Electrical Engineering

<p>Shiv Nadar University is a comprehensive, multidisciplinary, research-focused, and student-centric University offering a full range of academic programs at the undergraduate, postgraduate and doctoral level. With state-of-the-art infrastructure, the University comprises of academic wings, sophisticated labs, international standard sports facilities, amphitheaters, auditorium, conference rooms and smart classrooms. The University&rsquo;s goal is to become internationally recognized for the quality of its research and creative endeavors and their applicability to improving quality of life, generating new insights and expanding the boundaries of human knowledge creativity. Committed to excellence in teaching, research and service, the University aims to serve the higher education needs of India and the world beyond.</p>

Shiv Nadar University

In this paper, constrained controller design is proposed for a class of nonlinear discrete-time uncertain system having matched type system uncertainties, using the solution of HJB (Hamilton-Jaccobi-Bellman) equation. The discrete-time HJB equation is formulated using a suitable nonquadratic term in the performance functional to tackle constraints on the control input. Based on the non-quadratic functional, a greedy HDP algorithm is used to obtain the constrained robust-optimal controller. The constrained robust controller requires knowledge of the upper bound of system uncertainties. For facilitating the implementation of the iterative algorithm, two neural networks are used to approximate the value function and to compute the optimal control policy, respectively. Their weights have been tuned using least squares method. Proposed algorithm has been applied on a nonlinear discrete-time system with matched uncertainties. © 2009 AACC.

World University root node

(SoE) School of Engineering

Proceedings of the American Control Conference

Constrained controller design for a class of nonlinear discrete-time uncertain systems

Orthogonal Frequency Division Multiplexing (OFDM) is seen as one of the most promising solution to broadband wireless communications. Its performance depends on the channel state information (CSI) which can be estimated using different channel estimation algorithms. This paper proposes a Set Member Feasibility (SMF) formulation to govern the updating of the adaptive-filter coefficients. Here, twodimensional Set-Membership Normalized LMS (2D-SM-NLMS) algorithm is proposed. The 2D-RLS adaptive channel estimation algorithm is also simulated for comparison. Matlab simulations show that 2D-SM-NLMS and 2D-RLS algorithms have similar Bit Error Rate (BER) performance. But the proposed algorithm has computational complexity of O(N) which is less than that of conventional 2D-RLS algorithm having order O(N2) while compromising on the convergence speed.

Proceedings of INDICON 2009 - An IEEE India Council Conference

Two-dimensional set membership normalized least mean square adaptive channel estimation for OFDM systems

This paper presents Orthogonal Frequency Division Multiplexing (OFDM) based two-way relaying scheme to communicate between two nodes S1 and S2 via MIMO Relay Station (RS) in a frequency selective environment. This scheme relies on two-hop relaying approach which uses two orthogonal channel resources to transmit and receive a signal. This scheme requires the channel state information (CSI) only at RS. Spatial filtering is implemented at the RS for transmit and receive processing which is evaluated using zero forcing (ZF) and minimum mean squared error (MMSE) criterion. MATLAB simulations are conducted to evaluate the validity of the algorithm using both criterions in presence of frequency selective environment. The simulation results depict that the algorithm under defined conditions performs better in a frequency selective fading environment with the use of OFDM. Also, the BER performance evaluated through spatial filtering done using MMSE is found to be better than its ZF counterpart. © 2009 IEEE.

Proceedings - 2009 1st UK-India International Workshop on Cognitive Wireless Systems, UKIWCWS 2009

MIMO two-way relaying in frequency selective environment using OFDM

The effect of material type, yarn count, pick density, type of mordant and layers of fabrics on the electromagnetic shielding properties of textile materials has been studied. The shielding effectiveness of the fabric is measured by estimating the insertion loss incurred to the signal when the sample is placed in the path of the signal at the frequency range 100 MHz-3 GHz inside the coaxial transmission holder. There are some effects of type of material, yarn count, number of fabric layers and type of mordant on electromagnetic shielding effectiveness. But the number of apertures and thread density do not have significant effect on electromagnetic shielding effectiveness, particularly in case of metallic sheets.

