Faculty Publications

Now showing 1 - 4 of 4

Shape optimization of steel reinforced concrete beams
(Techno Press, 2007) Babu Narayan, K.S.B.; Venkataramana, K.
Steel reinforced concrete is perhaps the most versatile and widely used construction material. The versatility is attributed to mouldability of concrete to any conceivable shape. The inherent property of cracking of concrete is the reason for its low tensile strength and hence the design approach of RCC sections in flexure adopts the cracked section theory where in concrete in tension zone is ignored. Means, modes and methods of exploitation of concrete strength by conceiving shapes other than rectangular whereby ineffective concrete in tension zone is reduced and incorporated in compression zone where it is effective needs consideration. Shape optimization of beams is attempted in this analytical investigation employing Sequential Unconstrained Minimization Technique (SUMT). The results clearly show that trapezoidal beams happen to be less costlier than their rectangular counterparts, their usage needs serious reconsideration by the designers.
Image despeckling with non-local total bounded variation regularization
(Elsevier Ltd, 2018) Padikkal, P.; Banothu, B.
A non-local total bounded variational (TBV) regularization model is proposed for restoring images corrupted with data-correlated speckles and linear blurring artifacts. The energy functional of the model is derived using maximum a posteriori (MAP) estimate of the noise probability density function (PDF). The non-local total bounded variation prior regularizes the model while the data fidelity is derived using the MAP estimator of the noise PDF. The computational efficiency of the model is improved using a fast numerical scheme based on the Augmented Lagrange formulation. The proposed model is employed to restore ultrasound (US) and synthetic aperture radar (SAR) images, which are usually speckled and blurred. The numerical results are presented and compared. Furthermore, a detailed theoretical study of the model is performed in addition to the experimental analysis. © 2017 Elsevier Ltd
Optimizing Reinforcement Learning-Based Visual Navigation for Resource-Constrained Devices
(Institute of Electrical and Electronics Engineers Inc., 2023) Vijetha, U.; Geetha, V.
Existing work on Deep reinforcement learning-based visual navigation mainly focuses on autonomous agents with ample power and compute resources. However, Reinforcement learning for visual navigation on resource-constrained devices remains an under-explored area of research, primarily due to challenges posed by processing high-dimensional visual inputs and making prompt decisions in realtime scenarios. To address these hurdles, we propose a State Abstraction Technique (SAT) designed to transform high-dimensional visual inputs into a compact representation, enabling simpler reinforcement learning agents to process the information and learn effective navigation policies. The abstract representation generated by SAT effortlessly serves as a versatile intermediary that bridges the gap between simulation and reality, enhancing the transferability of learned policies across various scenarios. Additionally, our reward shaping strategy uses the data provided by SAT to maintain a safe distance from obstacles, further improving the performance of navigation policies on resource-constrained devices. Our work opens up opportunities for navigation assistance and other applications in a variety of resource-constrained domains, where computational efficiency is crucial for practical deployment, such as guiding miniature agents on embedded devices or aiding visually impaired individuals through smartphone-integrated solutions. Evaluation of proposed approach on the AI2-Thor simulated environment demonstrates significant performance improvements over traditional state representations. The proposed method provides 84.18% fewer collisions, 28.96% fewer movement instructions and 11.3% higher rewards compared to the best alternative options available. Furthermore, we carefully account for real-world challenges by considering noise and motion blur during training, ensuring optimal performance during deployment on resource-constrained devices. © 2013 IEEE.
Constrained radar waveform optimization for a cooperative radar-communication system
(Elsevier B.V., 2023) Mahipathi, A.C.; Gunnery, S.; Srihari, P.; D'Souza, J.; Jena, P.
The coexistence of radar-sensing and communication systems research has received a surge of interest in recent times to tackle the issue of spectrum inadequacy. Designing an optimized radar waveform for a coexistence scenario has been a challenging task for accomplishing the convergence of radar-sensing and communication functionalities, without degrading the performance at either end. This paper proposes a novel global optimization-based Spatial Branch and Bound (SBnB) approach to optimize the phase coefficients of a Non-Linear Frequency Modulated (NLFM) waveform in a CRCS framework. In addition, the Modified-Power Ratio Constraint-Cramér–Rao Lower Bound (M-PRC-CRLB), a local optimization-based approach is proposed to optimize the phase coefficients of an NLFM waveform. The spectral energy distribution and auto-correlation characteristics of an NLFM waveform are comprehensively investigated for various values of polynomial order (N) and at different threshold Signal-to-Noise-Ratio (SNR) values. To compare the proposed waveform design approaches (M-PRC-CRLB, SBnB) with the existing waveform design approaches namely, Minimum Estimation Error Variance (MEEV) and PRC- CRLB, a Peak-to-Side-Lobe-Ratio (PSLR), and Integrated-Side-Lobe-Ratio (ISLR) are evaluated at various polynomial orders and threshold SNR values. Furthermore, the performance of a CRCS is assessed using the radar estimation rate and communication data rate. The simulation results reveal that the proposed optimized radar waveform design approaches provide improved performance compared to the existing radar waveform design approaches in terms of radar estimation rate. Further, the proposed global optimization-based SBnB approach achieves a comparable performance of the communication data rate. In addition, the proposed approaches accomplish enhanced spectral utilization, controlled side-lobe energy levels, reduced range-domain ambiguities, and a higher information rate in a CRCS. © 2022 Elsevier B.V.

Faculty Publications

Browse

Filters

Settings

Sort By

Results per page

Search Results