached in the scenarios injecting 25 Tg(S) yr−1 of SO2 with a default microphysical time step of 6 min, changing the call sequence from the default “condensation first” to “nucleation first” leads to a massive increase in the number densities of particles in the nucleation mode (R&lt;0.01 µm) and a small decrease in coarse-mode particles (R&gt;1 µm). As expected, the influence of the call sequence becomes negligible when the microphysical time step is reduced to a few seconds, with the model solutions converging to a size distribution with a pronounced nucleation mode. While the main features and spatial patterns of climate forcing by SO2 injections are not strongly affected by the microphysical configuration, the absolute numbers vary considerably. For the extreme injection with 25 Tg(S) yr−1, the simulated net global radiative forcing ranges from −2.3 to −5.3 W m−2, depending on the microphysical configuration. Nucleation first shifts the size distribution towards radii better suited for solar scattering (0.3 µm &lt;R&lt; 0.4 µm), enhancing the intervention efficiency. The size distribution shift, however, generates more ultrafine aerosol particles, increasing the surface area density and resulting in 10 DU (Dobson units) less ozone (about 3 % of the total column) in the northern mid-latitudes and 20 DU less ozone (6 %) over the polar caps compared to the condensation first approach. Our results suggest that a reasonably short microphysical time step of 2 min or less must be applied to accurately capture the magnitude of the H2SO4 supersaturation resulting from SO2 injection scenarios or volcanic eruptions. Taken together, these results underscore how structural aspects of model representation of aerosol microphysical processes become important under conditions of elevated stratospheric sulfur in determining atmospheric chemistry and climate impacts.

Individuals with Methamphetamine Use Disorder Show Reduced Directed Exploration and Learning Rates Independent of an Aversive Interoceptive State Induction.

Methamphetamine Use Disorder (MUD) is associated with substantially reduced quality of life. Yet, decisions to use persist, due in part to avoidance of anticipated withdrawal states. However, the specific cognitive mechanisms underlying this decision process, and possible modulatory effects of aversive states, remain unclear. Here, 56 individuals with MUD and 58 healthy comparisons (HCs) performed a decision task, both with and without an aversive interoceptive state induction. Computational modeling measured the tendency to test beliefs about uncertain outcomes (directed exploration) and the ability to update beliefs in response to outcomes (learning rates). Compared to HCs, those with MUD exhibited less directed exploration and slower learning rates, but these differences were not affected by aversive state induction. These results suggest novel, state-independent computational mechanisms whereby individuals with MUD may have difficulties in testing beliefs about the tolerability of abstinence and in adjusting behavior in response to consequences of continued use.

Research on Delamination Damage Localization of Carbon Fiber-Reinforced Polymer Curved Plate Using Lamb Wave

Carbon fiber-reinforced polymers (CFRPs) are extensively employed in the aerospace industry due to their excellent properties. Delamination damage occurring at critical locations in CFRPs can seriously reduce the safety of in-service components. The detection and localization of delamination damage using Lamb waves hold significant potential for widespread application in non-destructive testing. However, the choice of damage localization algorithm may produce different delamination damage localization results. This research presented an IRAPID (improved reconstruction algorithm for probabilistic inspection of defects) method derived from the RAPID (reconstruction algorithm for probabilistic inspection of defects) method, aiming to improve the accuracy and reliability of delamination damage localization. Three CFRP curved plates, including a healthy curved plate and two curved plates with delamination damage sizes of Φ20 mm and Φ40 mm, were prepared in the experiment. The detection experiment of the CFRP curved plate using lead zirconate titanate (PZT) as a transducer to excite and receive Lamb waves was conducted, and the influence of excitation signal frequency on the performance of the proposed method was discussed. Under the condition of an excitation signal frequency of 220~320 kHz and a step size of 10 kHz, the accuracy of the delamination damage localization method proposed in this paper was compared with that of existing methods. The experimental results indicate that the IRAPID algorithm exhibits good stability in the localization of delamination damage across the range of frequency variations considered. The localization error of the IRAPID algorithm for delamination damage is significantly lower than that of the DaS (delay-and-sum) algorithm and the RAPID algorithm. As the size of the delamination damage increases, so does the localization error. The accuracy of delamination damage localization is lower in the X-axis direction than in the Y-axis direction. By averaging the localization results across various frequencies, we can mitigate the potential localization errors associated with single-frequency detection to a certain extent. For the localization of delamination damage, Lamb waves at multiple frequencies can be employed for detection, and the detection results at each frequency are averaged to enhance the reliability of localization.

Musculoskeletal System and Gait Characteristics in Patients with Osteogenesis imperfecta.

