An adaptive wavelet differential neural networks based identifier and its stability analysis

Jahangiri, Fatemeh; Doustmohammadi, Ali; Menhaj, Mohammad Bagher

doi:10.1016/j.neucom.2011.07.013

Cited by 15 publications

(8 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The optimal weight derived from this stage is used to generate an initial 100 population for a GA, which was run for 500 generations to obtain the optimal values of the unknown parameters of the WNN. This is a case of non-linear MIMO system, which is more challenging than most recently reported SISO in literature [8,18,28,41] , and the training approach was adopted for the following reasons in order to avoid the problem of being trapped in a local minimum solution. Moreover, it was shown after some heuristic study that the proposed is a faster approach.…”

Section: The Term δϕX δXmentioning

confidence: 99%

A wavelet neural network based non-linear model predictive controller for a multi-variable coupled tank system

Owa

Sharma

Sutton

2014

Int. J. Autom. Comput.

View full text Add to dashboard Cite

Abstract:In this paper, a novel real time non-linear model predictive controller (NMPC) for a multi-variable coupled tank system (CTS) is designed. CTSs are highly non-linear and can be found in many industrial process applications. The involvement of multi-input multi-output (MIMO) system makes the design of an effective controller a challenging task. MIMO systems have inherent couplings, interactions in-between the process input-output variables and generally have an complex internal structure. The aim of this paper is to design, simulate, and implement a novel real time constrained NMPC for a multi-variable CTS with the aid of intelligent system techniques. There are two major formidable challenges hindering the success of the implementation of a NMPC strategy in the MIMO case. The first is the difficulty of obtaining a good non-linear model by training a non-convex complex network to avoid being trapped in a local minimum solution. The second is the online real time optimisation (RTO) of the manipulated variable at every sampling time. A novel wavelet neural network (WNN) with high predicting precision and time-frequency localisation characteristic was selected for an MIMO model and a fast stochastic wavelet gradient algorithm was used for initial training of the network. Furthermore, a genetic algorithm was used to obtain the optimised parameters of the WNN as well as the RTO during the NMPC strategy. The proposed strategy performed well in both simulation and real time on an MIMO CTS. The results indicated that WNN provided better trajectory regulation with less mean-squared-error and average control energy compared to an artificial neural network. It is also shown that the WNN is more robust during abnormal operating conditions.

show abstract

Section: The Term δϕX δXmentioning

confidence: 99%

A wavelet neural network based non-linear model predictive controller for a multi-variable coupled tank system

Owa

Sharma

Sutton

2014

Int. J. Autom. Comput.

View full text Add to dashboard Cite

show abstract

“…Methods based on neural networks are effective when applied to the identification of dynamic nonlinear systems, which are also effective in combinatorial optimization [15], optimization of non-stationary functions [16], multi-objective optimization [17], bioinformatics [18]. The WNN with the merits of local optimization, approximation and self-learning in the time-frequency domain provides a framework that can be successfully applied to the dynamic models [19][20][21][22] and the nonlinearity control [23][24][25][26]. In [19], the wavelet representation with high-efficiency coupled map lattices trains the basic structure of the WNN, and a novel two-stage hybrid training scheme is developed for constructing a parsimonious WNN to improve the identification efficiency.…”

Section: Introductionmentioning

confidence: 99%

“…In [19], the wavelet representation with high-efficiency coupled map lattices trains the basic structure of the WNN, and a novel two-stage hybrid training scheme is developed for constructing a parsimonious WNN to improve the identification efficiency. A differential evolution algorithm in [20] and a storage strategy in [21] are used to optimize the WNN parameters in various ways, which guarantees the performance and robustness of the system in the sense of Lyapunov stability.…”

Section: Introductionmentioning

confidence: 99%

“…In this paper, the identification method of the ADE-VSWNN is proposed. The crossover and mutation operation, which optimizes the scale factors, the translation factors, and the weights of the WNN [20] is realized by using the individual genetic difference algorithm. The DE algorithm has the advantages of fast speed, strong robustness and wide search ability in the real domain, but its search performance is affected by the individual variations.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

The identification of electric load simulator for gun control systems based on variable-structure WNN with adaptive differential evolution

Wang

Hou

RongzhongLiu

et al. 2016

Applied Soft Computing

View full text Add to dashboard Cite

Owing to the complex nonlinearities of the electric load simulator (ELS) for the gun control system (GCS), the surplus torque plays a great negative impact on the performance of the loading system. This paper proposes a variable-structure wavelet-neural-network (VSWNN) identification strategy based on adaptive differential evolution (ADE). First of all, a mathematical model is established based on the structure and the working principle of the ELS. Then an intelligent identification method is applied, where the wavelet function is chosen as the excitation function, which improves the generalization and approximation ability of the neural network. The ADE is used to optimize the parameters, which solves the difficulty of determining the structure of the WNN. In order to reduce the computation complexity and speed up the convergence of the identification system, the adaptive laws of the pitch adjusting rate (PAR), band width (BW) and variable numbers of neurons are proposed.Finally, a pseudo random multilevel signal and a linear frequency modulation signal are chosen as input signals for the hardware-in-the-loop simulation. The test results show that the proposed ADE-VSWNN algorithm has superior validity and practicability, especially when the identification algorithm is used in the working circumstances with different inertial torque. Further, the high precision and strong robustness of the identification algorithm are further verified.

show abstract

“…As a result, training algorithms for WNNs typically converge in a smaller number of iterations when compared to ANNs; (ii) It has been argued that for a given quality of approximation, fewer nodes may be required for WNNs, when compared to ANNs using sigmoidal functions. Moreover, many researchers have studied the properties and applications of the WNNs models to represent nonlinear systems [15][16][17][18][19][20][21].…”

Section: Introductionmentioning

confidence: 99%

Wavelet neural network approach applied to biomechanics of swimming

Puchalski

Schauenburg

Mariani

et al. 2013

2013 13th UK Workshop on Computational Intelligence (UKCI)

View full text Add to dashboard Cite

An artificial neural network (ANN) consists of a number of interconnecting artificial neurons and employs mathematical or computational models for information processing. ANNs are suitable for handling large amounts of dynamic, noisy and nonlinear data. On the other hand, the wavelet theory provides a multi-resolution approximation for discriminate functions. The combination of the wavelet transforms theory with the basic concept of ANNs leads to new mapping networks called wavelet neural networks (WNNs) or wavenets, which are proposed as an alternative to feedforward ANNs for approximating arbitrary nonlinear functions. Generalized from radial basis function ANNs, WNNs are in fact feed-forward neural networks with one hidden layer, radial wavelets as activation functions in the hidden nodes and a linear output layer. The contribution of this paper is to evaluate the WNNs to model a parathlete swimmer behavior. The parathlete swimmer swims the breaststroke style using biomechanics data generated by the software tool called SWUMSUIT, which was developed in Tokyo Technological Institute in Japan. The forecasted results clearly show that WNN has good prediction properties. The proposed WNN modeling approach can benefit disabled swimmers (parathletes) to gain competitive advantage by studying the biomechanics involved in the sport and considering the help of simulations systems.

show abstract

An adaptive wavelet differential neural networks based identifier and its stability analysis

Cited by 15 publications

References 20 publications

A wavelet neural network based non-linear model predictive controller for a multi-variable coupled tank system

A wavelet neural network based non-linear model predictive controller for a multi-variable coupled tank system

The identification of electric load simulator for gun control systems based on variable-structure WNN with adaptive differential evolution

Wavelet neural network approach applied to biomechanics of swimming

Contact Info

Product

Resources

About