Wyniki wyszukiwania - BazTech

1

Survey on multi-objective-based parameter optimization for deep learning

Chakraborty Mrittika, Pal Wreetbhas, Bandyopadhyay Sanghamitra, Maulik Ujjwal

Computer Science

|

2023

|

T. 24 (3)

327--359

EN

Deep learning models form some of the most powerful machine-learning models for the extraction of important features. Most of the designs of deep neural models ( i.e., the initialization of parameters) are still manually tuned; hence, obtaining a model with high performance is exceedingly time-consuming and occasionally impossible. Optimizing the parameters of deep networks, therefore requires improved optimization algorithms with high convergence rates. The single objective-based optimization methods that are generally used are mostly time-consuming and do not guarantee optimum performance in all cases. Mathematical optimization problems that contain multiple objective functions that must be optimized simultaneously fall under the category of multi-objective optimization (sometimes referred to as Pareto optimization). Multi-objective optimization problems form one of the alternatives yet useful options for parameter optimization; however, this domain is a bit underexplored. In this survey, we focus on exploring the effectiveness of multi-objective optimization strategies for parameter optimization in conjunction with deep neural networks. The case studies that are used in this study focus on how the two methods are combined to provide valuable insights into the generation of predictions and analysis in multiple applications.

2

Research on improved sparrow search algorithm for PID controller parameter optimization

Zhang Mingfeng, Xu Chuntian, Xu Deying, Ma Guoqiang, Han Han, Zong Xu

Bulletin of the Polish Academy of Sciences. Technical Sciences

|

2023

|

Vol. 71, nr 6

art. no. e147344

EN

PID controllers are crucial for industrial control because of their simple structure and good robustness. In order to further improve the accuracy of PID controllers, this paper proposes an improved sparrow search algorithm (ISSA) to prevent the problem of the algorithm being prone to falling into the local optimum at the late stage of iteration. Based on the standard sparrow search algorithm, the position update formula and the step size control parameter are optimized to help quickly jump out of the local, and to obtain the optimal solution in the whole domain. Finally, to verify the accuracy and stability of the improved algorithm, nine standard test functions are first simulated. Then, the PID parameter optimization tests are finished with the chilled water and battery charging systems, where the lifting load and applying perturbation are carried out. Both the simulation and test results show that ISSA improves the convergence speed and accuracy, and performs better in terms of stability.

3

Research on Forming Parameters Optimization of Incremental Sheet Forming Process for Commercially Pure Titanium Grade 2 Sheets

Szpunar Marcin, Trzepieciński Tomasz, Ostrowski Robert, Zwolak Marek

Archives of Metallurgy and Materials

|

2022

|

Vol. 67, iss. 4

1411--1418

EN

This study aims to determine optimal forming parameters for Incremental Sheet Forming process Commercially Pure titanium Grade 2 sheets in terms of formability improvement, force reduction, and efficiency of forming. Based on the central composite design, data were collected during 20 runs and then variation analysis was performed. The experiments were performed on a 3 axis CNC milling machine equipped with a Kistler dynamometer plate. Subsequently, regression models have been developed to describe process responses by input factors. As crucial parameters, the relative velocity and step size of the tool that affect the forming force and the height of the fracture have been determined. Finally, the application of optimization algorithm has emerged optimal input factors in terms of selected multi-criteria goal. The results of this study suggest that there is a process window that allows the formation of 45° wall angle drawpieces of commercially pure titanium Grade 2.

4

Optymalizacja parametrów cięcia laserowego materiałów obuwniczych

Alvarez Joanna, Pieknik Marcin

Technologia i Jakość Wyrobów

|

2020

|

R. 65

79--91

PL

Przedstawiono wyniki optymalizacji parametrów laserowego cięcia materiałów obuwniczych, tj. prędkości oraz mocy. Testom poddano wybrane materiały zróżnicowane surowcem i/lub sposobem wytwarzania, grubością, a także docelowym przeznaczeniem. Na podstawie przeprowadzonych testów stwierdzono, że właściwości fizyczne promieniowania lasera CO2 – Maximus PRO (model: JSM 90 x 60), pozwalają zapewnić wysoką jakość obróbki dowolnych materiałów obuwniczych. Efekty wizualne cięcia zależą od precyzji ustawień ogniskowej pomiędzy płaską powierzchnią cięcia i głowicą lasera oraz parametrów prędkości i mocy lasera. Konieczny jest więc ich każdorazowy dobór na podstawie przeprowadzonych testów kontrolnych.

