Real-time Experimental Data Research Articles

Investigation of Thermal Deformation Behavior in Boron Nitride-Reinforced Magnesium Alloy Using Constitutive and Machine Learning Models

Accurate flow stress prediction is vital for optimizing the manufacturing of lightweight materials under high-temperature conditions. In this study, a boron nitride (BN)-reinforced AZ80 magnesium composite was subjected to hot compression tests at temperatures of 300–400 °C and strain rates ranging from 0.01 to 10 s−1. A data-driven Support Vector Regression (SVR) model was developed to predict flow stress based on temperature, strain rate, and strain. Trained on experimental data, the SVR model demonstrated high predictive accuracy, as evidenced by a low mean squared error (MSE), a coefficient of determination (R2) close to unity, and a minimal average absolute relative error (AARE). Sensitivity analysis revealed that strain rate and temperature exerted the greatest influence on flow stress. By integrating machine learning with experimental observations, this framework enables efficient optimization of thermal deformation, supporting data-driven decision-making in forming processes. The results underscore the potential of combining advanced computational models with real-time experimental data to enhance manufacturing efficiency and improve process control in next-generation lightweight alloys.

Performance analysis of 500kW solar PV plant based on machine learning algorithms

ABSTRACT This research presents a comprehensive performance analysis of a 500 kW solar photovoltaic (PV) plant, utilizing machine learning (ML) algorithms, installed at PSNA College of Engineering and Technology, Dindigul, India. Real-time experimental data, collected from November 2021 to October 2022, was used to enhance the operational efficiency of the solar PV system. The study focuses on forecasting the output of the 500 kW plant by employing six machine learning algorithms, with a comparative analysis conducted to evaluate the predictive capabilities of each model. Kernel Ridge Regression (KRR) demonstrated superior performance among the algorithms tested, producing results closely aligned with the actual data. Key performance metrics were assessed, including forecasted output power, final yield, system efficiency, performance ratio, and capacity factor. Notably, the system’s efficiency increased from 12% to 13.6%.These findings provide critical insights for optimizing the operational performance of solar PV plants, offering valuable contributions to improving energy management and sustainability.

System Identification for Robust Control of an Electrode Positioning System of an Industrial Electric Arc Melting Furnace

Through system identification for robust control methods and utilizing real-time experimental field data, a comprehensive mathematical model is derived that represents the dynamic performance of a single electrode positioning system (EPS) in an industrial electric arc melting furnace (EAF). This EPS is characterized by large, time-varying dynamic parameters, which fluctuate based on operating conditions, specifically as the electrode weight changes within its operational range. The system identification methodology for robust control is developed in four main steps, progressing from experimental design to model validation. This approach yields a nominal model of the actual system and provides a trustworthy estimate of the region of uncertainty of the model, bounded by models of the real system under maximum and minimum electrode weight conditions (limit operating models). The methodology generates three fourth-order time-delay models using an ARMAX structure. The results are promising, as system identification for robust control enables the derivation of mathematical models specifically tailored for designing robust controllers. These controllers significantly enhance the EPS control system’s performance and substantially reduce energy consumption and environmental emissions.

Development of Semi-Empirical and Machine Learning Models for Photoelectrochemical Cells

We introduce a theoretical model for the photocurrent-voltage (I-V) characteristics designed to elucidate the interfacial phenomena in photoelectrochemical cells (PECs). This model investigates the sources of voltage losses and the distribution of photocurrent across the semiconductor–electrolyte interface (SEI). It calculates the whole exchange current parameter to derive cell polarization data at the SEI and visualizes the potential drop across n-type cells. The I-V model’s simulation outcomes are utilized to distinguish between the impacts of bulk recombination and space charge region (SCR) recombination within semiconductor cells. Furthermore, we develop an advanced deep neural network model to analyze the electron–hole transfer dynamics using the I-V characteristic curve. The model’s robustness is evaluated and validated with real-time experimental data, demonstrating a high degree of concordance with observed results.

