• Journal of Broadcast Engineering
• /
• v.25 no.3
• /
• pp.418-427
• /
• 2020

The measurement of meat freshness at meat markets is important for the health of consumers. Currently a variety of sensors have been studied for the measurement of the meat freshness. Therefore, the analysis of sensor data is needed for the reduction of measurement errors. In this paper, we analyze the freshness measurement data of ten sensors based on deep learning. The measured data are composed of beef, pork and chicken, whose reliability and noise-robustness are examined by a deep neural network. Further, to search for multiple sensors better than a torrymeter, PCA (principle component analysis) is carried. Then, we validated that the performance of the three sensors outperforms the torrymeter in the experiment.

• Journal of Institute of Control, Robotics and Systems
• /
• v.6 no.9
• /
• pp.758-767
• /
• 2000

The properties of linear time-varying(LTV) systems vary because of the time-varying property of plant parameters. The generalized controller design method for linear time-varying systems does not exit because the analytic soultion of dynamic equation has not been found yet. Hence, to design a controller for LTV systems, the robust control methods for uncertain LTI systems which are the approximation of LTV systems have been generally ised omstead. However, these methods are not sufficient to reflect the fast dynamics of the original time-varying systems such as missiles and supersonic aircraft. In general, both the performance and the robustness of the control system which is designed with these are not satisfactory. In addition, since a better model will give the more robustness to the controlled system, a gain scheduling technique based on LTI controller design methods has been uesd to solve time problem. Therefore, we propose a new gain scheduled QFT method for LTV systems based on neural networks in this paper. The gain scheduled QFT involves gain dcheduling procedured which are the first trial for QFT and are well suited consideration of the properties of the existing QFT method. The proposed method is illustrated by a numerical example.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.25 no.3
• /
• pp.389-395
• /
• 2021

In this paper, we propose a target image exchange model to improve performance of the object tracking algorithm based on a Siamese network. The object tracking algorithm based on the Siamese network tracks the object by finding the most similar part in the search image using only the target image specified in the first frame of the sequence. Since only the object of the first frame and the search image compare similarity, if tracking fails once, errors accumulate and drift in a part other than the tracked object occurs. Therefore, by designing a CNN(Convolutional Neural Network) based model, we check whether the tracking is progressing well, and the target image exchange timing is defined by using the score output from the Siamese network-based object tracking algorithm. The proposed model is evaluated the performance using the VOT-2018 dataset, and finally achieved an accuracy of 0.611 and a robustness of 22.816.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.24 no.11
• /
• pp.1477-1483
• /
• 2020

The stability guarantee of wireless network based control systems is still challenging due to the lossy links and node failures. This paper proposes a hierarchical cluster-based network protocol called robust wireless sensor and actuator network (R-WSAN) by combining time, channel, and space resource diversity. R-WSAN includes a scheduling algorithm to support the network resource allocation and a control task sharing scheme to maintain the control stability of multiple plants. R-WSAN was implemented on a real test-bed using Zolertia RE-Mote embedded hardware platform running the Contiki-NG operating system. Our experimental results demonstrate that R-WSAN provides highly reliable and robust performance against lossy links and node failures. Furthermore, the proposed scheduling algorithm and the task sharing scheme meet the stability requirement of control systems, even if the controller fails to support the control task.

• Journal of Broadcast Engineering
• /
• v.27 no.3
• /
• pp.283-294
• /
• 2022

The feature map used in the network for deep learning generally has larger data than the image and a higher compression rate than the image compression rate is required to transmit the feature map. This paper proposes a method for transmitting a pyramid feature map with high compression rate, which is used in a network with an FPN structure that has robustness to object size in deep learning-based image processing. In order to efficiently compress the pyramid feature map, this paper proposes a structure that predicts a pyramid feature map of a level that is not transmitted with pyramid feature map of some levels that transmitted through the proposed prediction network to efficiently compress the pyramid feature map and restores compression damage through the proposed reconstruction network. Suggested mAP, the performance of object detection for the COCO data set 2017 Train images of the proposed method, showed a performance improvement of 31.25% in BD-rate compared to the result of compressing the feature map through VTM12.0 in the rate-precision graph, and compared to the method of performing compression through PCA and DeepCABAC, the BD-rate improved by 57.79%.