The aim of the study was to assess the muscoloskeletal system and spatiotemporal gait parameters of patients in three types of osteogenesis imperfecta. Retrospective observational study. The Department of Rehabilitation, Children's Memorial Health Institute in Warsaw, Poland. This study investigated individuals with various types of osteogenesis imperfecta: 33 with osteogenesis imperfecta I (aged 13.9), 16 with osteogenesis imperfecta III (aged 10.4), and 14 with osteogenesis imperfecta IV (aged, 15.8), as well as a reference group of 400 healthy individuals. The musculoskeletal assessment included: medical record review, clinical evaluation, functional tests, long bone deformity assessment via clinical and X-ray examination, and objective gait analysis with the Vicon Motion Systems (Ltd, Oxford, UK). The study revealed notable differences in clinical presentation, deformities within the musculoskeletal system, gait parameters across the various types of osteogenesis imperfecta (p < 0.001). The most affected gait parameters were: cadence, gait speed and step length. The greatest deformities of lower limbs and spine were presented in patients with osteogenesis imperfecta type III. These findings are significant for understanding gait abnormalities in osteogenesis imperfecta patients and designing customized physiotherapy programs to help them participate fully in daily life. Improvement of muscle strength is one of the key for easier engagement in activities like walking or stair-climbing.

Consensus Paper: Cerebellum and Reward.

Cerebellum is a key-structure for the modulation of motor, cognitive, social and affective functions, contributing to automatic behaviours through interactions with the cerebral cortex, basal ganglia and spinal cord. The predictive mechanisms used by the cerebellum cover not only sensorimotor functions but also reward-related tasks. Cerebellar circuits appear to encode temporal difference error and reward prediction error. From a chemical standpoint, cerebellar catecholamines modulate the rate of cerebellar-based cognitive learning, and mediate cerebellar contributions during complex behaviours. Reward processing and its associated emotions are tuned by the cerebellum which operates as a controller of adaptive homeostatic processes based on interoceptive and exteroceptive inputs. Lobules VI-VII/areas of the vermis are candidate regions for the cortico-subcortical signaling pathways associated with loss aversion and reward sensitivity, together with other nodes of the limbic circuitry. There is growing evidence that the cerebellum works as a hub of regional dysconnectivity across all mood states and that mental disorders involve the cerebellar circuitry, including mood and addiction disorders, and impaired eating behaviors where the cerebellum might be involved in longer time scales of prediction as compared to motor operations. Cerebellar patients exhibit aberrant social behaviour, showing aberrant impulsivity/compulsivity. The cerebellum is a master-piece of reward mechanisms, together with the striatum, ventral tegmental area (VTA) and prefrontal cortex (PFC). Critically, studies on reward processing reinforce our view that a fundamental role of the cerebellum is to construct internal models, perform predictions on the impact of future behaviour and compare what is predicted and what actually occurs.

Human-like Learning in Temporally Structured Environments

Natural environments have correlations at a wide range of timescales. Human cognition is tuned to this temporal structure, as seen by power laws of learning and memory, and by spacing effects whereby the intervals between repeated training data affect how long knowledge is retained. Machine learning is instead dominated by batch iid training or else relatively simple nonstationarity assumptions such as random walks or discrete task sequences. The main contributions of our work are: (1) We develop a Bayesian model formalizing the brain's inductive bias for temporal structure and show our model accounts for key features of human learning and memory. (2) We translate the model into a new gradient-based optimization technique for neural networks that endows them with human-like temporal inductive bias and improves their performance in realistic nonstationary tasks. Our technical approach is founded on Bayesian inference over 1/f noise, a statistical signature of many natural environments with long-range, power law correlations. We derive a new closed-form solution to this problem by treating the state of the environment as a sum of processes on different timescales and applying an extended Kalman filter to learn all timescales jointly. We then derive a variational approximation of this model for training neural networks, which can be used as a drop-in replacement for standard optimizers in arbitrary architectures. Our optimizer decomposes each weight in the network as a sum of subweights with different learning and decay rates and tracks their joint uncertainty. Thus knowledge becomes distributed across timescales, enabling rapid adaptation to task changes while retaining long-term knowledge and avoiding catastrophic interference. Simulations show improved performance in environments with realistic multiscale nonstationarity. Finally, we present simulations showing our model gives essentially parameter-free fits of learning, forgetting, and spacing effects in human data. We then explore the analogue of human spacing effects in a deep net trained in a structured environment where tasks recur at different rates and compare the model's behavioral properties to those of people.

