Dynamic Machine Research Articles

Dynamic machine vision with retinomorphic photomemristor-reservoir computing

Dynamic machine vision requires recognizing the past and predicting the future of a moving object based on present vision. Current machine vision systems accomplish this by processing numerous image frames or using complex algorithms. Here, we report motion recognition and prediction in recurrent photomemristor networks. In our system, a retinomorphic photomemristor array, working as dynamic vision reservoir, embeds past motion frames as hidden states into the present frame through inherent dynamic memory. The informative present frame facilitates accurate recognition of past and prediction of future motions with machine learning algorithms. This in-sensor motion processing capability eliminates redundant data flows and promotes real-time perception of moving objects for dynamic machine vision.

Acoustic signature analysis of a bladeless blower

There has always been a concern about acoustic emissions in traditional machinery. For example, in the transportation and power generation sectors, the market demands efficient, compact, lightweight, and low-cost devices with very low noise levels. In such a challenging context, as opposed to conventional turbomachinery, Tesla type bladeless fluid machine has the advantage of reducing the overall noise emissions level.The originality of this paper lies in its experimental acoustic signature characterization of a Tesla type bladeless blower, which has never been done before, nor is there any comparison to conventional blowers. A tailored experimental setup is described in order to evaluate the acoustic behavior of Tesla blower prototypes at different rotational speeds and distances. To achieve an accurate acoustic characterization, several measurement methods, as well as post-processing techniques, have been evaluated and compared.The results of this study have been compared with those obtained by conventional blowers, while considering the same tip speeds at different distances. A Tesla blower does not exhibit the blade passage phenomenon, as opposed to a conventional blower. A significant contribution of acoustic energy can thus be avoided which is due to its characteristic frequency and its harmonic components, which instead are significant in conventional turbomachinery overall acoustic noise.As a result of this innovative investigation, it has been determined that the air Tesla blower is significantly quieter than conventional dynamic machines, thus paving the way to novel applications for bladeless machinery, especially in those cases aimed at achieving a heightened level of acoustic comfort.

A novel bearing fault diagnosis method based on multi-scale transfer symbolic dynamic entropy and support vector machine

In view of the problem that the generalization ability of traditional data-driven fault diagnosis model declines or even fails in mechanical system diagnosis, a fault diagnosis method based on multi-scale transfer symbolic dynamic entropy and support vector machine is proposed based on the idea of transfer learning. Firstly, multi-scale symbolic dynamic entropy is used to extract fault features from measured vibration signals. And then a feature projection technique based on transfer learning is proposed, which reduces the data distribution difference. Secondly, the parameters of the multi-scale transfer symbol dynamic entropy method are optimized to improve the final fault identification rate. Then, the support vector machine can implement the fault identification. Finally, through the test of bearing fault experimental signals, the rolling bearing diagnosis method based on multi-scale transfer symbol dynamic entropy can effectively improve the generalization ability of data-driven model and realize accurate identification of different fault types of rolling bearing under a small number of samples.

Effect of Handedness on Learned Controllers and Sensorimotor Noise During Trajectory-Tracking.

In human-in-the-loop control systems, operators can learn to manually control dynamic machines with either hand using a combination of reactive (feedback) and predictive (feedforward) control. This article studies the effect of handedness on learned controllers and performance during a trajectory-tracking task. In an experiment with 18 participants, subjects perform an assay of unimanual trajectory-tracking and disturbance-rejection tasks through second-order machine dynamics, first with one hand then the other. To assess how hand preference (or dominance) affects learned controllers, we extend, validate, and apply a nonparametric modeling method to estimate the concurrent feedback and feedforward controllers. We find that performance improves because feedback adapts, regardless of the hand used. We do not detect statistically significant differences in performance or learned controllers between hands. Adaptation to reject disturbances arising exogenously (i.e., applied by the experimenter) and endogenously (i.e., generated by sensorimotor noise) explains observed performance improvements.