• Fibers and Polymers
• /
• v.7 no.4
• /
• pp.344-351
• /
• 2006

This paper is intended to determine the optimal processing parameters applied to the dyeing procedure so that the desired color strength of a raw fabric can be achieved. Moreover, the processing parameters are also used for constructing a system to predict the fabric quality. The fabric selected is the nylon and Lycra blend. The dyestuff used for dyeing is acid dyestuff and the dyeing method is one-bath-two-section. The Taguchi quality method is applied for parameter design. The analysis of variance (ANOVA) is applied to arrange the optimal condition, significant factors and the percentage contributions. In the experiment, according to the target value, a confirmation experiment is conducted to evaluate the reliability. Furthermore, the genetic algorithm (GA) is combined with the back propagation neural network (BPNN) in order to establish the forecasting system for searching the best connecting weights of BPNN. It can be shown that this combination not only enhances the efficiency of the learning algorithm, but also decreases the dependency of the initial condition during the network training. Most of all, the robustness of the learning algorithm will be increased and the quality characteristic of fabric will be precisely predicted.

• International Journal of Control, Automation, and Systems
• /
• v.2 no.1
• /
• pp.83-91
• /
• 2004

The purpose of this paper is to obtain an accurate nonlinear system model to test various control schemes for a motion control system that requires high speed, robustness and accuracy. An industrial sewing machine equipped with a Brushless DC motor is considered. It is modeled by a neural network that is configured as an output-error dynamical system. The identified model is essentially a one step ahead prediction structure in which past inputs and outputs are used to calculate the current output. Using the model, a 2 degree-of-freedom PID controller to compensate the effects of disturbance without degrading tracking performance has been de-signed. In this experiment, it is not preferable for safety reasons to tune the controller online on the actual machinery. Experimental results confirm that the model is a good approximation of sewing machine dynamics and that the proposed control methodology is effective.

• The Journal of Korea Robotics Society
• /
• v.12 no.3
• /
• pp.313-321
• /
• 2017

One of the most frequently performed tasks in human-robot interaction (HRI), intelligent vehicles, and security systems is face related applications such as face recognition, facial expression recognition, driver state monitoring, and gaze estimation. In these applications, accurate head pose estimation is an important issue. However, conventional methods have been lacking in accuracy, robustness or processing speed in practical use. In this paper, we propose a novel method for estimating head pose with a monocular camera. The proposed algorithm is based on a deep neural network for multi-task learning using a small grayscale image. This network jointly detects multi-view faces and estimates head pose in hard environmental conditions such as illumination change and large pose change. The proposed framework quantitatively and qualitatively outperforms the state-of-the-art method with an average head pose mean error of less than $4.5^{\circ}$ in real-time.

• Journal of Power Electronics
• /
• v.18 no.5
• /
• pp.1380-1397
• /
• 2018

The synchronous reluctance motor (SynRM) servo-drive system has highly nonlinear uncertainties owing to a convex construction effect. It is difficult for the linear control method to achieve good performance for the SynRM drive system. The nonlinear backstepping control system using upper bound with switching function is proposed to inhibit uncertainty action for controlling the SynRM drive system. However, this method uses a large upper bound with a switching function, which results in a large chattering. In order to reduce this chattering, a nonlinear backstepping control system using an adaptive law is proposed to estimate the lumped uncertainty. Since this method uses an adaptive law, it cannot achiever satisfactory performance. Therefore, a nonlinear backstepping control system using a reformed recurrent Hermite polynomial neural network with an adaptive law and an error estimated law is proposed to estimate the lumped uncertainty and to compensate the estimated error in order to enhance the robustness of the SynRM drive system. Further, the reformed recurrent Hermite polynomial neural network with two learning rates is derived according to an increment type Lyapunov function to speed-up the parameter convergence. Finally, some experimental results and a comparative analysis are presented to verify that the proposed control system has better control performance for controlling SynRM drive systems.

• Advances in Computational Design
• /
• v.1 no.1
• /
• pp.1-25
• /
• 2016

Domes are architectural and elegant structures which cover a vast area with no interrupting columns in the middle, and with suitable shapes can be also economical. Domes are built in a wide variety of forms and specialized terms are available to describe them. According to their form, domes are given special names such as network, lamella, Schwedler, ribbed, and geodesic domes. In this paper, an optimum topology design algorithm is performed using the enhanced colliding bodies optimization (ECBO) method. The network, lamella, ribbed and Schwedler domes are studied to determine the optimum number of rings, the optimum height of crown and tubular sections of these domes. The minimum volume of each dome is taken as the objective function. A simple procedure is defined to determine the dome structures configurations. This procedure includes calculating the joint coordinates and element constructions. The design constraints are implemented according to the provision of LRFD-AISC (Load and Resistance Factor Design-American Institute of Steel Constitution). The wind loading act on domes according to ASCE 7-05 (American Society of Civil Engineers). This paper will explore the efficiency of various type of domes and compare them at the first stage to investigate the performance of these domes under different kind of loading. At the second stage the wind load on optimum design of domes are investigated for Schwedler dome. Optimization process is performed via ECBO algorithm to demonstrate the effectiveness and robustness of the ECBO in creating optimal design for domes.