Madhu T.,SIET |
Suryakalavathi M.,Jawaharlal Nehru Technological University
2013 IEEE Conference on Information and Communication Technologies, ICT 2013 | Year: 2013
Evolutionary Algorithms (EA) are used in many optimization problems such as the Artificial Neural Networks (ANN). But the main challenging issue of using these algorithms is the time taken for computing the function value, implementation and verification of hardware architecture design. In this paper we propose to implement the architecture design of Back Propagation Neural (BPN) networks using Very High Speed Integrated Circuits Hardware Description Language (VHDL). The simulation is carried out using Xilinx Spartan 3E. © 2013 IEEE.
Hota S.S.,SIET |
International Journal of Acoustics and Vibrations | Year: 2011
The hallmark of this paper is the adoption of the constraint technique, in combination with a subparametric, triangular-plate bending element of first-order shear deformation, to maintain uniform mesh size and shape even while dealing with cutouts of arbitrary shapes. The evolution of two cutout models in the present investigation is a distinct improvement over the existing practices of cutout analysis. The use of matching polynomials offers he scope of eliminating the hazards of locking and spurious zero energy modes, while solving problems of very thin plates. Benchmark examples, as well as the author's own problems on free vibration of rectangular plates with different shapes of cutouts, have been solved to exhibit versatility of these models. Mode shapes of plates with different shapes of cutout have also been provided.
Srinivas V.,SIET |
Santhi Rani C.H.,Dms And Svh Engineering College
ICIIECS 2015 - 2015 IEEE International Conference on Innovations in Information, Embedded and Communication Systems | Year: 2015
In this paper, we proposed a new technique presents an extraction method for robust speech recognition using the MVDR (Minimum Variance Distortionless Response) spectrum of short time autocorrelation sequence which can reduce the effects of leftover of the nonstationary additive noise that remains after filtering the autocorrelation. To produce a further robust front-end, we present the customized robust distortionless constraint of the MVDR spectral estimation method through revised weighting of the subband power spectrum values based on the sub-band signal to noise ratios (SNRs), which adjust it to the new proposed technique. The new proposed functions allow the components of the input signal at the frequencies with minimum affected by noise to pass with better weights and attenuate more effectively the noisy and unwanted components. This revision results in decrease of the noise residuals of the projected spectrum from the filtered autocorrelation sequence, thus advancing to a more robust algorithm. Our proposed technique, when analyzed on Aurora 2 task for recognition applications, best performed all MFCC (Mel frequency cepstral coefficients) as the fundamental, respective autocorrelation sequence MFCC (RAS MFCC), and the proposed MVDR related features in numerous different noisy conditions. © 2015 IEEE.
Kar B.P.,GIET |
Nayak S.K.,VSSUT |
Advances in Intelligent Systems and Computing | Year: 2016
Artificial neural network (ANN) based forecasting models have been established their efficiencies with improved accuracies over conventional models. Evolutionary algorithms (EA) are used most frequently by the researchers to train ANN models. Population initialization of EA can affect the convergence rate as well as the quality of optimal solution. Random population initialization of EAs is the most commonly used technique to generate candidate solutions. This paper presents an opposition-based genetic algorithm (OBGA) learning to generate initial candidate solutions for ANN based forecasting models. The present approach is based on the concept that, it is better to begin with some fitter candidate solutions when no a priori information about the solution is available. In this study both GA and OBGA optimizations are used to optimize the parameters of a multilayer perceptron (MLP) separately. The efficiencies of these methods are evaluated on forecasting the daily closing prices of some fast growing stock indices. Extensive simulation studies reveal that OBGA method outperforms other with better accuracies and convergence speed. © Springer India 2016.
Kumar D.,BIT |
Singh T.,SIET |
Dwivedi R.,BIT |
Proceedings - 4th International Conference on Computational Intelligence and Communication Networks, CICN 2012 | Year: 2012
This paper presents the design of compact CPW-FED Dual band notched UWB square ring antenna. The main objective of this proposed research work is to reduce the size of antenna and avoid interference between UWB and WLAN (5.5GHz) and WiMax (3.4 GHz) application. The antenna consists of a square-ring metal patch and 50O coplanar wave guide (CPW) strip with dimension of 35x31 mm2. The antenna is then modified to possess dual band rejection by etching two symmetrical slots in the ground plane nearby the feed line and rectangular split ring slot in the square-ring radiating patch. The geometry parameters of antenna are investigated and optimized with HFSS. The result show that the proposed antenna achieves an impedance bandwidth of 3.1-10.6GHz with VSWR<2, except in the band of 4.8 - 6.2 GHz and 3.2-3.6GHz with an omnidirectional radiation pattern. © 2012 IEEE.