Robust current control of IPMSM drives under uncertain and varying inductances

Interior permanent magnet synchronous machines (IPMSMs) are gaining popularity in e-mobility applications due to their wide constant power speed range when compared with other electric motors. The high-performance of IPMSMs is due to the pronounced degree of magnetic saliency, which originates from embedding their magnets inside the rotor. Yet, this feature complicates their control since it increases nonlinear phenomena such as cross-magnetization and saturation. This results in pronounced variations of the machine inductances, which are highly dependent on the machine current. This behaviour is of particular relevance in e-mobility applications, where a highly dynamic operation of the machine is required. In contrast to most existing current control approaches, where precise inductance knowledge is required, we address this challenge by deriving a controller that exploits the machine dynamics to achieve exponential current tracking in the presence of unknown and varying inductances. By interpreting the flux linkage as a modelled disturbance, we propose a dynamic compensator based on the internal model principle. As a by-product, the method provides an estimate of the flux linkage. The gain in performance compared to the standard PI controller is illustrated in simulation.

Global analysis and prediction scenario of infectious outbreaks by recurrent dynamic model and machine learning models: A case study on COVID-19

It is essential to evaluate patient outcomes at an early stage when dealing with a pandemic to provide optimal clinical care and resource management. Many methods have been proposed to provide a roadmap against different pandemics, including the recent pandemic disease COVID-19. Due to recurrent epidemic waves of COVID-19, which have been observed in many countries, mathematical modeling and forecasting of COVID-19 are still necessary as long as the world continues to battle against the pandemic. Modeling may aid in determining which interventions to try or predict future growth patterns. In this article, we design a combined approach for analyzing any pandemic in two separate parts. In the first part of the paper, we develop a recurrent SEIRS compartmental model to predict recurrent outbreak patterns of diseases. Due to its time-varying parameters, our model is able to reflect the dynamics of infectious diseases, and to measure the effectiveness of the restrictive measures. We discuss the stable solutions of the corresponding autonomous system with frozen parameters. We focus on the regime shifts and tipping points; then we investigate tipping phenomena due to parameter drifts in our time-varying parameters model that exhibits a bifurcation in the frozen-in case. Furthermore, we propose an optimal numerical design for estimating the system’s parameters. In the second part, we introduce machine learning models to strengthen the methodology of our paper in data analysis, particularly for prediction scenarios. We use MLP, RBF, LSTM, ANFIS, and GRNN for training and evaluation of COVID-19. Then, we compare the results with the recurrent dynamical system in the fitting process and prediction scenario. We also confirm results by implementing our methods on the released data on COVID-19 by WHO for Italy, Germany, Iran, and South Africa between 1/22/2020 and 7/24/2021, when people were engaged with different variants including Alpha, Beta, Gamma, and Delta. The results of this article show that the dynamic model is adequate for long-term analysis and data fitting, as well as obtaining parameters affecting the epidemic. However, it is ineffective in providing a long-term forecast. In contrast machine learning methods effectively provide disease prediction, although they do not provide analysis such as dynamic models. Finally, some metrics, including RMSE, R-Squared, and accuracy, are used to evaluate the machine learning models. These metrics confirm that ANFIS and RBF perform better than other methods in training and testing zones.

Characterization of the mechanism of bile salt hydrolase substrate specificity by experimental and computational analyses

Bile salt hydrolases (BSHs) are currently being investigated as target enzymes for metabolic regulators in humans and as growth promoters in farm animals. Understanding structural features underlying substrate specificity is necessary for inhibitor design. Here, we used a multidisciplinary workflow including mass spectrometry, mutagenesis, molecular dynamic simulations, machine learning, and crystallography to demonstrate substrate specificity in Lactobacillus salivarius BSH, the most abundant enzyme in human and farm animal intestines. We show the preference of substrates with a taurine head and a dehydroxylated sterol ring for hydrolysis. A regression model that correlates the relative rates of hydrolysis of various substrates in various enzyme mutants with the residue-substrate interaction energies guided the identification of structural determinants of substrate binding and specificity. In addition, we found T208 from another BSH protomer regulating the hydrolysis. The designed workflow can be used for fast and comprehensive characterization of enzymes with a broad range of substrates.

