• Journal of Power System Engineering
• /
• v.6 no.1
• /
• pp.96-101
• /
• 2002

In this paper, we address the problem of controlling an end-effector to track and grab a moving target using the visual servoing technique. A visual servo mechanism based on the image-based servoing principle, is proposed by using visual feedback to control an end-effector without calibrated robot and camera models. Firstly, we consider the control problem as a nonlinear least squares optimization and update the joint angles through the Taylor Series Expansion. And to track a moving target in real time, the Jacobian estimation scheme(Dynamic Broyden's Method) is used to estimate the combined robot and image Jacobian. Using this algorithm, we can drive the objective function value to a neighborhood of zero. To show the effectiveness of the proposed algorithm, simulation results for a six degree of freedom robot are presented.

• Proceedings of the Korean Information Science Society Conference
• /
• 2004.04b
• /
• pp.721-723
• /
• 2004

A trust-region method is a quite attractive optimization technique. It is, in general, faster than the steepest descent method and is free of a learning rate unlike the gradient-based methods. In addition to its convergence property (between linear and quadratic convergence), ifs stability is always guaranteed, in contrast to the Newton's method. In this paper, we present an efficient implementation of the maximum likelihood independent component analysis (ICA) using the trust-region method, which leads to trust-region-based ICA (TR-ICA) algorithms. The useful behavior of our TR-ICA algorithms is confimed through numerical experimental results.

• The Journal of the Convergence on Culture Technology
• /
• v.9 no.3
• /
• pp.647-654
• /
• 2023

Deep Learning is a useful method for classifying and recognizing complex data such as images and text, and the accuracy of the deep learning method is the basis for making artificial intelligence-based services on the Internet useful. However, the vast amount of user da vita used for training in deep learning has led to privacy violation problems, and it is worried that companies that have collected personal and sensitive data of users, such as photographs and voices, own the data indefinitely. Users cannot delete their data and cannot limit the purpose of use. For example, data owners such as medical institutions that want to apply deep learning technology to patients' medical records cannot share patient data because of privacy and confidentiality issues, making it difficult to benefit from deep learning technology. In this paper, we have designed a privacy preservation technique-applied deep learning technique that allows multiple workers to use a neural network model jointly, without sharing input datasets, in multi-party system. We proposed a method that can selectively share small subsets using an optimization algorithm based on modified stochastic gradient descent, confirming that it could facilitate training with increased learning accuracy while protecting private information.

• Earthquakes and Structures
• /
• v.1 no.1
• /
• pp.43-67
• /
• 2010

A gradient-based evolutionary optimization methodology is presented for finding the optimal design of both the added dampers and their supporting members to minimize an objective function of a linear multi-storey structure subjected to the critical ground acceleration. The objective function is taken as the sum of the stochastic interstorey drifts. A frequency-dependent viscoelastic damper and the supporting member are treated as a vibration control device. Due to the added stiffness by the supplemental viscoelastic damper, the variable critical excitation needs to be updated simultaneously within the evolutionary phase of the optimal damper placement. Two different models of the entire damper unit are investigated. The first model is a detailed model referred to as "the 3N model" where the relative displacement in each component (i.e., the spring and the dashpot) of the damper unit is defined. The second model is a simpler model referred to as "the N model" where the entire damper unit is converted into an equivalent frequency-dependent Kelvin-Voigt model. Numerical analyses for 3 and 10-storey building models are conducted to investigate the characters of the optimal design using these models and to examine the validity of the proposed technique.

• Journal of Information Processing Systems
• /
• v.16 no.1
• /
• pp.145-154
• /
• 2020

Maximum likelihood (ML) is the best estimator asymptotically as the number of training samples approaches infinity. This paper deduces an adaptive algorithm for blind signal processing problem based on gradient optimization criterion. A parametric density model is introduced through a parameterized generalized distribution family in ML framework. After specifying a limited number of parameters, the density of specific original signal can be approximated automatically by the constructed density function. Consequently, signal separation can be conducted without any prior information about the probability density of the desired original signal. Simulations on classical biomedical signals confirm the performance of the deduced technique.

• Transactions of the Korean Society of Mechanical Engineers A
• /
• v.28 no.3
• /
• pp.267-273
• /
• 2004

The purpose of this paper is to develop a systematic design method for the suspension system hard points and compliance elements, which have great influence on the handling stability of a vehicle. In this paper, a method to optimize J-turn responses is presented based on the principles of design of experiments, multi-body dynamic analysis and optimum design technique. The design variables associated with the J-turn maneuver are selected through the experimental design sensitivity analysis using the perturbation method. An objective function is defined as an approximate function for the J-turn characteristics using the TSA(Taylor series approximation). The values of the design variables, which make the optimized J-turn characteristics, are obtained using the conjugate gradient method. The result of the J-turn simulation shows that the optimized vehicle has more improved handling stability than the optimized vehicle.