A Novel Method for Estimating the Thermal Performance of Multi-Block Wall Systems Using Thermal Impedance Z-Value under Transient Uncontrolled Heat Transfer Conditions

Climate change is one of the biggest challenges today. An increasing population accelerates the construction of concrete houses and the use of air conditioners, thereby leading to an increase in energy consumption. When the walls of buildings are well-designed and insulated, energy consumption can be reduced. Therefore, it is important to measure the thermal performance of wall systems accurately. The existing traditional methods of measuring R- and U-values provide acceptable solutions for steady-state controlled, uncontrolled or transient state-controlled conditions. However, a need to develop a novel approach for transient state-uncontrolled realistic conditions has been identified. The present study involves both experimental and numerical investigations. An in situ model room with dimensions of 1.60 m × 1.73 m × 1.50 m was built for the experimental work, and a series of experiments were conducted. For numerical work, two models using Ansys Fluent 2021/2022 and MATLAB Simulink 2021/2022 were developed. The real-time experimental data were fed into numerical models to predict the thermal behavior of the wall system. The results include the evaluation of a concept called ‘Time-Lag’ for all three models. ‘Time-Lag’ is the time taken for the heat energy to flow across the wall system. The Time-Lag for the experimental model was 8 h 45 min, while for MATLAB and Ansys models, it was 8 h 22 min. (average) and 7 h 30 min, respectively. Minor variations validate the accuracy of the numerical models. Further, a novel method using a new parameter in building systems called ‘thermal impedance Z-value’ was developed to estimate the real-time thermal performance of walls using MATLAB Simulink. The Z-value measures the ability of a wall system to resist the flow of heat (thermal resistance, R-value) combined with its ability to store heat energy (thermal capacitance, Cth-value). It is evaluated for steady-state and dynamic (transient) systems. For the steady-state system, the Z-values on the outer and inner walls were 18.2683 K/W and 18.6761 K/W, respectively with a minor difference of 0.4078 K/W at the end of 72 h. For the dynamic system, the Z-value did not reach a constant value and fluctuated in a particular pattern during 24 h of the solar cycle with average values of 3.2969 K/W on the outer and 1.2886 K/W on the inner walls at the end of 72 h, thus presenting more accurate and realistic thermal performance results of a wall system.

Digital twin for subcritical extraction driven by data-mechanism fusion

Abstract Addressing the key issue of low automation in the production process due to the complex associations involved in sub-critical extraction, this study focuses on sub-critical extraction as the subject matter. It integrates data models with mechanism models to construct a multidimensional digital twin model and conduct systematic research. By combining simulation methods with real-time experimental data, the study establishes multilevel machine characteristics of data, behavior, and mechanisms, and identifies data patterns. From the perspectives of the physical production line, virtual production line, and data transmission, it builds an interactive mapping platform for multi-source data and carries out corresponding application research.

A 3-DOF caudal fin for precise maneuvering of thunniform-inspired unmanned underwater vehicles

Unmanned underwater vehicles (UUVs) will see increased use in scientific research, military operations and maintenance of industrial infrastructure. Many of these applications require that the vehicle possess a long range while retaining precise maneuvering or station-keeping capabilities. Both current UUVs and biological swimmers, often considered the basis for the next generation of UUVs, face a trade-off between the two characteristics. Here, we introduce a novel hybrid propeller concept that enables thunniform-inspired vehicles, which imitate nature’s most efficient swimmers, to also achieve high maneuverability. The propeller can produce enhanced three-dimensional kinematics of the caudal fin. An optimization procedure based on real-time experimental data is used to obtain the best kinematics to maneuver the vehicle, and a 4-step strategy is uncovered that results in a 49% increase in maneuverability with respect to conventional 2-D kinematics. The proposed mechanism is shown to be effective for a wide range of fin geometries and stiffness values.

