Search
Search results for: GLOBAL SURROGATE MODELING · NEURAL NETWORKS · MODEL UNCERTAINTY · ERROR BASED EXPLORATION.
-
Global Surrogate Modeling by Neural Network-Based Model Uncertainty
PublicationThis work proposes a novel adaptive global surrogate modeling algorithm which uses two neural networks, one for prediction and the other for the model uncertainty. Specifically, the algorithm proceeds in cycles and adaptively enhances the neural network-based surrogate model by selecting the next sampling points guided by an auxiliary neural network approximation of the spatial error. The proposed algorithm is tested numerically...
-
Iterative Global Sensitivity Analysis Algorithm with Neural Network Surrogate Modeling
PublicationGlobal sensitivity analysis (GSA) is a method to quantify the effect of the input parameters on outputs of physics-based systems. Performing GSA can be challenging due to the combined effect of the high computational cost of each individual physics-based model, a large number of input parameters, and the need to perform repetitive model evaluations. To reduce this cost, neural networks (NNs) are used to replace the expensive physics-based...
-
Efficient uncertainty quantification using sequential sampling-based neural networks
PublicationUncertainty quantification (UQ) of an engineered system involves the identification of uncertainties, modeling of the uncertainties, and the forward propagation of the uncertainties through a system analysis model. In this work, a novel surrogate-based forward propagation algorithm for UQ is proposed. The proposed algorithm is a new and unique extension of the recent efficient global optimization using neural network (NN)-based...
-
Trust-Based Model for the Assessment of the Uncertainty of Measurements in Hybrid IoT Networks
PublicationThe aim of this paper is to introduce a NUT model (NUT: network-uncertainty-trust) that aids the decrease of the uncertainty of measurements in autonomous hybrid Internet of Things sensor networks. The problem of uncertainty in such networks is a consequence of various operating conditions and varied quality of measurement nodes, making statistical approach less successful. This paper presents a model for decreasing the uncertainty...
-
Neural Network-Based Sequential Global Sensitivity Analysis Algorithm
PublicationPerforming global sensitivity analysis (GSA) can be challenging due to the combined effect of the high computational cost, but it is also essential for engineering decision making. To reduce this cost, surrogate modeling such as neural networks (NNs) are used to replace the expensive simulation model in the GSA process, which introduces the additional challenge of finding the minimum number of training data samples required to...
-
Session-Based Recommendation with Graph Neural Networks with an Examination of the Impact of Local and Global Vectors
PublicationThis study investigates the application of graph neural networks (GNN) in session-based recommendation systems (SR), focusing on a key modification involving the use of a global vector. Session-based recommendation systems often face challenges in accurately capturing user behavior due to the limited data available within individual sessions. The SR-GNN model, originally designed for automatic feature extraction from session graphs...
-
Fundamentals of Physics-Based Surrogate Modeling
PublicationChapter 1 was focused on data-driven (or approximation-based) modeling methods. The second major class of surrogates are physics-based models outlined in this chapter. Although they are not as popular, their importance is growing because of the challenges related to construction and handling of approximation surrogates for many real-world problems. The high cost of evaluating computational models, nonlinearity of system responses,...
-
Triangulation-based Constrained Surrogate Modeling of Antennas
PublicationDesign of contemporary antenna structures is heavily based on full-wave electromagnetic (EM) simulation tools. They provide accuracy but are CPU-intensive. Reduction of EM-driven design procedure cost can be achieved by using fast replacement models (surrogates). Unfortunately, standard modeling techniques are unable to ensure sufficient predictive power for real-world antenna structures (multiple parameters, wide parameter ranges,...
-
Accurate Modeling of Antenna Structures by Means of Domain Confinement and Pyramidal Deep Neural Networks
PublicationThe importance of surrogate modeling techniques has been gradually increasing in the design of antenna structures over the recent years. Perhaps the most important reason is a high cost of full-wave electromagnetic (EM) analysis of antenna systems. Although imperative in ensuring evaluation reliability, it entails considerable computational expenses. These are especially pronounced when carrying out EM-driven design tasks such...