EN

The results of optimizing laser cutting parameters of footwear materials, i.e. speed and power, are presented. Selected materials were tested, differing in terms of raw material and/or manufacturing method, thickness and intended use. On the basis of the tests, it was found that the physical properties of CO2 laser radiation – Maximus PRO (model: JSM 90 x 60), ensure high quality processing of any footwear materials. The visual effects of the cut depend on the precision of the focal length settings between the flat cutting surface and the laser head, as well as the speed and laser power parameters. Therefore, it is necessary to always select them on the basis of the control tests.

5

Aerodynamic performance of a new double-flap wing sail

Li Dongqin, Zhang Yili, Li Peng, Dai Jingjing, Li Guohuan

Polish Maritime Research

|

2019

|

nr 4

61--68

EN

As a type of natural energy resource, wind power is used in the modern implementation of wind-assisted technologies as a method for reducing the fuel consumption and environmental pollution of ocean-going ships. In order to promote the full usage of ocean wind energy for cargo ships, an innovative type of ship propulsion-assisted wing sail is proposed in this paper. The propulsion efficiency of this new wing sail can be increased by enlarging its area in both the transverse and vertical directions in good weather conditions, and it can be folded up automatically in poor weather conditions, improving the sailing safety of the ship. The sail parameters relating to the gaps and rotation angles between different parts of the wing sail are compared, and the values giving the best aerodynamic performance are identified using CFD simulation technology. The results for the lift and drag coefficients for the new wing sail at different attack angles are also compared with those of traditional aerofoil sails, including an arc-shaped rigid sail and a variable-camber sail proposed in 2015. From the viewpoint of the sailing performance of the vessel, our results demonstrate that this new type of wing sail has good aerodynamic performance and can reduce fuel costs for commercial vessels.

6

Fault diagnosis of power transformer based on improved particle swarm optimization OS-ELM

Li Yuancheng, Ma Longqiang

Archives of Electrical Engineering

|

2019

|

Vol. 68, nr 1

161--172

EN

A transformer is an important part of power transmission and transformation equipment. Once a fault occurs, it may cause a large-scale power outage. The safety of the transformer is related to the safe and stable operation of the power system. Aiming at the problem that the diagnosis result of transformer fault diagnosis method is not ideal and the model is unstable, a transformer fault diagnosis model based on improved particle swarm optimization online sequence extreme learning machine (IPSO-OS-ELM) algorithm is proposed. The improved particle swarmoptimization algorithm is applied to the transformer fault diagnosis model based on the OS-ELM, and the problems of randomly selecting parameters in the hidden layer of the OS-ELM and its network output not stable enough, are solved by optimization. Finally, the effectiveness of the improved fault diagnosis model in improving the accuracy is verified by simulation experiments.

7

Sensitivity of optimized transient storage model parameters to spatial and temporal resolution

Wallis Steve, Manson Russell

Acta Geophysica

|

2019

|

Vol. 67, no. 3

951--960

EN

The transient storage model is a popular tool for modelling solute transport along rivers. Its use requires values for the velocity and shear flow dispersion coefficient in the main channel of the river together with two exchange rates between the main channel and transient storage zones, which surround the main channel. Currently, there is insufficient knowledge to enable these parameters to be predicted from the type of hydraulic variables that may typically be available. Hence, recourse is made to tracer experiments, which provide temporal solute concentration profiles that can be used to estimate the parameters by optimizing model output to observations. The paper explores the sensitivity of such parameters to the spatial and temporal resolutions used in the optimization of the model. Data from 25 tracer experiments covering a river flow rate range of 300–2250 L/s in a single reach of the river Brock in north-west England were used. The shear flow dispersion coefficient was found to be the most sensitive parameter; the velocity was found to be the least sensitive parameter. When averaged over all the experiments, mean percentage differences in parameter values between a coarse resolution case and a fine resolution case were of the order of 2% for the velocity, 70% for the shear flow dispersion coefficient and 30% and 20% for the two exchange rates. Since the shear flow dispersion coefficient was found to be small, both in numerical terms and in comparison with an estimate of the total dispersion in the reach, it is suggested that it may be viable to omit the shear flow dispersion term from the model.

8

Optymalizacja parametrów systemu odometrii wizyjnej RGB-D metodami populacyjnymi

Kostusiak A., Skrzypczyński P.

Prace Naukowe Politechniki Warszawskiej. Elektronika

|

2018

|

z. 196

461--470

PL

Dokładna lokalizacja jest bardzo ważna w wielu praktycznych zagadnieniach robotyki mobilnej. Często korzysta się z odometrii wizyjnej, obecnie powszechnie wykorzystującej kamery (sensory) RGB-D. W niniejszym artykule badana jest możliwość automatycznego doboru optymalnych parametrów prostego systemu odometrii wizyjnej RGB-D. Wykorzystano i porównano dwie metody optymalizacji oparte na populacji rozwiązań: algorytm genetyczny oraz algorytm roju cząstek. Zbadano wpływ poszczególnych parametrów na dokładność otrzymywanych estymat trajektorii sensora. Na tej podstawie wyciągnięto wnioski zarówno co do skuteczności i efektywności zastosowanych metod optymalizacji, jak i co do wpływu poszczególnych parametrów na dokładność estymat. Eksperymenty przeprowadzono przy wykorzystaniu publicznie dostępnych zestawów danych, aby zapewnić weryfikowalność prezentowanych wyników.

