Comparing the performance of neural networks developed by using Levenberg–Marquardt and Quasi-Newton with the gradient descent algorithm for modelling a multiple response grinding process

Mukherjee, Indrajit; Routroy, Srikanta

doi:10.1016/j.eswa.2011.08.087

Cited by 136 publications

(54 citation statements)

References 43 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…In classical ANNs, the training of the weights is accomplished using Gradient Descent (GD), a gradient based algorithm. Recently, however, Levenberg-Marquardt (LM) algorithm has instead been used for the training purpose as it has shown to outperform GD in a variety of problems [7,8,9,10]. As LM is still a gradient based technique, it can still converge to the local minimum depending on the initial weight values.…”

Section: Modellingmentioning

confidence: 99%

Modelling of Decision Making in the Production of Stator Core Using a GA-ANN Approach

Rajora¹,

Zou²,

Xu³

et al. 2017

dtcse

View full text Add to dashboard Cite

Abstract. In the production of stator cores, it is relied upon experienced engineers to make time sensitive decisions on the number of compensation sheets to be added to achieve uniform pressure distribution though out the laminations. However, this method yields inconsistent results as humans are unable to store and analyze large amounts of data. In this paper ANNs have been employed to help the engineers with the decision making process. The ANNs are trained using a hybrid Genetic Algorithm (GA) -Levenberg-Marquardt (LM) to avoid local convergence. When used on testing data sets, the ANNs displayed a high degree of prediction accuracy indicating their ability to simulate the decision making process of these experienced engineers.

show abstract

Section: Modellingmentioning

confidence: 99%

Modelling of Decision Making in the Production of Stator Core Using a GA-ANN Approach

Rajora¹,

Zou²,

Xu³

et al. 2017

dtcse

View full text Add to dashboard Cite

show abstract

“…MLPs utilize the back-propagation (BP) learning technique in conjunction with an optimization method such as gradient descent and Levenberg-Marquardt for training. 105 At completion of a training process, the MLP is capable of giving output solution for any new input based on the generalized mapping that has been developed. 106 2.…”

Section: Ann-based Modelsmentioning

confidence: 99%

Multidimensional prognostics for rotating machinery: A review

Duan

Mba

et al. 2017

Advances in Mechanical Engineering

View full text Add to dashboard Cite

Determining prognosis for rotating machinery could potentially reduce maintenance costs and improve safety and availability. Complex rotating machines are usually equipped with multiple sensors, which enable the development of multidimensional prognostic models. By considering the possible synergy among different sensor signals, multivariate models may provide more accurate prognosis than those using single-source information. Consequently, numerous research papers focusing on the theoretical considerations and practical implementations of multivariate prognostic models have been published in the last decade. However, only a limited number of review papers have been written on the subject. This article focuses on multidimensional prognostic models that have been applied to predict the failures of rotating machinery with multiple sensors. The theory and basic functioning of these techniques, their relative merits and drawbacks and how these models have been used to predict the remnant life of a machine are discussed in detail. Furthermore, this article summarizes the rotating machines to which these models have been applied and discusses future research challenges. The authors also provide seven evaluation criteria that can be used to compare the reviewed techniques. By reviewing the models reported in the literature, this article provides a guide for researchers considering prognosis options for multi-sensor rotating equipment.

show abstract

“…The former require some partial derivatives to be calculated to tune the parameters of FNN, while the latter do not need derivative information in order to update the parameters of FNN. Gradient descent (GD) (Mukherjee and Routroy, 2012) and genetic algorithms (GAs) (MartinezMartinez et al, 2015) are the most widely used approaches among the existing derivative-based and derivative-free training methods, respectively. However, GD training algorithms are based on the first-order Taylor expansion of a nonlinear function.…”

Section: Accepted Manuscriptmentioning

confidence: 99%

Novel Levenberg–Marquardt based learning algorithm for unmanned aerial vehicles

Sarabakha¹,

Imanberdiyev²,

Kayacan

et al. 2017

Information Sciences

View full text Add to dashboard Cite

In this paper, a Levenberg-Marquardt inspired sliding mode control theory based adaptation laws are proposed to train an intelligent fuzzy neural network controller for a quadrotor aircraft. The proposed controller is used to control and stabilize a quadrotor unmanned aerial vehicle in the presence of periodic wind gust. A proportional-derivative controller is firstly introduced based on which fuzzy neural network is able to learn the quadrotor's control model on-line. The proposed design allows handling uncertainties and lack of modelling at a computationally inexpensive cost. The parameter update rules of the learning algorithms are derived based on a Levenberg-Marquardt inspired approach, and the proof of the stability of two proposed control laws are verified by using the Lyapunov stability theory. In order to evaluate the performance of the proposed controllers extensive simulations and real-time experiments are conducted. The 3D trajectory tracking problem for a quadrotor is considered in the presence of time-varying wind conditions.

show abstract

Comparing the performance of neural networks developed by using Levenberg–Marquardt and Quasi-Newton with the gradient descent algorithm for modelling a multiple response grinding process

Cited by 136 publications

References 43 publications

Modelling of Decision Making in the Production of Stator Core Using a GA-ANN Approach

Modelling of Decision Making in the Production of Stator Core Using a GA-ANN Approach

Multidimensional prognostics for rotating machinery: A review

Novel Levenberg–Marquardt based learning algorithm for unmanned aerial vehicles

Contact Info

Product

Resources

About