Indian Journal of Fibre and Textile Research

Effect of various parameters on electromagnetic shielding effectiveness of textile fabrics

Reinforcement learning (RL) is a popular learning paradigm to adaptive learning control of nonlinear systems, and is able to work without an explicit model. However, learning from scratch, i.e., without any a priori knowledge, is a daunting undertaking, which results in long training time and instability of learning process with large continuous state space. For physical systems, one must consider that the design of controller is very rarely a tabula rasa: some approximate mathematical model of the system is always available. In this paper, our focus is on control applications wherein the system to be controlled is a physical system. We can always obtain at least an approximate mathematical model of the plant to be controlled. We propose a method for hybridization of model-based approach with RL, which is the right solution for such control problems. The superiority of proposed hybrid approach has been established through simulation experiments on a cart-pole balance bench mark problem, comparing it with model-free RL system. We have used fuzzy inference system for function approximation; it can deal with continuous action space in Q-Iearning. Comparison with other function approximators has shown its superiority in terms of robustness of the controller to parameter variations in the plant. ©2009 IEEE.

IEEE International Symposium on Industrial Electronics

Hybridization of model-based approach with reinforcement fuzzy system design

Considerable attention has been given to the design of stable controllers for robot manipulators, in the presence of uncertainties. We investigate here the robust tracking performance of reinforcement learning control of manipulators, subjected to parameter variations and extraneous disturbances. Robustness properties in terms of average error, absolute maximum errors and absolute maximum control efforts, have been compared for reinforcement learning systems using various parameterized function approximators, such as fuzzy, neural network, decision tree, and support vector machine. Simulation results show the importance of fuzzy Q-learning control. Further improvements in this control approach through dynamic fuzzy Q-learning have also been highlighted.

Proceedings of the IEEE International Conference on Industrial Technology

Reinforcement learning control of robot manipulators in uncertain environments

In this paper, a new algorithm is proposed for the constrained control of weakly coupled nonlinear systems. The controller design problem is solved by solving Hamilton-Jacobi-Bellman(HJB) equation with modified cost to tackle constraints on the control input and unknown coupling. In the proposed controller design framework, coupling terms have been formulated as model uncertainties. The bounded controller requires the knowledge of the upper bound of the uncertainty. In the proposed algorithm, Neural Network (NN) is used to approximate the solution of HJB equation using least squares method. Necessary theoretical and simulation results are presented to validate proposed algorithm. © 2009 Springer-Verlag Berlin Heidelberg.

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

Constrained control of weakly coupled nonlinear systems using neural network

Selected Mapping (SLM) technique is one of the promising PAPR reduction techniques for Orthogonal Frequency Division Multiplexing (OFDM). But this reduction technique results in a huge amount of computation complexity which is costly and time consuming. In this paper, Riemann sequences are used as a phase vectors along with the thresholding of the power amplifier(PA). The default index of the phase sequence is used to reduce the computational complexity of SLM technique. The default index of a phase sequence refers to the starting index for the phase vectors in the SLM scheme. This index is derived from the experimentation results. The Matlab simulations show that the complexity has been reduced by 98 percent if the threshold value of PA is around 3.0dB.

A low complexity SLM technique for PAPR reduction in OFDM using Riemann sequence and thresholding of power amplifier

A novel technique, frequency modulated thermal wave imaging (FMTWI) has recently been proposed and demonstrated for non-destructive subsurface defect detection in solids. Its theory and mathematical analysis is provided in this paper illustrating, how the technique simultaneously combines advantages of conventional pulse based thermography as well as lock-in thermography (LT) The specimen is heated for launching thermal waves into the sample, not at a single frequency (Lock-in) or at all frequencies (pulse), but in a desired range of frequencies. While peak power requirement is reduced, phase images obtained retain known advantages. Experimental results are also presented in support. © 2008 American Institute of Physics.

AIP Conference Proceedings

Applications of frequency modulated thermal wave imaging for non-destructive characterization

This paper presents a modeling, simulation and experimentation for detection of material losses and deposited scales in boiler tube specimen by adopting Pulsed Thermography (PT). A data analysis approach, based on variations of slopes in Temperature-Time responses has been adopted, to estimate material losses and subsurface scale detection in the tubes. The slopes are calculated for different pixels in the field of view of the captured image sequence during transient heating cycle of the specimen. Relative thickness variations due to material loss and the deposited scales of boiler tube have been estimated from the calculated slopes. Experimental results are presented in support. Copyright © 2008 American Scientific Publishers. All rights reserved.