EN

In this paper, we investigate two population-based optimization methods as the means for optimization of the selected parameters in a visual odometry system using RGB-D data. One of the simplest and most used approaches to localization with RGB-D data is feature-based visual odometry that computes frame-to-frame rigid transformations of the sensor upon a sparse set of features and then concatenates these transformations into an estimate of the trajectory. This approach, yet simple, requires careful tuning of a number of parameters that control both the behavior of the feature detector, and the frame-to-frame rota-translation estimation algorithm. Therefore, we propose to employ robust and efficient soft computing optimization methods to find the best parameters for an exemplary RGB-D visual odometry system. We investigate and compare two approaches: the simple to implement particle swarm optimization algorithm, and a more complicated variant of the genetic algorithm. We seek a set of parameters that not only provide good results in terms of the estimated trajectory residual errors but are also applicable to different RGB-D datasets. Moreover, the optimization experiments make it possible to draw more general conclusions as to the role of particular building blocks of the visual odometry system (e.g. RANSAC) in achieving accurate trajectory estimates.

9

Using particle swarm optimization to accurately identify syntactic phrases in free text

Tambouratzis G.

Journal of Artificial Intelligence and Soft Computing Research

|

2018

|

Vol. 8, No. 1

63--77

EN

The present article reviews the application of Particle Swarm Optimization (PSO) algorithms to optimize a phrasing model, which splits any text into linguistically-motivated phrases. In terms of its functionality, this phrasing model is equivalent to a shallow parser. The phrasing model combines attractive and repulsive forces between neighbouring words in a sentence to determine which segmentation points are required. The extrapolation of phrases in the specific application is aimed towards the automatic translation of unconstrained text from a source language to a target language via a phrase-based system, and thus the phrasing needs to be accurate and consistent to the training data. Experimental results indicate that PSO is effective in optimising the weights of the proposed parser system, using two different variants, namely sPSO and AdPSO. These variants result in statistically significant improvements over earlier phrasing results. An analysis of the experimental results leads to a proposed modification in the PSO algorithm, to prevent the swarm from stagnation, by improving the handling of the velocity component of particles. This modification results in more effective training sequences where the search for new solutions is extended in comparison to the basic PSO algorithm. As a consequence, further improvements are achieved in the accuracy of the phrasing module.

10

Nonoverlapping Schwarz Waveform Relaxation Algorithm for a Class of Time-Fractional Heat Equations

Wu S.-L., Wu G.-C.

Fundamenta Informaticae

|

2017

|

Vol. 151, nr 1/4

231--240

EN

In this paper, we analyze the convergence properties of the Schwarz waveform relaxation (SWR) algorithm with Robin transmission conditions (TCs) for a class of heat equations with Riemann-Liouville fractional derivative. The Robin TCs contain a free parameter, which has a significant effect on the convergence rate of the SWR algorithm, and optimizing this parameter is an important step for the convergence analysis of the SWR algorithm. By studying the monotonic properties of the convergence factor obtained by applying the Fourier transform to the error functions, we provide a realiable choice of the Robin parameter in the nonoverlapping case. Numerical results are provided, which show that the analyzed Robin parameter results in satisfactory convergence rate.

11

Optimizing the parameters of a small standalone hybrid power plant

Bogalecka E., Michoński A.

Acta Energetica

|

2017

|

nr 2

4--10

EN

In recent years, interest has grown in Poland in the installation of renewable energy sources (RES), including small standalone hybrid power plants aiming at full independence of energy supply from the power grid. A hybrid power plant consists of renewable energy sources, such as a solar and/or wind power plant, an energy storage facility providing the system’s autonomy, a discharge load for surplus energy in the system, and an emergency power supply. The power plant is equipped with an energy management system. Power plant parameters are tailored to meet the requirements of continuity of supply, cost minimization, return on investment period, and system capacity utilization. The paper presents a methodology for selecting power plant parameters with a larger number of decision criteria. The task is solved as a single-criterion optimization task with a weighted quality indicator. The user priority reflecting indicator weights were determined using the multi-criteria hierarchical method for analysing decision problems, in other words the Saaty’s analytic hierarchy process (AHP). The climatic data typical for Polish territory and the energy needs of a selected household were selected for the study.

