Deep Convolutional Neural Network (CNN) is a special type of Neural Networks, which has shown exemplary performance on several competitions related to Computer Vision and Image Processing. Some of the exciting application areas of CNN include Image Classification and Segmentation, Object Detection, Video Processing, Natural Language Processing, and Speech Recognition. The powerful learning ability of deep CNN is primarily due to the use of multiple feature extraction stages that can automatically learn representations from the data. The availability of a large amount of data and improvement in the hardware technology has accelerated the research in CNNs, and recently interesting deep CNN architectures have been reported. Several inspiring ideas to bring advancements in CNNs have been explored, such as the use of different activation and loss functions, parameter optimization, regularization, and architectural innovations. However, the significant improvement in the representational capacity of the deep CNN is achieved through architectural innovations. Notably, the ideas of exploiting spatial and channel information, depth and width of architecture, and multi-path information processing have gained substantial attention. Similarly, the idea of using a block of layers as a structural unit is also gaining popularity. This survey thus focuses on the intrinsic taxonomy present in the recently reported deep CNN architectures and, consequently, classifies the recent innovations in CNN architectures into seven different categories. These seven categories are based on spatial exploitation, depth, multi-path, width, feature-map exploitation, channel boosting, and attention. Additionally, the elementary understanding of CNN components, current challenges, and applications of CNN are also provided.

https://link.springer.com/content/pdf/10.1007/s10462-020-09825-6.pdf

A survey of the recent architectures of deep convolutional neural networks

The dynamic behavior of rigid-block structures resting on a rigid foundation subjected to horizontal harmonic excitation is examined. For slender structures, the nonlinear equation of motion is approximated by a piecewise linear equation. Using this approximation for an initially quiescent structure, safe or no-toppling and unsafe regions are identified in an excitation amplitude versus excitation frequency plane. Furthermore, several possible modes of steady-state response are detected, and analytical procedures are developed for determining the amplitudes of the predominant modes and for performing stability analyses. It is shown that the produced stability diagrams can be beneficial to assessing the toppling potential of a rigid-block structure under a given amplitude-frequency combination of harmonic excitation; in this manner the integration of the equation of motion is circumvented.

Closure of "Rocking of Rigid Blocks Due to Harmonic Shaking"

This paper assesses the seismic fragility of single degree of freedom rocking structures within a probabilistic framework. The focus is on slender rigid structures that exhibit negative stiffness during rocking. The analysis considers ground motions with near-fault characteristics, either solely coherent pulses or synthetic ground motions that include, in addition, a stochastic high-frequency component. The study offers normalized fragility curves that estimate the overturning tendency, as well as the peak response rotation of a rocking structure. It shows that the use of bivariate intensity measures (IM s) can lead to superior fragility curves compared with conventional univariate IM s. Regardless, the study advocates the use of dimensionless–orientationless IM s that offer an approximately ‘universal’ description of rocking behavior/fragility, a normalized description almost indifferent to the amplitude and the predominant frequency of the excitation or the size and the slenderness of the rocking structure. Importantly, the analysis unveils hidden order in rocking response. There exists a critical peak ground acceleration, below and above which, peak rocking response scales differently. In particular, when the structure does not overturn, the peak rotation follows approximately a biplanar pattern with respect to the intensity and the predominant frequency of the excitation. Finally, the analysis verifies that rocking overturning depends primarily on the velocity characteristics of the ground motion. Copyright © 2015 John Wiley & Sons, Ltd.

Dimensionless fragility curves for rocking response to near‐fault excitations

This paper presents the results of a series of shake table tests of two 460-mm-diameter columns supported on 1.5-meter-square shallow rocking foundations. The tests were conducted using the Large Outdoor High-Performance Shake Table of the Network for Earthquake Engineering Simulation at the University of California at San Diego. The first specimen was aligned with the uniaxial direction of shaking, and the second was positioned in a skew configuration. The specimens were placed inside a soil-confining box 10.1 m long and 4.6 m wide with a 3.4-m height of clean sand compacted at 90% relative density. Three series of tests were performed; each had different groundwater and backfill conditions. The test protocols included up to six historical ground motions and resulted in peak drift ratios up to 13.8%. For peak drift ratios up to 6.9%, the rocking foundations performed very well, with residual drift ratios between 0.5 and 0.9% depending on the backfill conditions and with minimal settlements and no...

Shake Table Test of Large-Scale Bridge Columns Supported on Rocking Shallow Foundations

Deep learning techniques for estimation of the yield and size of citrus fruits using a UAV

This paper numerically investigates the three-dimensional seismic response of six reinforced concrete bridges hypothetically located in Oakland, California, 3 km from the Hayward fault. Three of the bridges are 17 m tall and three are 8 m tall. Three types of column-foundation designs are studied: (1) columns that form flexural plastic hinges, which are conventionally designed according to CALTRANS seismic design criteria; (2) columns on rocking pile foundations that are designed to remain elastic; and (3) columns designed to remain elastic that are supported on rocking shallow foundations. The bridges with rocking foundations use lead rubber bearings at the abutments to enhance strength, stiffness, and hysteretic energy dissipation. Using two components of horizontal excitation, three-dimensional nonlinear response history analyses are performed for two seismic hazard levels with return periods of 975 and 2,475 years, respectively. At both levels of shaking, the conventionally designed bridges experience substantial inelastic deformations and damage in the columns, whereas the bridges with rocking foundations result in negligible structural damage and a nominally elastic response and small residual deformations.

Seismic Response of Bridges with Rocking Foundations Compared to Fixed-Base Bridges at a Near-Fault Site

.......................................................................................................................... iii ACKNOWLEDGMENTS ............................................................................................................ v TABLE OF CONTENTS ........................................................................................................... vii LIST OF FIGURES ..................................................................................................................... ix LIST OF TABLES ..................................................................................................................... xiii

Seismic Design and Performance of Bridges with Columns on Rocking Foundations

The objectives of this dissertation is to investigate the use of rocking foundations in bridges for enhanced seismic design and performance and the reduction of post-earthquake damage. The seismic response of bridge systems was studied numerically using three-dimensional nonlinear models, whereas bridge columns with rocking foundations and superstructure mass were studied both numerically and experimentally. The experimental part consisted of the shake-table testing of large scale bridge columns with shallow rocking foundations using physical modeling of the underlying soil. Using the data from these tests, a three-dimensional model with Winkler springs was modified and validated for rocking shallow foundations designed with high factors of safety against vertical loads. The proposed model was then used on a parametric study to investigate the seismic demand on a large variety of bridge piers with rocking shallow foundations.

/pdf/numerical-and-experimental-investigation-of-bridges-with-1t4j6lohpu.pdf

Grigorios Antonellis

Papers

Shake Table Test of Large-Scale Bridge Columns Supported on Rocking Shallow Foundations

Seismic Response of Bridges with Rocking Foundations Compared to Fixed-Base Bridges at a Near-Fault Site

Seismic Design and Performance of Bridges with Columns on Rocking Foundations

Numerical and Experimental Investigation of Bridges with Columns Supported on Rocking Foundations

Shake table test response of large-scale bridge columns supported on rocking shallow foundations