scholarly journals Artificial Neural Network as a Universal Model of Nonlinear Dynamical Systems

2021 ◽  
Vol 17 (1) ◽  
pp. 5-21
Author(s):  
P. V. Kuptsov ◽  
◽  
A. V. Kuptsova ◽  
N. V. Stankevich ◽  
◽  
...  
Keyword(s):  
Neural Network ◽  
Artificial Neural Network ◽  
Dynamical Systems ◽  
Network Model ◽  
Numerical Solutions ◽  
Nonlinear Dynamical Systems ◽  
Universal Model ◽  
Dynamical Equations ◽  
Wide Range ◽  
Artificial Neural

We suggest a universal map capable of recovering the behavior of a wide range of dynamical systems given by ODEs. The map is built as an artificial neural network whose weights encode a modeled system. We assume that ODEs are known and prepare training datasets using the equations directly without computing numerical time series. Parameter variations are taken into account in the course of training so that the network model captures bifurcation scenarios of the modeled system. The theoretical benefit from this approach is that the universal model admits applying common mathematical methods without needing to develop a unique theory for each particular dynamical equations. From the practical point of view the developed method can be considered as an alternative numerical method for solving dynamical ODEs suitable for running on contemporary neural network specific hardware. We consider the Lorenz system, the Rössler system and also the Hindmarch – Rose model. For these three examples the network model is created and its dynamics is compared with ordinary numerical solutions. A high similarity is observed for visual images of attractors, power spectra, bifurcation diagrams and Lyapunov exponents.

scholarly journals The Effect of Different Cover and Flooring Materials on Climatic Comfort in Landscape Design

2022 ◽  
Vol 9 (sp) ◽  
pp. 2571-2580
Author(s):  
Orhun Soydan ◽  
Ahmet Benliay
Keyword(s):  
Neural Network ◽  
Artificial Neural Network ◽  
Network Model ◽  
Temperature Difference ◽  
Neural Network Model ◽  
Artificial Neural Network Model ◽  
Ambient Air ◽  
Maximum Temperature ◽  
Wide Range ◽  
Artificial Neural

In this study, it is aimed to understand the effects of structural and vegetative elements that can be used in landscape designs on the temperature factor, which will greatly affect the climatic comfort, by using artificial neural networks. In this context, measurements were carried out in the morning (08:00-09:00), noon (13:00-14:00) and evening (17:00-18:00) of a total of 100 days, 50 days in each of the winter and summer seasons, at 7 randomly selected points in the Akdeniz University Campus. In these measurements, the temperature difference values of 11 cover elements on 7 different floor covering types were measured, and the ambient air temperature, humidity and wind values were also determined. The temperature differences between the areas where the flooring elements are exposed to direct sun and the shadow effect of different plant and cover elements were determined using an infrared laser thermometer. These values were processed with Neural Designer software and possible temperature difference prediction values were created for 57.750 different alternatives with the help of artificial neural network model from 837 sets of data. Evaluation shows that the maximum temperature difference is 15.6°C at noon in the summer months in the red tartan flooring material and Callistemon viminalis cover material. While the artificial neural network model predicts that there will be a high 2-3° C temperature difference for the alternatives, it has made predictions for temperature differences between 0-10°C in winter and 0-16°C in summer months. Although the temperature differences that will occur in the noon hours are distributed over a wide range of values, it seems that the morning and evening forecasts are concentrated between 0-7°C values. Also, it has been determined that the wind and humidity in the environment are more important factors than the ambient temperature in terms of temperature differences.

scholarly journals Predictive modelling of soils’ hydraulic conductivity using artificial neural network and multiple linear regression

2021 ◽  
Vol 3 (2) ◽  
Author(s):  
Charles Gbenga Williams ◽  
Oluwapelumi O. Ojuri
Keyword(s):  
Neural Network ◽  
Artificial Neural Network ◽  
Hydraulic Conductivity ◽  
Linear Regression ◽  
Multiple Linear Regression ◽  
Coarse Grained ◽  
Soil Types ◽  
Wide Range ◽  
Artificial Neural ◽  
Input Variables