PL

W ostatnich latach wzrosło w Polsce zainteresowanie instalacją odnawialnych źródeł energii (OZE), w tym małych autonomicznych elektrowni hybrydowych mających na celu pełne uniezależnienie od dostaw energii z sieci elektroenergetycznej. Elektrownia hybrydowa składa się z odnawialnych źródeł energii, np. elektrowni słonecznej i/lub wiatrowej, magazynu energii zapewniającego systemowi autonomię, odbiornika zrzutowego wykorzystującego nadwyżki energii w systemie oraz z zasilania awaryjnego. Elektrownia jest wyposażona w układ zarządzania zasobami energetycznymi. Parametry elektrowni dobierane są tak, aby zaspokoić wymagania: ciągłość zasilania, minimalizację kosztów, określony czas zwrotu inwestycji, wykorzystanie potencjału instalacji. W pracy przedstawiono metodologię doboru parametrów elektrowni przy większej liczbie kryteriów decyzyjnych. Zadanie rozwiązuje się jako zadanie optymalizacji jednokryterialnej z ważonym wskaźnikiem jakości. Wagi wskaźnika odzwierciedlające priorytety użytkownika wyznaczono, stosując wielokryterialną metodę hierarchiczną analizy problemów decyzyjnych Saaty’ego (AHP). Do badań wybrano dane klimatyczne typowe dla terenu Polski i potrzeby energetyczne wybranego gospodarstwa domowego.

12

Badanie procesu spulchniania gleby wibracyjnym zespołem podkopującym

Rapinchuk A.L., Baranovskiy I. A., Kassar F. H., Tanaś W., Szymanek M.

Journal of Research and Applications in Agricultural Engineering

|

2016

|

Vol. 61, nr 1

66--69

PL

W publikacji przedstawiono analizę wyników badań eksperymentalnych wibracyjnego zespołu podkopującego. Analizie poddano wpływ amplitudy, częstotliwości drgań podkopujących łap oraz prędkości roboczej na zapotrzebowanie mocy. Przeprowadzono optymalizację parametrów deformacji gruntu.

EN

The paper presents an analysis of experimental study results of vibrating digging unit. The impact of amplitude, frequency of vibrating digging shares and the operating speed on power consumption were analyzed. The soil deformation parameters were optimized.

13

Statistical Optimization of the Preparation of HNIW Nanoparticles via Oil in Water Microemulsions

Bayat Y., Zarandi M., Khadiv-Parsi P., Salimi Beni A.

Central European Journal of Energetic Materials

|

2015

|

Vol. 12, no. 3

459--472

EN

HNIW (2,4,6,8,10,12-hexanitro-2,4,6,8,10,12-hexaazaisowurtzitane) is a family member of high-energy density cage nitramines which have so many versatile applications. In this paper, HNIW nanoparticles were prepared by the oil in water microemulsion route. The effects of various experimental parameters on this reaction were investigated using the Taguchi method. The effects of different variables: organic phase, water/organic phase (W1/W2), organic phase/ propanol (W3/W4) and HNIW weight percent, on the particle size of the HNIW were investigated at three distinct levels. Optimal conditions for obtaining HNIW nanoparticles were determined. Performing the process under the optimal conditions proposed by the Taguchi method leads to the production of HNIW nanoparticles with an average size of about 80 nm. The HNIW nanoparticles were characterized using Scanning Electron Microscopy (SEM), Dynamic Light Scattering (DLS), Differential Thermal Analysis (DTA) and X-Ray Diffraction (XRD).

14

Model for selection of involute gear parameters using the area of possible solutions method

Herma S.

Management and Production Engineering Review

|

2012

|

Vol. 3, No. 2

13-20

EN

The paper presents a method of selecting the parameters of external involute spur gearing components with straight and helical teeth for general - purpose gears, using the method of searching an area of possible solutions, understood as a space for values that meet several geometric, strength and operating criteria. Possible effects of this method in the context of gear production process organization and optimization, were also described.

15

Prosty sposób optymalizacji parametrycznej w metodzie (FOC) dla silnika indukcyjnego

Sieklucki G., Orzechowski T., Sykulski R.

Elektrotechnika i Elektronika

|

2005

|

T. 24, z. 1

93-99

PL

W artykule przedstawiono standardową strukturę układu sterowania polowo zorientowanego (FOC) regulacji prędkości silnika indukcyjnego zasilanego z falownika napięcia. Zaprezentowano prostą metodę doboru regulatorów prądów i prędkości (typu PI). Zamieszczono wyniki badań symulacyjnych.

EN

Standard structure of field orientated control (FOC) system of induction motor is presented. Simple method of parameters selection of currents and speed controller (type PI) are proposed. Simulation result of this control system are discussed.