The Levenberg-Marquardt Neural Network model of the PEMFC's MEA

Author(s):  
Erkan Dursun ◽  
Osman Kilic
2011 ◽  
Vol 71-78 ◽  
pp. 4103-4108
Author(s):  
Yu Zhou Jiang ◽  
Rui Hong Wang ◽  
Jie Bing Zhu

Rheological experiments were carried out for sandstone and marble specimens from left bank high slope of Jingping First Stage Hydropower Project by using the rock servo-controlling rheology testing machine. Typical triaxial rheological curves under step loading and temperature curves in the process of rheological experiment were gained. BP neural network is improved by Levenberg-Marquardt algorithm. Improved neural network model for rock rheology is established in accordance with the rheology experimental results of rock specimen. The improved neural network model was used to forecast rock rheological experimental curves, and the result shows that the forecasted rock rheology curves are closely accorded with the experimental result. The improved neural network model takes into account the influence of loading history and temperature difference on the rock rheological deformation, and the forecasted result can reflect better the rheology deformation behavior of rock material.


Author(s):  
Revathy Jayaseelan ◽  
Gajalskshmi Pandulu ◽  
Ashwini G

This paper presents the prediction of fresh concrete properties and compressive strength of flowable concrete through neural network approach. A comprehensive data set was generated from the experiments performed in the laboratory under standard conditions. The flowable concrete was made with two different types of micro particles and with single nano particles. The input parameter was chosen for the neural network model as cement, fine aggregate, coarse aggregate, superplasticizer, water-cement ratio, micro aluminium oxide particles, micro titanium oxide particles, and nano silica. The output parameter includes the slump Flow, L-Box flow, V Funnel flow and compressive strength of the flowable concrete. To develop a suitable neural network model, several training algorithms were used such as BFGS Quasi- Newton back propagation, Fletcher-Powell conjugate gradient back propagation, Polak - Ribiere conjugate gradient back propagation, Gradient descent with adaptive linear back propagation and Levenberg-Marquardt back propagation. It was found that BFGS Quasi- Newton back propagation and Levenberg-Marquardt back propagation algorithm provides more than 90% on the prediction accuracy. Hence, the model performance was agreeable for prediction purposes for the fresh properties and compressive strength of flowable concrete.


2013 ◽  
Vol 2013 ◽  
pp. 1-14 ◽  
Author(s):  
A. J. Litta ◽  
Sumam Mary Idicula ◽  
U. C. Mohanty

Forecasting thunderstorm is one of the most difficult tasks in weather prediction, due to their rather small spatial and temporal extension and the inherent nonlinearity of their dynamics and physics. Accurate forecasting of severe thunderstorms is critical for a large range of users in the community. In this paper, experiments are conducted with artificial neural network model to predict severe thunderstorms that occurred over Kolkata during May 3, 11, and 15, 2009, using thunderstorm affected meteorological parameters. The capabilities of six learning algorithms, namely, Step, Momentum, Conjugate Gradient, Quick Propagation, Levenberg-Marquardt, and Delta-Bar-Delta, in predicting thunderstorms and the usefulness for the advanced prediction were studied and their performances were evaluated by a number of statistical measures. The results indicate that Levenberg-Marquardt algorithm well predicted thunderstorm affected surface parameters and 1, 3, and 24 h advanced prediction models are able to predict hourly temperature and relative humidity adequately with sudden fall and rise during thunderstorm hour. This demonstrates its distinct capability and advantages in identifying meteorological time series comprising nonlinear characteristics. The developed model can be useful in decision making for meteorologists and others who work with real-time thunderstorm forecast.


Sign in / Sign up

Export Citation Format

Share Document