-
Performance-Based Nested Surrogate Modeling of Antenna Input Characteristics
PublicationUtilization of electromagnetic (EM) simulation tools is mandatory in the design of contemporary antenna structures. At the same time, conducting designs procedures that require multiple evaluations of the antenna at hand, such as parametric optimization or yield-driven design, is hindered by a high cost of accurate EM analysis. To certain extent, this issue can be addressed by utilization of fast replacement models (also referred...
-
A MODEL FOR FORECASTING PM10 LEVELS WITH THE USE OF ARTIFICIAL NEURAL NETWORKS
PublicationThis work presents a method of forecasting the level of PM10 with the use of artificial neural networks. Current level of particulate matter and meteorological data was taken into account in the construction of the model (checked the correlation of each variable and the future level of PM10), and unidirectional networks were used to implement it due to their ease of learning. Then, the configuration of the network (built on the...
-
Classification of objects in the LIDAR point clouds using Deep Neural Networks based on the PointNet model
PublicationThis work attempts to meet the challenges associated with the classification of LIDAR point clouds by means of deep learning. In addition to achieving high accuracy, the designed system should allow the classification of point clouds covering an area of several dozen square kilometers within a reasonable time interval. Therefore, it must be characterized by fast processing and efficient use of memory. Thus, the most popular approaches...
-
Results of implementation of Feed Forward Neural Networks for modeling of heat transfer coefficient during flow condensation for low and high values of saturation temperature
Open Research DataThis database present results of implementation of Feed Forward Neural Networks for modeling of heat transfer coefficient during flow condensation for low and high values of saturation temperature. Databse contain one table and 7 figures.
-
Wiktoria Wojnicz dr hab. inż.
PeopleDSc in Mechanics (in the field of Biomechanics) - Lodz Univeristy of Technology, 2019 PhD in Mechanics (in the field of Biomechanics) - Lodz Univeristy of Technology, 2009 (with distinction) List of papers (2009 - ) Wojnicz W., Wittbrodt E., Analysis of muscles' behaviour. Part I. The computational model of muscle. Acta of Bioengineering and Biomechanics, Vol. 11, No.4, 2009, p. 15-21 Wojnicz W., Wittbrodt E., Analysis of...
-
Traffic Modeling in IMS-based NGN Networks
PublicationIn the modern world the need for accurate and quickly delivered information is becoming more and more essential. In order to fulfill these requirements, next generation telecommunication networks should be fast introduced and correctly dimensioned. For this reason proper traffic models must be identified, which is the subject of this paper. In the paper standardization of IMS (IP Multimedia Subsystem) concept and IMS-based NGN...
-
Modeling of Surface Roughness in Honing Processes by UsingFuzzy Artificial Neural Networks
PublicationHoning processes are abrasive machining processes which are commonly employed to improve the surface of manufactured parts such as hydraulic or combustion engine cylinders. These processes can be employed to obtain a cross-hatched pattern on the internal surfaces of cylinders. In this present study, fuzzy artificial neural networks are employed for modeling surface roughness parameters obtained in finishing honing operations. As...
-
Classification of Glacial and Fluvioglacial Landforms by Convolutional Neural Networks Using a Digital Elevation Model
PublicationThe rise of artificial neural networks (ANNs) has revolutionized various fields of research, demonstrating their effectiveness in solving complex problems. However, there are still unexplored areas where the application of neural networks, particularly convolutional neural network (CNN) models, has yet to be explored. One area is where the application of ANNs is even expected is geomorphology. One of the tasks of geomorphology...
-
Adaptive Hyperparameter Tuning within Neural Network-based Efficient Global Optimization
PublicationIn this paper, adaptive hyperparameter optimization (HPO) strategies within the efficient global optimization (EGO) with neural network (NN)-based prediction and uncertainty (EGONN) algorithm are proposed. These strategies utilize Bayesian optimization and multiarmed bandit optimization to tune HPs during the sequential sampling process either every iteration (HPO-1itr) or every five iterations (HPO-5itr). Through experiments using...