Amelioration of gait and balance disorders by rosuvastatin is associated with changes in cerebrovascular reactivity in older patients with hypertensive treatment.

To investigate the effect of rosuvastatin on gait and balance disorder progression and elucidate the role of cerebrovascular reactivity (CVR) on this effect. From April 2008 to November 2010, 943 hypertensive patients aged ≥60 years were enrolled from the Shandong area of China. Patients were randomized into rosuvastatin and placebo groups. Gait, balance, CVR, fall and stroke were assessed. During an average 72 months of follow-up, the decreasing trends for step length, step speed, and Berg balance scale scores and the increasing trends for step width and chair rising test were slower in the rosuvastatin group when compared to the placebo group. The hazard ratio of incident balance impairment and falls was 0.542 [95% confidence interval (CI) 0.442-0.663] and 0.532 (95% CI 0.408-0.694), respectively, in the rosuvastatin group compared with placebo group. For CVR progression, the cerebrovascular reserve capacity and breath-holding index were increased and the pulsatility index decreased in the rosuvastatin group, while the cerebrovascular reserve capacity and breath-holding index were decreased, and pulsatility index increased in the placebo group. The changes in gait stability and balance function were independently associated with the changes in the CVR. The odds risks of balance impairment and falls were 2.178 (95% CI: 1.491-3.181) and 3.227 (95% CI: 1.634-6.373), respectively, in the patients with CVR impairment and patients without CVR impairment. Rosuvastatin ameliorated gait and balance disorder progression in older patients with hypertension. This effect might result from the improvement in the CVR. This double-blind clinical trial recruited 943 hypertensive patients aged ≥60 years who were randomly administered rosuvastatin and placebo interventions. The data indicates that rosuvastatin significantly ameliorated the progressions of gait and balance disorders in older hypertensive patients. The cerebrovascular reactivity might play an important mediating role in this amelioration.

Integrating gated recurrent unit in graph neural network to improve infectious disease prediction: an attempt.

This study focuses on enhancing the precision of epidemic time series data prediction by integrating Gated Recurrent Unit (GRU) into a Graph Neural Network (GNN), forming the GRGNN. The accuracy of the GNN (Graph Neural Network) network with introduced GRU (Gated Recurrent Units) is validated by comparing it with seven commonly used prediction methods. The GRGNN methodology involves multivariate time series prediction using a GNN (Graph Neural Network) network improved by the integration of GRU (Gated Recurrent Units). Additionally, Graphical Fourier Transform (GFT) and Discrete Fourier Transform (DFT) are introduced. GFT captures inter-sequence correlations in the spectral domain, while DFT transforms data from the time domain to the frequency domain, revealing temporal node correlations. Following GFT and DFT, outbreak data are predicted through one-dimensional convolution and gated linear regression in the frequency domain, graph convolution in the spectral domain, and GRU (Gated Recurrent Units) in the time domain. The inverse transformation of GFT and DFT is employed, and final predictions are obtained after passing through a fully connected layer. Evaluation is conducted on three datasets: the COVID-19 datasets of 38 African countries and 42 European countries from worldometers, and the chickenpox dataset of 20 Hungarian regions from Kaggle. Metrics include Average Root Mean Square Error (ARMSE) and Average Mean Absolute Error (AMAE). For African COVID-19 dataset and Hungarian Chickenpox dataset, GRGNN consistently outperforms other methods in ARMSE and AMAE across various prediction step lengths. Optimal results are achieved even at extended prediction steps, highlighting the model's robustness. GRGNN proves effective in predicting epidemic time series data with high accuracy, demonstrating its potential in epidemic surveillance and early warning applications. However, further discussions and studies are warranted to refine its application and judgment methods, emphasizing the ongoing need for exploration and research in this domain.

Comparative Analysis of the Performance of Artificial Neural Networks in Assessing the Technical Condition of Steel Ropes