Phase-field modeling of stored-energy-driven grain growth with intra-granular variation in dislocation density

We present a phase-field (PF) model to simulate the microstructure evolution occurring in polycrystalline materials with a variation in the intra-granular dislocation density. The model accounts for two mechanisms that lead to the grain boundary migration: the driving force due to capillarity and that due to the stored energy arising from a spatially varying dislocation density. In addition to the order parameters that distinguish regions occupied by different grains, we introduce dislocation density fields that describe spatial variation of the dislocation density. We assume that the dislocation density decays as a function of the distance the grain boundary has migrated. To demonstrate and parameterize the model, we simulate microstructure evolution in two dimensions, for which the initial microstructure is based on real-time experimental data. Additionally, we applied the model to study the effect of a cyclic heat treatment (CHT) on the microstructure evolution. Specifically, we simulated stored-energy-driven grain growth during three thermal cycles, as well as grain growth without stored energy that serves as a baseline for comparison. We showed that the microstructure evolution proceeded much faster when the stored energy was considered. A non-self-similar evolution was observed in this case, while a nearly self-similar evolution was found when the microstructure evolution is driven solely by capillarity. These results suggest a possible mechanism for the initiation of abnormal grain growth during CHT. Finally, we demonstrate an integrated experimental-computational workflow that utilizes the experimental measurements to inform the PF model and its parameterization, which provides a foundation for the development of future simulation tools capable of quantitative prediction of microstructure evolution during non-isothermal heat treatment.

Hybrid Cathode Lithium Battery Discharge Simulation for Implantable Cardioverter Defibrillators Using a Coupled Electro-Thermal Dynamic Model.

This paper explores the effect of the load imposed by implantable cardioverter defibrillators (ICDs) on their lithium battery power sources longevity using a simulation approach that incorporates a coupled electro-thermal dynamic model. ICDs are one of the effective treatments available to significantly improve survival of patients with fatal arrhythmia (abnormal heart rhythm) disorders. Using a lithium battery power source, this life-saving device sends electrical shocks or pulses to regulate the heartbeat. The service life and reliability of an ICD is primarily expressed by its battery's lifespan and performance. In this paper we investigate the terminal voltage, depth of discharge and temperature dynamics of the implantable lithium battery with a combined cathode material, namely silver vanadium oxide and carbon-monofluoride (Li/SVO-CFx). Modeling the implantable batteries characteristics is a well-established topic in literature; however, to the best of the author's knowledge, the impact of the high-energy shocks (defibrillation) and low-energy device power supply (housekeeping) on the ICD's battery operation is relatively less-explored. Our analysis reveals that the battery terminal voltage is primarily influenced by the continuous low-level housekeeping discharge current within the range of micro amps, rather than the intermittent high-level demand of defibrillation current within the range of several amperes. A Monte Carlo simulation and model prediction comparison with real-time experimental data from literature were used to assess the accuracy and applicability of the model. The results can be used to improve the device battery design, control and operation, thus extending the service life in patients and reducing the need for invasive replacement surgery.

Experimental Advances in the Real-Time Recording of Cross-Linking Alginate In Situ Gelation: A Review.

Alginate-based hydrogels are promising smart materials widely employed in the food, bioengineering, and energy sectors. The development and optimization of their production require a thorough knowledge of gelation. In recent years, advanced experimental procedures have been developed for real-time cross-linking alginate reaction monitoring. Novel methods, such as customized rheometric setups, enable the recording of mechanical properties and morphological changes during hydrogel formation. These innovative techniques provide important insights into the gelation stages, the reaction rate, the diffusion of cross-linker to polymer chains, and the homogeneity of the gelling structures. Based on real-time experimental data, kinetic models are developed to enhance comprehension of the reaction mechanism and, eventually, to predict the gelation progress. The aim is to enable better control of the characterization of both the complex gelation and the propagated structures. This review aspires to present a comprehensive overview and evaluation of the breakthrough innovations of the real-time in situ recording of cross-linking alginate hydrogels and bead formation. A detailed analysis of the pioneering experimental developments provides a deep comprehension of the alginate gelation, including the parameters controlling the reaction.

Internet of things (IoT) for smart agriculture: Assembling and assessment of a low-cost IoT system for polytunnels.