Collaboration methods for ensembles of dispatching rules for the dynamic unrelated machines environment

Dynamic scheduling represents an important combinatorial optimisation problem that often appears in the real world. The difficulty in solving these problems arises from their dynamic nature, which limits the applicability of improvement based metaheuristics. Dynamic problems are usually solved using dispatching rules (DRs), which iteratively construct the schedule. Recently, such heuristics have been constructed using various hyperheuristic methods, most notably genetic programming. Although automatically designed DRs achieve good performance, it is still very difficult to design a single DR that would perform a good decision at every decision point. As a remedy, DRs were combined into ensembles to improve their performance. For that purpose it is required to define how ensembles are constructed and how DRs in the ensemble collaborate. This paper proposes a novel ensemble collaboration method based on a similar method applied for static scheduling problems and adapts it for dynamic problems. The goal is to obtain a collaboration method that produces better results than standard collaboration methods. Additionally, the paper investigates the application of novel ensemble construction methods for dynamic scheduling. The proposed methods are validated on dynamic unrelated machines scheduling problem and compared with existing ensemble construction and collaboration methods. The obtained results demonstrate that the proposed collaboration method performs better than standard ones. Further analyses provide additional insights into the proposed methods and outline several potential research directions in the area of hyper-heuristic ensemble construction.

Modeling and optimizing the maintenance schedule using dynamic quality and machine condition monitors in an unreliable single production system

This work attempted to propose an integrated approach for the joint control of production, quality and maintenance where a single machine would produce one type of product. The machine would be subject to degradation process that would produce non-conform units. Based on the simultaneous operations of assessment of the machine degradation and that of the product quality, decisions to conduct maintenance activities would be proposed to improve the machine reliability. The strategy adopted in this work was to establish a link between the information from the quality indicator and the level of the machine degradation using a coupling function h(x,y) to generate a maintenance schedule. The ultimate objective of this work was to minimize the average total cost per unit of time in function of cost of maintenance and cost of production. For this purpose, a mathematical model was developed and numerical examples were implemented where optimal values of decision variables were computed.

Multi-Class Skin Cancer Classification Using a Hybrid Dynamic Salp Swarm Algorithm and Weighted Extreme Learning Machines with Transfer Learning

Multi-Class Skin Cancer Classification Using a Hybrid Dynamic Salp Swarm Algorithm and Weighted Extreme Learning Machines with Transfer Learning

Single-molecule tracking (SMT): a window into live-cell transcription biochemistry.

How molecules interact governs how they move. Single-molecule tracking (SMT) thus provides a unique window into the dynamic interactions of biomolecules within live cells. Using transcription regulation as a case study, we describe how SMT works, what it can tell us about molecular biology, and how it has changed our perspective on the inner workings of the nucleus. We also describe what SMT cannot yet tell us and how new technical advances seek to overcome its limitations. This ongoing progress will be imperative to address outstanding questions about how dynamic molecular machines function in live cells.

Investigation of the cutting-edge radius size effect on dynamic forces in micro end milling of brass260

Rapid increases in the number of industries that need a sub-micron surface finish on 3D micro/meso technological innovation parts are being seen. Due to the high rotating speeds of activities, chatter behaviour also occurs during micro-cutting. Changes in bearing stiffness, gyroscopic effects, and misalignment are only some of the factors that need to be considered while studying the spindle's dynamics. By regulating the movement of the cutting edge, the cutting action may be made more consistent. The structure of the tools also affects the machine's dynamics. Since the modes of different machine-tool components interact with one another, the dynamics of the tool tip are affected. In this research, we consider a model of a micro spindle system to enhance machining accuracy. Timoshenko beam theory is used to account for shear deformation and rotational inertia effects due to the short and thick beam-type designs of each component of the micro tool. A detailed dynamical model of the moving micro end mill is created using an extended form of Hamilton's Principle. A two-degree-of-freedom model of the micro-milling process considers the modal dynamic properties of the tool-holder-spindle assembly and the micro-milling cutting forces. A three-dimensional finite element model of the system is built, and its dynamic response is gathered, to validate the full order finite element approach. Cutting force coefficients are then modelled as a function of instantaneous uncut chip thickness, which is principally affected by the tool edge radius and feed per tooth at different radial immersion depths.