AbstractAs a result of heterogeneity nature of soils and variation in its hydraulic conductivity over several orders of magnitude for various soil types from fine-grained to coarse-grained soils, predictive methods to estimate hydraulic conductivity of soils from properties considered more easily obtainable have now been given an appropriate consideration. This study evaluates the performance of artificial neural network (ANN) being one of the popular computational intelligence techniques in predicting hydraulic conductivity of wide range of soil types and compared with the traditional multiple linear regression (MLR). ANN and MLR models were developed using six input variables. Results revealed that only three input variables were statistically significant in MLR model development. Performance evaluations of the developed models using determination coefficient and mean square error show that the prediction capability of ANN is far better than MLR. In addition, comparative study with available existing models shows that the developed ANN and MLR in this study performed relatively better.

scholarly journals Burst Pressure Prediction of API 5L X-Grade Dented Pipelines Using Deep Neural Network

2020 ◽  
Vol 8 (10) ◽  
pp. 766
Author(s):  
Dohan Oh ◽  
Julia Race ◽  
Selda Oterkus ◽  
Bonguk Koo
Keyword(s):  
Neural Network ◽  
Machine Learning ◽  
Artificial Neural Network ◽  
Network Model ◽  
Neural Network Model ◽  
Deep Neural Network ◽  
Machine Learning Techniques ◽  
Burst Pressure ◽  
Comparison Results ◽  
Artificial Neural

Mechanical damage is recognized as a problem that reduces the performance of oil and gas pipelines and has been the subject of continuous research. The artificial neural network in the spotlight recently is expected to be another solution to solve the problems relating to the pipelines. The deep neural network, which is on the basis of artificial neural network algorithm and is a method amongst various machine learning methods, is applied in this study. The applicability of machine learning techniques such as deep neural network for the prediction of burst pressure has been investigated for dented API 5L X-grade pipelines. To this end, supervised learning is employed, and the deep neural network model has four layers with three hidden layers, and the neural network uses the fully connected layer. The burst pressure computed by deep neural network model has been compared with the results of finite element analysis based parametric study, and the burst pressure calculated by the experimental results. According to the comparison results, it showed good agreement. Therefore, it is concluded that deep neural networks can be another solution for predicting the burst pressure of API 5L X-grade dented pipelines.

scholarly journals Using Artificial Neural Networks to Predict Influences of Heterogeneity on Rock Strength at Different Strain Rates

Materials ◽  
2021 ◽  
Vol 14 (11) ◽  
pp. 3042
Author(s):  
Sheng Jiang ◽  
Mansour Sharafisafa ◽  
Luming Shen
Keyword(s):  
Neural Network ◽  
Artificial Neural Network ◽  
Strain Rate ◽  
Network Model ◽  
Neural Network Model ◽  
Artificial Neural Network Model ◽  
Rock Strength ◽  
Strain Rates ◽  
Filling Material ◽  
Artificial Neural

Pre-existing cracks and associated filling materials cause the significant heterogeneity of natural rocks and rock masses. The induced heterogeneity changes the rock properties. This paper targets the gap in the existing literature regarding the adopting of artificial neural network approaches to efficiently and accurately predict the influences of heterogeneity on the strength of 3D-printed rocks at different strain rates. Herein, rock heterogeneity is reflected by different pre-existing crack and filling material configurations, quantitatively defined by the crack number, initial crack orientation with loading axis, crack tip distance, and crack offset distance. The artificial neural network model can be trained, validated, and tested by finite 42 quasi-static and 42 dynamic Brazilian disc experimental tests to establish the relationship between the rock strength and heterogeneous parameters at different strain rates. The artificial neural network architecture, including the hidden layer number and transfer functions, is optimized by the corresponding parametric study. Once trained, the proposed artificial neural network model generates an excellent prediction accuracy for influences of high dimensional heterogeneous parameters and strain rate on rock strength. The sensitivity analysis indicates that strain rate is the most important physical quantity affecting the strength of heterogeneous rock.

Sign in / Sign up

Export Citation Format

Share Document