Introduction. Currently, artificial neural networks (ANN) are successfully used for technical diagnostics of steel ropes. Expensive software products with an adapted neural network implementation environment, such as STATISTICA, Amygdala, MatLab Simulink, are often used for this purpose. The most affordable way to build and train an ANN, from a financial point of view, is to write your own program code using interactive libraries such as TensorFlow, PyTorch, Scikit-learn. However, such libraries are not fully adapted for building an ANN, and to use them you need to have basic programming skills. As a result, the quality of an ANN depends not only on its architecture, training data, and composition, but also on the environment in which it is built. The aim of the work was to compare the quality of the ANN, built and trained by various methods according to the criterion of test network performance, confidence levels for assessing the technical condition of the rope, as well as the complexity and speed of training. For this purpose, new software has been developed to solve the problem of assessing the technical condition of a steel rope using a combination of various rejection indicators. Materials and Methods. The basis for an ANN training was a statistical database of typical damages of steel ropes and, an expert assessment of the technical condition of steel ropes. The software was written in the Python programming language. Various methods of programming a neural network were presented: an ANN built on the basis of the STATISTICA software package and an ANN built using the interactive Scikit-learn library. Ten test samples were prepared to verify the operation of the ANN. The ANN quality was assessed based on the test network performance and confidence probabilities (activation levels of the “winning” neuron) of determining the technical condition of the rope. Results. The construction of the ANN using the interactive library Scikit-learn showed a relatively high complexity of construction and a relatively low learning rate of the ANN. Test performance of the network, with a test sample size of ten, turned out to be the same for both built ANNs. At the same time, there was a difference in the indicator of the average confidence level for determining the technical condition of a steel rope between the results of the ANN built on the basis of the STATISTICA software package and the ANN built using the Scikit-learn interactive library. Discussion and Conclusion. The results showed that the ANN built using the STATISTICA software package with the same architecture and network learning parameters had more optimal software algorithms according to the criteria of confidence probability and network learning speed in comparison with the ANN built using the free Skicit-learn library. However, the indicator of the ANN test performance turned out to be the same for both ANNs. This result justified the use of TensorFlow, PyTorch, and Skicit-learn libraries by the world's leading research and commercial centers in the field of artificial intelligence. The obtained scientific result allows us to numerically evaluate and compare the quality of an ANN having the same architecture and learning parameters, but built using different methods. This will be useful for future scientific research in the field and for selecting the optimal environment for constructing ANNs in industrial applications.

Comparison of Learning Golf Putting Skill in Different Choices: A Test of the OPTIMAL Theory

Background: The OPTIMAL theory of motor learning posits that autonomy support (AS), enhanced expectancies (EE), and an external focus of attention (EF) facilitate improved motor learning and performance. However, the applicability of this theory has not been investigated by simultaneously considering all three factors in task-relevant and irrelevant choices in learning golf skills. Objectives: The purpose of this study was to investigate the sequential implementation of AS, EE and EF factors in relevant and irrelevant choices on golf putting tasks. Methods: During a pre-test under neutral conditions and acquisition phase (six blocks), thirty-six participants attempted to putt golf balls into a hole cup placed in the center of four circles with different diameters on the green. The participants were divided into two groups for optimized practices: The optimal group with task-relevant choices and the optimal group with task-irrelevant choices, both groups performed the optimized factors (EE, AS, and EF) successively and identically to each other. After 24 hours, learning was measured by a retention test. Results: The optimized group with relevant choices performed better than the optimized group with irrelevant choices in the autonomy variable. In the retention test, the aiming accuracy score of both groups increased compared to the pre-test, but no significant difference was observed in the learning rate between the two optimized groups. Conclusions: The present findings demonstrate that the key factors of OPTIMAL theory can be further benefited by task-related choices during practice for beginners' learning.

An analysis of high order FEM and IGA for explicit dynamics: Mass lumping and immersed boundaries

SummaryWe investigate the behavior of different shape functions for the discretization of hyperbolic problems. In particular, we consider classical Lagrange polynomials and B‐splines. The studies focus on the performance of the these functions as a spatial discretization approach combined with an explicit time marching scheme. In this regard, a major concern is the maximum eigenvalue that imposes restrictions on the critical time step size and suitable lumping techniques that yield a diagonal mass matrix. The accuracy of the discretization methods is assessed in an asymptotic manner in terms of the convergence of eigenvalues and eigenvectors. Further, the global accuracy is investigated in terms of the full spectrum. The results show that B‐spline discretization with a consistent mass matrix are more accurate than those based on Lagrange shape functions, which holds true in the boundary‐fitted as well as in the immersed setting. On the other hand, Lagrange shape functions are more robust with respect to standard lumping techniques, which cannot be directly applied for B‐splines without loss of accuracy. In general, we observe that none of the standard lumping schemes yields optimal results for B‐splines, even in the boundary‐fitted setting. For the immersed setting, also Lagrange shape functions show a drop in accuracy which depends on the position of the boundary that cuts the element. Several remedies are considered in order to overcome these issues, including interpolatory B‐spline bases as well as eigenvalue stabilization methods. While accuracy and stability can be improved using these remedies, we conclude from our study that it is still an open question, how to design a discretization method that achieves large critical time step sizes in combination with a diagonal mass matrix and high accuracy in the immersed setting. We note that these considerations primarily relate to linear structural dynamics applications, such as for example, structural acoustics. In nonlinear problems, such as automotive crash dynamics, other considerations predominate. An example of a one‐dimensional elastic‐plastic bar impacting a rigid wall is illustrative.