Adaptive secure malware efficient machine learning algorithm for healthcare data

AbstractMalware software now encrypts the data of Internet of Things (IoT) enabled fog nodes, preventing the victim from accessing it unless they pay a ransom to the attacker. The ransom injunction is constantly accompanied by a deadline. These days, ransomware attacks are too common on IoT healthcare devices. On the other hand, IoT‐based heartbeat digital healthcare applications have been steadily increasing in popularity. These applications make a lot of data, which they send to the fog cloud to be processed further. In healthcare networks, it is critical to examine healthcare data for malicious intent. The malware is a peace code with polymorphic and metamorphic attack forms. Existing malware analysis techniques did not find malware in the content‐aware heartbeat data. The Adaptive Malware Analysis Dynamic Machine Learning (AMDML) algorithm for content‐aware heartbeat data in fog cloud computing is described in this article. Based on heartbeat data from health records, an adaptive method can train both pre‐ and post‐train malware models. AMDML is based on a rule called ‘federated learning,’ which says that malware analysis models are made at both the local fog node and the remote cloud to meet the performance workload safely. The simulation results show that AMDML outperforms machine learning malware analysis models in terms of accuracy by 60%, delay by 50%, and detection of original heartbeat data by 66% compared to existing malware analysis schemes.

Determination of dynamic and traction-energy indicators of all-wheel-drive traction-transport machine

The results of a study of dynamic and traction-energy indicators of an all-wheel drive wheeled traction-transport machine are presented. A diagram of a dynamic transmission model for an all-wheel drive wheeled traction and transport machine and a system of equations for the transmission dynamics in the Cauchy form have been compiled. This made it possible to determine the dependences of the angular speeds of rotation of the transmission elements of the traction-transport machine, the dependences of the torques and the dependences of the contacting traction forces on the wheels on time. The method allows to determine the optimal transmission parameters, differential designs and gear ratios to improve the traction and coupling and fuel-economic performance of the machine. It has been established that the angular speeds of rotation of the front wheels of the traction-transport machine 1.29 rad/s, 1.27 rad/s are higher than the angular speeds of rotation of the rear wheels 1.24 rad/s, 1.25 rad/s, which leads to the appearance of a kinematic discrepancies and additional energy losses. The torques of the front drive wheels are 6972 Nm, the rear drive wheels are 4622 Nm. The contacting traction forces on the front wheels of the machine are 5478 N after the end of the acceleration of the machine, on the rear wheels – 3473 N. Experimental studies were carried out on the example of an all-wheel drive wheeled tractor with an articulated frame to validate the method for assessing the dynamics of the traction-transport transmission. The difference between the values of the angular speeds of rotation of the wheels and the tangential traction forces on the wheels, determined theoretically and obtained during experimental studies, is 2 %. The developed method for assessing the transmission dynamics of an all-wheel drive traction and transport machine should be considered valid. The method proposed in the paper can be used to assess the dynamics of wheeled machines

Dynamic NIR Fluorescence Imaging and Machine Learning Framework for Stratifying High vs. Low Notch-Dll4 Expressing Host Microenvironment in Triple-Negative Breast Cancer.

Delta like canonical notch ligand 4 (Dll4) expression levels in tumors are known to affect the efficacy of cancer therapies. This study aimed to develop a model to predict Dll4 expression levels in tumors using dynamic enhanced near-infrared (NIR) imaging with indocyanine green (ICG). Two rat-based consomic xenograft (CXM) strains of breast cancer with different Dll4 expression levels and eight congenic xenograft strains were studied. Principal component analysis (PCA) was used to visualize and segment tumors, and modified PCA techniques identified and analyzed tumor and normal regions of interest (ROIs). The average NIR intensity for each ROI was calculated from pixel brightness at each time interval, yielding easily interpretable features including the slope of initial ICG uptake, time to peak perfusion, and rate of ICG intensity change after reaching half-maximum intensity. Machine learning algorithms were applied to select discriminative features for classification, and model performance was evaluated with a confusion matrix, receiver operating characteristic curve, and area under the curve. The selected machine learning methods accurately identified host Dll4 expression alterations with sensitivity and specificity above 90%. This may enable stratification of patients for Dll4 targeted therapies. NIR imaging with ICG can noninvasively assess Dll4 expression levels in tumors and aid in effective decision making for cancer therapy.