-
Heavy Duty Vehicle Fuel Consumption Modelling Based on Exploitation Data by Using Artificial Neural Networks
PublicationOne of the ways to improve the fuel economy of heavy duty trucks is to operate the combustion engine in its most efficient operating points. To do that, a mathematical model of the engine is required, which shows the relations between engine speed, torque and fuel consumption in transient states. In this paper, easy accessible exploitation data collected via CAN bus of the heavy duty truck were used to obtain a model of a diesel...
-
A compact smart sensor based on a neural classifier for objects modeled by Beaunier's model
PublicationA new solution of a smart microcontroller sensor based on a simple direct sensor-microcontroller interface for technical objects modeled by two-terminal networks and by the Beaunier’s model of anticorrosion coating is proposed. The tested object is stimulated by a square pulse and its time voltage response is sampled four times by the internal ADC of microcontroller. A neural classifier based on measurement data classifies the...
-
Cost-Efficient Surrogate Modeling of High-Frequency Structures Using Nested Kriging with Automated Adjustment of Model Domain Lateral Dimensions
PublicationSurrogate models are becoming popular tools of choice in mitigating issues related to the excessive cost of electromagnetic (EM)-driven design of high-frequency structures. Among available techniques, approximation modeling is by far the most popular due to its versatility. In particular, the surrogates are exclusively based on the sampled simulation data with no need to involve engineering insight or problem-specific knowledge....
-
System for monitoring road slippery based on CCTV cameras and convolutional neural networks
PublicationThe slipperiness of the surface is essential for road safety. The growing number of CCTV cameras opens the possibility of using them to automatically detect the slippery surface and inform road users about it. This paper presents a system of developed intelligent road signs, including a detector based on convolutional neural networks (CNNs) and the transferlearning method employed to the processing of images acquired with video...
-
Customization of UWB 3D-RTLS Based on the New Uncertainty Model of the AoA Ranging Technique
PublicationThe increased potential and effectiveness of Real-time Locating Systems (RTLSs) substantially influence their application spectrum. They are widely used, inter alia, in the industrial sector, healthcare, home care, and in logistic and security applications. The research aims to develop an analytical method to customize UWB-based RTLS, in order to improve their localization performance in terms of accuracy and precision. The analytical...
-
Artificial Neural Networks in Microwave Components and Circuits Modeling
PublicationArtykuł dotyczy wykorzystania sztucznych sieci neuronowych (SNN) w projektowaniu i optymalizacji układów mikrofalowych.Zaprezentowano podstawowe zasady i założenia modelowania z użyciem SNN. Możliwości opisywanej metody opisano wykorzystując przykładowyprojekt anteny łatowej. Przedstawiono różne strategie modelowania układów, które wykorzystują możliwości opisywanej metody w połączeniu zwiedzą mikrofalową. Porównano również dokładność...
-
Fast Low-fidelity Wing Aerodynamics Model for Surrogate-Based Shape Optimization
PublicationVariable-fidelity optimization (VFO) can be efficient in terms of the computational cost when compared with traditional approaches, such as gradient-based methods with adjoint sensitivity information. In variable-fidelity methods, the directoptimization of the expensive high-fidelity model is replaced by iterative re-optimization of a physics-based surrogate model, which is constructed from a corrected low-fidelity model. The success...
-
Blood Pressure Estimation Based on Blood Flow, ECG and Respiratory Signals Using Recurrent Neural Networks
PublicationThe estimation of systolic and diastolic blood pressure using artificial neural network is considered in the paper. The blood pressure values are estimated using pulse arrival time, and additionally RR intervals of ECG signal together with respiration signal. A single layer recurrent neural network with hyperbolic tangent activation function was used. The average blood pressure estimation error for the data obtained from 21 subjects...