Internet of things (IoT) applications in smart agricultural systems vary from monitoring climate conditions, automating irrigation systems, greenhouse automation, crop monitoring and management, and crop prediction, up to end-to-end autonomous farm management systems. One of the main challenges to the advancement of IoT systems for the agricultural domain is the lack of training data under operational environmental conditions. Most of the current designs are based on simulations and artificially generated data. Therefore, the essential first step is studying and understanding the finely tuned and highly sensitive mechanism plants have developed to sense, respond, and adapt to changes in their environment, and their behavior under field and controlled systems. Therefore, this study was designed to achieve two specific objectives; to develop low-cost IoT components from basic building blocks, and to study the performance of the developed systems, and generate real-time experimental data, with and without plants. Low-cost IoT devices developed locally were used to convert existing basic polytunnels to semi-controlled and monitoring-only polytunnels. Their performances were analyzed and compared with each other based on several matrices while maintaining the planted tomato variety and agronomic practices similar. The developed system performed as expected suggesting the possibility of commercial applications and research purposes.

Machine learning model for predicting the hardness of additively manufactured acrylonitrile butadiene styrene

Additive manufacturing is an incipient technology with great potential to achieve excellent mechanical properties with minimal material wastage. Additive manufacturing is well known for its quick fabrication of end-use components. The process parameter described during slicing process will influence the mechanical properties of the additively fabricated components. In this present study, machine learning models such as linear regression, decision tree, random forest, adaboost were used to optimize the process parameters and to predict hardness value for other process parameters. Machine learning models such as linear regression, decision tree, random forest, and adaboost were used to predict the output response. These models were selected to illustrate the variance in performance of various kinds of machine learning model on real-time experimental data. The test specimens were fabricated through fused filament fabrication (FFF). Acrylonitrile Butadiene Styrene (ABS) is an engineering thermoplastic used as a feedstock material. The hardness values were determined using a shore D durometer, which is specifically meant to measure the hardness of plastics and rubbers. The design of experiments was evaluated by considering four parameters at three different levels. Infill density, layer thickness, print orientation, and raster orientation are the four parameters, which was taken into account. The robustness and acceptability of four different machine learning models were evaluated at various error metrics. From the results, the adequacy of random forest model is highlighted over other machine learning models. From observation, the value of the co-efficient of determination(R2) for various models ranged from 0.8437 to 0.9136, with the random forest model ranking the highest (0.9136). This research will be helpful in predicting the level of input parameters needed to obtain the maximum hardness of ABS material.

State-of-charge estimation for lithium-ion batteries based on incommensurate fractional-order observer

This paper presents a new state-of-charge (SOC) estimation method for lithium-ion batteries (LIBs) based on incommensurate fractional-order (FO) observer. First, a FO RC equivalent circuit model is constructed to describe the charging and discharging dynamic characteristics of the LIB. An adaptive FO particle swarm optimization (AFOPSO) algorithm is employed to identify the parameters of the established FO model. The accuracy of the parameterized model is verified, showing its superiority over the available ones. Then, by using the comparison principle of FO systems with multi-order and stability of FO incommensurate linear systems with multi-order, a new order-dependent incommensurate FO observer is designed for battery SOC estimation. Finally, the validity and feasibility of the proposed SOC observer are examined with real time experimental data of LIBs.

Design and Application of Collaborative Experiment Management Platform for Innovation and Entrepreneurship Education Based on an Intelligent Sensor Network