• Journal of Animal Reproduction and Biotechnology
• /
• v.35 no.2
• /
• pp.142-148
• /
• 2020

The research work was undertaken to determine an effective fertilization medium, sperm separation method and sperm capacitating agent for optimum in vitro fertilization (IVF) rates of indigenous zebu cow oocytes. In experiment 1, tissue culture medium (TCM 199), Tyrode's albumin lactate pyruvate (TALP) and Brackett and Oliphant (BO) medium were used as basic medium for IVF of oocytes of indigenous zebu cows. In experiment 2, three sperm separation methods namely centrifugation, swim up and percoll gradient methods were used for separation of motile and viable spermatozoa for IVF. In experiment 3, for capacitation of spermatozoa, IVF medium supplemented with the heparin, mixture of penicillamine, hypotaurine and epinephrine (PHE) or the combination of heparin with PHE were used for fertilization. In vitro culture (IVC) of presumptive zygotes was done in modified synthetic oviduct fluid (mSOF) medium using standard procedure 24 h after sperm-oocytes co-culture. The cleavage rate was determined to evaluate the efficacy of fertilization medium, sperm separation method and sperm capacitating agent 24 h after IVC. The cleavage rate was higher in oocytes fertilized in TALP (63.3%) than in TCM 199 (47.5%) (p < 0.05). The cleavage rate was higher in oocytes fertilized by spermatozoa separated by percoll gradient method (62.3%) than by centrifugation (51.6%) (p < 0.05). The cleavage rate of oocytes was higher when insemination was done with spermatozoa capacitated in TALP supplemented with heparin and PHE (61.3%) compared to control (40.9%) (p < 0.05). In conclusions, TALP based medium and percoll gradient sperm separation followed by capacitation with combination of heparin and PHE are suitable for IVF of indigenous zebu cow oocytes in Bangladesh.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.21 no.6
• /
• pp.103-108
• /
• 2020

Generally, the cargo tank of LPG carriers corresponds to an independent tank Type A defined by the International Maritime Organization (IMO). The outside of the tank is insulated by polyurethane foam, and the tank is made of expensive low temperature steel that can withstand temperatures as low as -50℃. The cargo tank is composed of outer shell plates, bulkheads, stiffeners, web frames, and stringers. Among them, the outer shell plates, bulkheads, and stiffeners can be designed without structural analysis by the Classification Rules and are constructed easily through optimal design. On the other hand, optimal design, including numerous structural analysis, is not performed because web frames and stringers should be designed and approved through structural analysis. Only adequate design, which determines the design dimensions through several dozen structural analysis, is performed. In this study, for finite element analysis, eight loading conditions were applied, and the deformation of the entire ship for each loading condition was considered. The minimum weight design was performed for the web frames of cargo tanks in the 82,000 ㎥ LNG carrier through the gradient-based optimization technique, and the weight was reduced by approximately 108 tons per ship.

• Proceedings of the KIEE Conference
• /
• 2011.07a
• /
• pp.991-992
• /
• 2011

This study was designed to satisfy induced voltage limits considering drive's specifications and optimize design reducing usage of permanent magnet, by increasing salient poles ratio, when designing 150kW IPMSM. In order to achieve these objectives, design plans were determined, based on Ld and Lq parameters of a basic design model, according to changes in salient poles ratio and flux linkage using IPMSM's voltage equation and torque equation and then, required torque and induced voltage were analyzed using Sensitivity Analysis. Based on analysis data, the optimum design was performed and basic model's characteristics were compared to final model's through Gradient-Based Optimization Technique.

• Korean Journal of Remote Sensing
• /
• v.33 no.4
• /
• pp.423-436
• /
• 2017

This paper presents a deep learning-based road segmentation framework from very high-resolution orthophotos. The proposed method uses Deep Convolutional Autoencoders for end-to-end mapping of orthophotos to road segmentations. In addition, a set of post-processing steps were applied to make the model outputs GIS-ready data that could be useful for various applications. The optimization of the model's parameters is explained which was conducted via grid search method. The model was trained and implemented in Keras, a high-level deep learning framework run on top of Tensorflow. The results show that the proposed model with the best-obtained hyperparameters could segment road objects from orthophotos at an average accuracy of 88.5%. The results of optimization revealed that the best optimization algorithm and activation function for the studied task are Stochastic Gradient Descent (SGD) and Exponential Linear Unit (ELU), respectively. In addition, the best numbers of convolutional filters were found to be 8 for the first and second layers and 128 for the third and fourth layers of the proposed network architecture. Moreover, the analysis on the time complexity of the model showed that the model could be trained in 4 hours and 50 minutes on 1024 high-resolution images of size $106{\times}106pixels$, and segment road objects from similar size and resolution images in around 14 minutes. The results show that the deep learning models such as Convolutional Autoencoders could be a best alternative to traditional machine learning models for road segmentation from aerial photographs.