-
Accurate Modeling of Frequency Selective Surfaces Using Fully-Connected Regression Model with Automated Architecture Determination and Parameter Selection Based on Bayesian Optimization
PublicationSurrogate modeling has become an important tool in the design of high-frequency structures. Although full-wave electromagnetic (EM) simulation tools provide an accurate account for the circuit characteristics and performance, they entail considerable computational expenditures. Replacing EM analysis by fast surrogates provides a way to accelerate the design procedures. Unfortunately, modeling of microwave passives is a challenging...
-
Efficient Surrogate Modeling and Design Optimization of Compact Integrated On-Chip Inductors Based on Multi-Fidelity EM Simulation Models
PublicationHigh-performance and small-size on-chip inductors play a critical role in contemporary radio-frequency integrated circuits. This work presents a reliable surrogate modeling technique combining low-fidelity EM simulation models, response surface approximations based on kriging interpolation, and space mapping technology. The reported method is useful for the development of broadband and highly accurate data-driven models of integrated...
-
Artificial Neural Networks as an architectural design tool- generating new detail forms based on the Roman Corinthian order capital
PublicationThe following paper presents the results of the research in the field of the machine learning, investigating the scope of application of the artificial neural networks algorithms as a tool in architectural design. The computational experiment was held using the backward propagation of errors method of training the artificial neural network, which was trained based on the geometry of the details of the Roman Corinthian order capital....
-
Application of Feed Forward Neural Networks for Modeling of Heat Transfer Coefficient During Flow Condensation for Low and High Values of Saturation Temperatur
PublicationMost of the literature models for condensation heat transfer prediction are based on specific experimental parameters and are not general in nature for applications to fluids and non-experimental thermodynamic conditions. Nearly all correlations are created to predict data in normal HVAC conditions below 40°C. High temperature heat pumps operate at much higher parameters. This paper aims to create a general model for the calculation...
-
Comparative study of neural networks used in modeling and control of dynamic systems
PublicationIn this paper, a diagonal recurrent neural network that contains two recurrent weights in the hidden layer is proposed for the designing of a synchronous generator control system. To demonstrate the superiority of the proposed neural network, a comparative study of performances, with two other neural network (1_DRNN) and the proposed second-order diagonal recurrent neural network (2_DRNN). Moreover, to confirm the superiority...
-
Sensing Direction of Human Motion Using Single-Input-Single-Output (SISO) Channel Model and Neural Networks
PublicationObject detection Through-the-Walls enables localization and identification of hidden objects behind the walls. While numerous studies have exploited Channel State Information of Multiple Input Multiple Output (MIMO) WiFi and radar devices in association with Artificial Intelligence based algorithms (AI) to detect and localize objects behind walls, this study proposes a novel non-invasive Through-the-Walls human motion direction...
-
Adrian Bekasiewicz dr hab. inż.
PeopleAdrian Bekasiewicz received the MSc, PhD, and DSc degrees in electronic engineering from Gdansk University of Technology, Poland, in 2011, 2016, and 2020, respectively. In 2014, he joined Engineering Optimization & Modeling Center at Reykjavik University, Iceland, where he held a Research Associate and a Postdoctoral Fellow positions, respectively. Currently, he is an Associate Professor and the head of Teleinformation Networks...
-
Neural networks based NARX models in nonlinear adaptive control
Publication -
Monitoring the gas turbine start-up phase on the platform using a hierarchical model based on Multi-Layer Perceptron networks
PublicationVery often, the operation of diagnostic systems is related to the evaluation of process functionality, where the diagnostics is carried out using reference models prepared on the basis of the process description in the nominal state. The main goal of the work is to develop a hierarchical gas turbine reference model for the estimation of start-up parameters based on multi-layer perceptron neural networks. A functional decomposition...
-
Nuclear Power Plant Steam Turbine - Modeling for Model Based Control Purposes
PublicationThe nature of the processes taking place in a nuclear power plant (NPP) steam turbine is the reason why their modeling is very difficult, especially when the model is intended to be used for on-line optimal model based process control over a wide range of operating conditions, caused by changing electrical power demand e.g. when combined heat and power mode of work is utilized. The paper presents three nonlinear models of NPP steam...