Innovation and entrepreneurship is the soul of entrepreneurship, and realizing the integration of innovation and entrepreneurship education and collaborative experiment platform in colleges and universities is an important way to promote the development of entrepreneurship and facilitate higher quality employment. In this paper, the authors propose the 3D-IVF intelligent sensor overlay algorithm wisdom laboratory system structure and application mode through in-depth investigation of collaborative laboratory, build the macroenvironment and construction demand of collaborative experiment management platform of innovation and entrepreneurship education, comprehensively analyze the current situation, problems, and effectiveness of its integration of innovation and entrepreneurship education and professional education. In addition, we explore the operation mechanism of its organizational system, curriculum system, management system, and platform construction in the practice of integrating innovation and entrepreneurship education with professional education. The average coverage rate of the 3D-IVF-based 3D coverage algorithm is 92.15%, and the deployment time is 2.5857 s; the average coverage rate is improved by 0.76%, and the deployment time is reduced by 0.1712 s. Design FIR filter using MATLAB and Hilbert transform filter, generate HDL code, and implement it on FPGA. Use the designed low-pass and band-pass filters to filter the 100 KHz square wave signal sampled by the ADC. When the number of iterations of the algorithm reaches 30, the coverage rates of the four-node deployment strategies will stabilize, i.e., the forces between the nodes of the 3D-VF-based 3D coverage algorithm of this paper. The equilibrium state is reached. We use resources and real-time experimental data sharing to realize collaborative design among team members, further optimize the curriculum structure of innovation and entrepreneurship education, adopt diversified forms of curriculum implementation, and implement a diversified curriculum evaluation system, etc.

Extraction optimization of neem bioactives from neem seed kernel by ultrasonic assisted extraction and profiling by UPLC-QTOF-ESI-MS

BackgroundEfficient utilization of energy is the need of the hour. Being a time saving and energy efficient technique, ultrasonic assisted extraction (UAE) was utilised for optimization of sequential extraction of neem oil and azadirachtinoids from neem seed kernel (NSK) using response surface methodology (RSM). ResultsYield of neem oil from the NSK was 32.5%, whereas 0.19% yield of azadirachtin-A was obtained under optimised conditions. Standardised factors for the oil extraction included solvent:matrix ratio (4.81), time (18.41 min) and amplitude (45.8%), with hexane as solvent, whereas, for the extraction of azadirachtinoids, the corresponding optimised values of process factors were 4.78, 20 min and 50%, using ethanol as extraction solvent. First, azadirachtinoids and neem oil meliacins were characterised by UPLC-QTOF-ESI-MS. Identified neem meliacins were deacetylnimbin, nimbin, deacetylsalanin and salannin with their corresponding m/z value of 499.2333, 541.2438, 555.2955 and 597.3066 for [M+H]+ peak, respectively. Similarly, azadirachtin A, B and H were identified by their sodiated adduct with corresponding value of 743.2520, 685.2466, 685.2468. Aza-A, B and H present in the extract were quantified by UPLC-TQD-MS/MS using MRM technique at transitions 720.20 > 685.25, 662.40 > 645.20 and 663.10 > 57.15, respectively. Optimised method was validated by the real time experimental data and the yield of neem oil and azadirachtin-A was recorded as 31.8 and 0.182%. Greenness of the analytical method for estimation of Aza-A, B and H was calculated by Analytical Eco Scale and Green analytical procedure index metrics and found sufficiently well in terms of its greenness. ConclusionSequential UAE method for extraction of neem oil and azadirachtinoids developed in the present work will prove helpful for neem oil and biopesticide industry in better utilization of NSK as a valuable resource. Analytical method was significantly well in terms of greenness.

Collaborative smartphone experiments for large audiences with phyphox

We present methods to implement collaborative experimentation with smartphone sensors for larger audiences ranging from typical university undergraduate courses of hundreds of students in a lecture hall to world-wide outreaches on the Internet. These methods are based on the app ‘phyphox’, which is being developed by the authors, and encompass simple data collection via web forms as well as a new network interface for ‘phyphox’, allowing to collect real-time experiment data from an audience on-site or easy data submission for remote participants. Examples are given with practical considerations derived from first implementations of this method in a lecture hall with 350 undergraduate students as well as a global experiment to determine the Earth’s axial tilt with smartphones.