Mechanism of the week: An approach for improving student engagement, interest, and understanding of course material

A program called mechanism of the week was created for use within a third-year mechanical engineering course, machine dynamics, to improve the student engagement and interest in the course content as well as to help aid student understanding of the course concepts. This program was implemented across three semesters, including some virtual sessions, and evaluated using a student feedback survey. The results of this survey indicated strong student support of the program. In addition to this particular application, the general principles of this program could be applied to other courses and disciplines to gain the same overall benefit.

Energy-aware QoS-based dynamic virtual machine consolidation approach based on RL and ANN

Energy-aware QoS-based dynamic virtual machine consolidation approach based on RL and ANN

Dynamic Machine Learning Global Optimization Algorithm and Its Application to Aerodynamics

A dynamic extreme gradient boosting (XGBoost) and MaxLIPO trust region parallel global optimization algorithm is proposed in this paper, and it is applied to the turbomachinery blade aerodynamic optimization coupled with an in-house graphics processing unit (GPU) heterogeneous accelerated compressible flow solver, AeroWhale. The algorithm combines an accurate machine learning regression model, an efficient nongradient optimization method with no hyperparameters, a dynamic update regression strategy, and double convergence criteria to achieve high optimization accuracy and efficiency. The optimization results on the test function indicate that the number of objective function calls is less than 2% of that required by a traditional genetic algorithm, which greatly reduces the optimization time. The dynamic XGBoost model ensures that the regression model accuracy near the optimum is relatively high, which is attributed to the update strategy. The error between the optimal value identified by the proposed algorithm and the theoretical value is only 0.52% after several objective function calls. Finally, the aerodynamic optimization algorithm is applied to the LS89 high-pressure turbine, and the total pressure loss is reduced by 13.16%. The sensitivity of each optimization feature to the objective function is determined, showing that the blade suction surface control point near the trailing edge has the greatest impact on aerodynamic performance.

Long-term mechanical performance of 3D printed thermoplastics in seawater environments

This study investigates the effects of 3D-printed thermoplastics' long-term exposure to ocean water. The work focuses on seawater diffusivity rates and mechanical performance to discuss material candidates for long-term undersea structures. Several commonly available thermoplastics were printed using Fused Deposition Modeling and then submerged in an ocean water solution for a prolonged period at elevated temperatures. The thermoplastics studied included Nylon, ABS, PLA, PCTG, PETG, and ASA. The different temperatures corresponded with distinct accelerated rates of water mass diffusion. The solution's water level, temperature, and salinity were monitored and maintained during the accelerated weathering process. The total mass of each specimen and its elastic modulus were recorded daily using a precision scale and a Dynamic Mechanical Analyzer machine. The change in mass was used to calculate the activation energy of each material system by defining the degradation to have an Arrhenius behavior and using a new polynomial fit-based technique. The new polynomial fit-based method was robust compared to a one-dimensional system that uses Fick's law of diffusion. Results also showed that the 3D printed structures, regardless of their composition, are susceptible to extremely high diffusivity rates compared to their bulk material behavior. Lastly, mechanical properties decreased for all polymers, and the stiffness of the aged specimens was proportionally affected by the change in mass.

Light-Driven Aqueous Dissipative Pseudorotaxanes with Tunable Fluorescence Enabling Deformable Nano-Assemblies

Developing an artificial dynamic nanoscale molecular machine that dissipatively self-assembles far from equilibrium is fundamentally important but is significantly challenging. Herein, we report dissipatively self-assembling light-activated convertible pseudorotaxanes (PRs) that show tunable fluorescence and enable deformable nano-assemblies. A pyridinium-conjugated sulfonato-merocyanine derivative (EPMEH) and cucurbit[8]uril (CB[8]) form the 2EPMEH ⊂ CB[8] [3]PR in a 2:1 stoichiometry, which phototransforms into a transient spiropyran containing 1:1 EPSP ⊂ CB[8] [2]PR when exposed to light. The transient [2]PR thermally relaxes (reversibly) to the [3]PR in the dark accompanied by periodic fluorescence changes that include near-infrared emission. Moreover, octahedral and spherical nanoparticles are formed through the dissipative self-assembly of the two PRs, and the Golgi apparatus is dynamically imaged using fluorescent dissipative nano-assemblies.