-
Neural Networks Based on Ultrafast Time-Delayed Effects in Exciton Polaritons
Publication -
Accidental wow evaluation based on sinusoidal modeling and neural nets prediction
PublicationReferat przedstawia opis algorytmu do określenia charakterystyki zniekształcenia kołysania dźwięku. Prezentowane podejście wykorzystuje sinusoidalną analizę dźwięku bazującą zarówno na amplitudowym jak i fazowym widmie sygnału fonicznego. Trajektorie poszczególnych składowych tonalnych, obrazujące zniekształcenie kołysania, określane są na podstawie analizy ich chwilowych amplitud, częstotliwości i faz. Dodatkowo referat przedstawia...
-
Automated speech-based screening of depression using deep convolutional neural networks
Publication -
Artificial Neural Networks in Classification of Steel Grades Based on Non-Destructive Tests
Publication -
Prediction of Early Childhood Caries Based on Single Nucleotide Polymorphisms Using Neural Networks
Publication -
Mathematical Modeling of Hydrodynamics in Bioreactor by Means of CFD-Based Compartment Model
Publication -
Modeling the economic dependence between town development policy and increasing energy effectiveness with neural networks. Case study: The town of Zielona Góra
Publication -
Exploration of the Solubility Hyperspace of Selected Active Pharmaceutical Ingredients in Choline- and Betaine-Based Deep Eutectic Solvents: Machine Learning Modeling and Experimental Validation
PublicationDeep eutectic solvents (DESs) are popular green media used for various industrial, pharmaceutical, and biomedical applications. However, the possible compositions of eutectic systems are so numerous that it is impossible to study all of them experimentally. To remedy this limitation, the solubility landscape of selected active pharmaceutical ingredients (APIs) in choline chloride- and betaine-based deep eutectic solvents was...
-
On dynamic modeling of piezomagnetic/flexomagnetic microstructures based on Lord–Shulman thermoelastic model
PublicationWe study a time-dependent thermoelastic coupling within free vibrations of piezomagnetic (PM) microbeams considering the flexomagnetic (FM) phenomenon. The flexomagneticity relates to a magnetic field with a gradient of strains. Here, we use the generalized thermoelasticity theory of Lord–Shulman to analyze the interaction between elastic deformation and thermal conductivity. The uniform magnetic field is permeated in line with...
-
Modeling and control of a redundantly actuated variable mass 3RRR planar manipulator controlled by a model-based feedforward and a model-based-proportional-derivative feedforward–feedback controller
PublicationIn the paper, dynamics of a complex mechatronics system is considered. A redundantly actuated planar manipulator is the base of the mechanical part of it. It is a 3RRR 1 platform based parallel manipulator. To control its trajectory, a model-based feedforward controller is employed. Three aspects are fundamental in the presented investigations. The first focus is on development of an accurate numerical model used to solve the inverse...
-
A new fuzzy model of multi-criteria decision support based on Bayesian networks for the urban areas' decarbonization planning
Publication -
Activation maps of convolutional neural networks as a tool for brain degeneration tracking in early diagnosis of dementia in Parkinson's disease based on magnetic resonance imaging
Publication -
Modeling and optimizing the removal of cadmium by Sinapis alba L. from contaminated soil via Response Surface Methodology and Artificial Neural Networks during assisted phytoremediation with sewage sludge
Publication -
Adaptive Sampling for Non-intrusive Reduced Order Models Using Multi-Task Variance
PublicationNon-intrusive reduced order modeling methods (ROMs) have become increasingly popular for science and engineering applications such as predicting the field-based solutions for aerodynamic flows. A large sample size is, however, required to train the models for global accuracy. In this paper, a novel adaptive sampling strategy is introduced for these models that uses field-based uncertainty as a sampling metric. The strategy uses...