An Experimental Study to Locate Leakage in the Water Distribution Network using Real-time Wireless Sensor and Machine Learning Algorithm

Objective: To pinpoint the leakage location using a machine learning algorithm on a real-time basis. Method: A laboratory experimental model was developed using wireless sensors for real-time data collection and monitoring the changes in the pressure and flow in presence of leakage under a different scenario. Modification in laboratory model was made considering Loop network of distribution pipes as compared to the simple experimental model. The model has been validated in EPANET software. The machine learning algorithm, along with the K-fold approach has been used to locate leakage and it was compared with other algorithms like Support Vector Machine (SVM), logistic regression, multilayer perceptron, Recurrent Neural Network-Long Short-Term Memory (RNN-LSTM), Principal Component Analysis (PCA), Artificial Neural Network (ANN), Random Forest, Gradient Boosting Tree (GBT) Classification, and Convolutional Neural Network (CNN). Findings: In this research, a system was developed based on real-time data received by Wireless Sensor Network (WSN) was demonstrated and it is able to find even a small leak by monitoring pressure and flow. The K-fold approach in machine learning algorithm has been used to locate the leakage in different variations made in the experimental model in terms of pressure, leak size, cross-section of pipe, and profile level of the pipe network. A comparison was also presented with other machine learning algorithms of recent research in terms of accuracy, size of the opening for the leak, type of experimental model used, and variables considered to locate the leakage. Novelty: It is the first research kind of work that shows the average accuracy of 78% to locate leakage based on the real-time experimental data in a loop network using the K-Fold approach. Keywords: Leakage; Leakage Detection; Kfold approach; Pressure Measurements; Water Distribution Network; Wireless sensor

X-Ray Imaging of Complex Flow Patterns during Tungsten Inert Gas Welding

Fusion welding techniques such as tungsten inert gas (TIG) welding process have been widely used in industrial and construction applications. The molten metal flow in the weld pool has a major impact on the microstructure evolution, chemical element distribution and defects formation during solidification, which subsequently determines the performance of the welds. However, limited real-time experimental data availability of internal flow behavior has been considered as a major barrier to achieve a thorough understanding and development of accurate weld pool prediction models. In situ x-ray imaging with the tracking particles facilitated us to visualize the flow evolution during the solid–liquid–solid transformation. Experimental results indicated the flow patterns are progressively becoming complicated with the expansion of the melt pool. The shape of the melt pool also changed according to this flow evolution. Our analysis of flow patterns concerning the underlying variation of the driving forces suggests that gravity-derived buoyancy has a considerable effect on determining fluid flow at the melt pool periphery compared to other regions.

Applications and Techniques for Fast Machine Learning in Science.

In this community review report, we discuss applications and techniques for fast machine learning (ML) in science—the concept of integrating powerful ML methods into the real-time experimental data processing loop to accelerate scientific discovery. The material for the report builds on two workshops held by the Fast ML for Science community and covers three main areas: applications for fast ML across a number of scientific domains; techniques for training and implementing performant and resource-efficient ML algorithms; and computing architectures, platforms, and technologies for deploying these algorithms. We also present overlapping challenges across the multiple scientific domains where common solutions can be found. This community report is intended to give plenty of examples and inspiration for scientific discovery through integrated and accelerated ML solutions. This is followed by a high-level overview and organization of technical advances, including an abundance of pointers to source material, which can enable these breakthroughs.

WaterSafe: A Water Network Benchmark for Fault Diagnosis Research

Currently, in the water distribution systems literature, fault detection methods are typically evaluated on benchmark water networks that do not include real-time experimental data, or on private commercial datasets, which prohibit the reproducibility of the results. Moreover, realistic modeling of faults on hydraulic system components, sensors and actuators is often unavailable. In this work, we provide a framework for the application of fault-diagnosis methodologies on WaterSafe, a water network benchmark for fault diagnosis. The WaterSafe benchmark is a small scale replica of a water transport network constructed using industrial components and devices, while the communications are implemented in a way that resemble a water utility's Supervisory Control and Data Acquisition system. A general problem formulation for fault-diagnosis on water systems is provided, in accordance to the mathematical model of the benchmark. Moreover, we provide a calibrated simulation model including system, sensor and actuator faults, based on observations from the real system. Finally, we provide open access to the datasets generated from the experiments containing the aforementioned faults.