Rout M.,SIET |
Noise and Vibration Worldwide | Year: 2010
Development of an early damage detection method for delamination is one of the most important keys in maintaining the integrity and safety of composite structures. The present study focuses on the effects of delamination on the vibration characteristics of beams with single through-width delamination. Comparison is made between experimental study on aluminum specimens and corresponding finite element models created in ANSYS using brick elements. Specimens are made by bonding thin aluminum strips using very thin adhesive layers, with delaminations introduced by inserting cellophane paper at pre-determined locations. Vibration tests are carried out on the specimens, the results of which are compared with those of the finite element analysis. The finite element model was then used to demonstrate the comparison of natural frequency, mode shapes and curvature mode shapes of the intact and delaminated models. Further, the analysis is extended into layered composites of orthotropic materials. Comparisons are made between published experimental results of laminates and results of finite element analyses on models with layered brick elements using ANSYS. Eigen value extraction was done on 'free' models, with no constrains between the delaminated layers and models with linear spring connecting the delaminated layers.
Sahu B.N.,Siksha ‘O’ Anusandhan University |
Mohanty M.N.,Siksha ‘O’ Anusandhan University |
Padhi S.K.,Siksha ‘O’ Anusandhan University |
2015 International Conference on Communication and Signal Processing, ICCSP 2015 | Year: 2015
Tasks of system identification has occupied an important space in research field for development of automated system. Artificial neural network (ANN) model is most suitable for analysis of dynamic systems. It has been exploited in this work as an alternative approach for such task. The objective of this paper is to design a novel technique to improve the performance of the existing techniques. Adaptive learning algorithm is applied with the sliding mode strategy for the neuron models. It is considered for the first-order dynamic system with adjustable parameters. It can perform for faster convergence with robust characteristics. It has been chosen as suitable alternative for nonlinear system identification as it has good function approximation capabilities. It has been shown that the proposed ANN model can be used to model the complex dynamic systems. Also the performance analysis has been done using different methods like Sliding Mode strategy, MLP-Back propagation, FLANN-LMS and compared for system identification. © 2015 IEEE.
Ramesh Raju N.,SIET |
Linga Reddy P.,Koneru Lakshmaiah College of Engineering
International Journal of Engineering and Technology | Year: 2015
PID controller is mostly used in process plants to control the system performance by properly choosing its parameters. The optimum PID parameters can be obtained in offline using genetic algorithm if the mathematical model of the system is exactly known. In all process plants the process parameters such as properties of materials like thermal conductivity, electrical conductivity, physical dimensions such as diameter, length of the pipes, parameters of valves and pumps will change as time runs. This happens due to corrosion, scaling, aging, repairs during the maintenance, wear and tear. When the system is robust these changes slightly affect the performance of the system. When the system is not robust they make the system performance worst. Due to above reasons the process plant parameters changes as time runs. It is not easy to measure the changes in system parameters while plant is running and could not be evaluated optimum PID parameters through mathematical model. In this paper a new approach using genetic algorithm and neural network is established for optimum self tuning of PID parameters by observing the time response of the system at any time while plant is running.
Singh N.,SIET |
Tripathi A.,Motilal Nehru National Institute of Technology
Advances in Intelligent Systems and Computing | Year: 2015
As the software size grows, the maintenance become challenging. To make it easier, there is a need to measure some quality parameters in earlier phases of software development. Understandability has a major contribution to control the maintainability. Coupling and cohesion are two well-accepted parameters to measure the software quality parameters. In this paper, a model is proposed to measure the understandability that is based on coupling and cohesion. © Springer India 2015.
Ramesh Raju N.,SIET |
Linga Reddy P.,Koneru Lakshmaiah College of Engineering
International Journal of Engineering and Technology | Year: 2016
In this paper a Fractional order PID controller is proposed for AVR system and its parameters are optimised through Genetic Algorithm. Results are obtained by simulation in MATLAB/SIMULINK environment with FOMCON software. The results show that the AVR system with fractional order PID controller is faster and robust compared to integer order PID controller.