An enhanced sparrow search swarm optimizer via multi-strategies for high-dimensional optimization problems

With the development of science and technology, high-dimensional global optimization problems have become increasingly prevalent for scientific research and engineering, such as gene recognition, vehicle routing, job scheduling, and network topology. These problems are typically characterized by enormous and complex search spaces and numerous local minima, making it challenging to find the global optimal solution with limited computing resources. This paper introduces an enhanced sparrow search swarm optimizer (ESSSO) based on a bio-mimetic method. The ESSSO employs an adaptive sinusoidal walk strategy based on the von Mises distribution, a learning strategy utilizing roulette wheel selection, a two-stage evolution strategy, and a selection mutation strategy to address these issues. The proposed sinusoidal walk strategy, grounded in the von Mises distribution, supports a balanced evolutionary search. This mechanism disperses the individuals in a swarm in various directions based on a circular normal distribution. It then leads the search and adaptively adjusts their step sizes according to the size of the search domain during each generation of evolution. The learning strategy, based on roulette wheel selection, enhances the diversity of the population and improves the global search capability of the algorithm during the initial iterations. The two-stage evolution strategy involves a sine-learning mechanism based on the von Mises distribution and an adaptive mutation mechanism. The former is designed to boost the convergence speed of ESSSO, while the latter prevents ESSSO from getting trapped in a local optimum. Additionally, the selection mutation strategy further enhances convergence speed while maintaining population diversity. These strategies promote exploration in the early stages of evolution and exploitation in the later stages, enabling a well-balanced search for optimal solutions. We conducted comprehensive experiments two standard benchmark sets (i.e., CEC2010 and CEC2013), antenna array optimization, feature selection, and four engineering design problems. The results indicate that ESSSO outperforms ten comparison algorithms, especially in scenarios with smaller population sizes. This confirms its effectiveness in high-dimensional global optimization tasks and demonstrates that it can achieve better results with less computational resource consumption.

A novel framework using 3D-CNN and BiLSTM model with dynamic learning rate scheduler for visual speech recognition

A novel framework using 3D-CNN and BiLSTM model with dynamic learning rate scheduler for visual speech recognition

Dynamical stability and chaos in artificial neural network trajectories along training

The process of training an artificial neural network involves iteratively adapting its parameters so as to minimize the error of the network’s prediction, when confronted with a learning task. This iterative change can be naturally interpreted as a trajectory in network space–a time series of networks–and thus the training algorithm (e.g., gradient descent optimization of a suitable loss function) can be interpreted as a dynamical system in graph space. In order to illustrate this interpretation, here we study the dynamical properties of this process by analyzing through this lens the network trajectories of a shallow neural network, and its evolution through learning a simple classification task. We systematically consider different ranges of the learning rate and explore both the dynamical and orbital stability of the resulting network trajectories, finding hints of regular and chaotic behavior depending on the learning rate regime. Our findings are put in contrast to common wisdom on convergence properties of neural networks and dynamical systems theory. This work also contributes to the cross-fertilization of ideas between dynamical systems theory, network theory and machine learning.

The Relationship between Take-Off Parameters and Relative Vertical Momentum of Free Limbs at the Take-Off in Hurdle Clearance

This study aimed to clarify the kinetics of the relative vertical momentum to the proximal joint of each free limb and their contribution to the increase in the centre of mass height at the take-off of hurdle clearance, as well as their relationship with take-off variables. Thirteen male hurdlers cleared one hurdle at the height of their centre of mass, and their attempts were filmed using six high-speed cameras. The hurdle height was 96.54 ± 2.63 cm (55.35 ± 0.29% of body height). The approach distance was set at 15 m and adjusted by each hurdler in the range of 10–50 cm so as not to involve any noticeable step length adjustment before the take-off. The combined free limb relative vertical momentum tended to increase until mid-support and was maintained until the toe off. The smaller the whole-body vertical momentum at the toe off and the increase in relative vertical momentum of the lead leg during the take-off, the higher the take-off velocity, the shorter the support time, and the smaller the deceleration. The higher the relative vertical momentum of the forward swing arm during touchdown and the smaller the relative vertical momentum increase of the combined free limb and the forward swing arm during the take-off, the smaller the deceleration. In conclusion, hurdlers should reduce the increase in whole-body vertical momentum at the take-off by suppressing the increase in relative vertical momentum of the lead leg and the forward swing arm.