• Korean Journal of Fisheries and Aquatic Sciences
• /
• v.8 no.1
• /
• pp.1-6
• /
• 1975

The sea noise of moving trawl net was recorded by an underwater tape recorder which was set wireless, and was analyzed by a sound level meter and an octave-band analyzer. The frequency distribution of sea noise of the moving otter trawl net ranged from DC to 5000 Hz, and the dominant frequency zone ranged from 500 Hz to 700 Hz, and the maximum sound pressure is about 22 dB at the otter trawl net. The main sound source of the sea noise from the moving trawl net was found to be sea noise due to the resistance of the ground rope against the sea bottom.

• ETRI Journal
• /
• v.41 no.6
• /
• pp.771-781
• /
• 2019

We propose the VS³-NET model to solve the task of question answering questions with machine-reading comprehension that searches for an appropriate answer in a given context. VS³-NET is a model that trains latent variables for each question using variational inferences based on a model of a simple recurrent unit-based sentences and self-matching networks. The types of questions vary, and the answers depend on the type of question. To perform efficient inference and learning, we introduce neural question-type models to approximate the prior and posterior distributions of the latent variables, and we use these approximated distributions to optimize a reparameterized variational lower bound. The context given in machine-reading comprehension usually comprises several sentences, leading to performance degradation caused by context length. Therefore, we model a hierarchical structure using sentence encoding, in which as the context becomes longer, the performance degrades. Experimental results show that the proposed VS³-NET model has an exact-match score of 76.8% and an F1 score of 84.5% on the SQuAD test set.

• IEMEK Journal of Embedded Systems and Applications
• /
• v.15 no.4
• /
• pp.197-204
• /
• 2020

The ImageNet dataset is a large scale dataset and contains various natural scene images. In this paper, we propose a convolutional neural network (CNN)-based weighted ensemble technique for the ImageNet classification task. First, in order to fuse several models, our technique uses weights for each model, unlike the existing average-based ensemble technique. Then we propose an algorithm that automatically finds the coefficients used in later ensemble process. Our algorithm sequentially selects the model with the best performance of the validation set, and then obtains a weight that improves performance when combined with existing selected models. We applied the proposed algorithm to a total of 13 heterogeneous models, and as a result, 5 models were selected. These selected models were combined with weights, and we achieved 3.297% Top-5 error rate on the ImageNet test dataset.

• Journal of the Korean Society of Fisheries and Ocean Technology
• /
• v.16 no.1
• /
• pp.1-15
• /
• 1980

An underwater recording system was designed to measure the sound spectra of the underwater noises produced by fishing gears in operation. Recorded were noi~es from three types of fishing gears: an anchovy set net, three anchovy boat seine net and a stern trawlnet. Acoustic analysis were made using a heterodyne analyzer, a digital frequency analyzer and a level recorder. The no;'e produced by the anchovy set net was found in the high frequency region of the onset of ambient noise spectrum with a slope of - 6 dB/octave. Here the ambient noise spectrum is higher, though similar in shape, than Knudsen spectrum, and is attributed to the breaking action of the coastal wave. Measured noise spectra during the fishing operations of the anchovy boat seine nets are attributed to the background noise of the sea in the presence of the fishing vessels. The frequency distribution of the noise was 20~5, 000 Hz in the case of two steel anchovy boat seiners, and 20-3,000 Hz in the case of the wooden anchovy boat seiner. The predominant frequency range was 250~350 Hz and maximum sound pressure level was 122 dB (re $1\muPa$) in the case of the steel boat and ] 17 dB in the case of the wooden boat. The noises produced by the trawl fishing gears are remarkably higher than the background noi~e in the presence of the fishing vessel. The frequency distribution of the noi~e was 20-6,300 Hz. The predominant frequency range was 100~200 Hz and maximum sound pressure level was 137 dB ( re $1\muPa$) . The noise spectra were not so much different from that caused by vibrations of the towing cable and the structure of the ground rope of the trawl net towed in an experimental tank.

• Journal of the Korean Society of Fisheries and Ocean Technology
• /
• v.27 no.4
• /
• pp.247-254
• /
• 1991

Mackerel purse seine fishery is one of the most important fisheries in Korea and its annual mean catch from 1982 to 1988 is 350 thousand tons. Between 1982 and 1988, the characteristics of catch at 5 sectors of the sea area was analyzed according to operating sea block and the catch by fishes based on the data obtained from daily report of fishing condition that had been made out by 48 Korean fishing boats engaged in the fishery of purse seine. In these 5 sectors of the sea area, catch per net hauling in the coastal area of Cheju Island is less than that in the other sectors. Accordingly, the shape of fishing gear during operation has been measured to analyzed the deformation of net shape caused by the bottom current. (The lunar calender is used in the paper) The results obtained are as follows; The average CPUE of purse seines is 31.6 tons and it is higher from October to April than that from May to September. In the 5 purse seine fishing grounds, the ratio of net haul in the coastal sea area of cheju Island is 42.4% and it is the highest level, but the CPUE is 25.7 tons which is relatively low level. Each CPUE in the sea areas is significant differences with 1% level. Index of seasonal variation of mackerel CPUE from May to October is lower than the standard value, whereas index from November to January is higher. The diameter of circle by net shooting becomes shorter to the minimum caused by the bottom current during loose set, becomes narrower and the efficiency of fish encirclement becomes lower. The shape of leadline by bottom current becomes long oval type along with current direction in the tight set, while it becomes long oval type in the direction of a right angle to the current direction and the enclosed area is reduced hurriedly, and the deformation of net is most serious in the loose set.

• Journal of the Korean Society of Marine Environment & Safety
• /
• v.25 no.5
• /
• pp.535-541
• /
• 2019

The purpose of this study was to investigate the possibility of fishery in offshore wind farms and evaluate the risk linked to the presence of turbines and submarine cables in these areas. With this objective, we studied an offshore wind farm in the Southwest Sea and the current state of vessels in the surrounding National Fishing Port. The risk assessment criteria for 22 fishing gears and methods were set by referring to the fishing boats; thereafter, the risk was assessed by experts. The fishing gears and methods that could be safely operated (i.e., associated with low risk) in the offshore wind farm were: single-line fishing, jigging, and the anchovy lift net. The risk was normal so that it is possible to operate, but the fishing gears and methods that need attention are: the set long line, drifting long line, troll line, squid rip hook, octopus pot, webfoot octopus pot, coastal fish pot, stow net on stake, winged stow net, stationary gill net, and drift gill net. Moreover, the fishing gears and methods difficult to operate in the of shore wind farm (i.e., associated with high risk) were: the dredge, beam trawl, and purse seine. Finally, those associated with very high risk and that should not be allowed in offshore wind farms were: the stow net, anchovy drag net, otter trawl, Danish seine, and bottom pair trawl.

• Korean Journal of Computational Design and Engineering
• /
• v.2 no.4
• /
• pp.237-243
• /
• 1997

A new method is introduced for the interpolation in NURBS Surface. This method uses the basis functions to assign the parameter values to the arbitrary set of geometric data and uses the iteration method to compute the control net. The advantages of this method are the feasible transformation of the data set to the matrix form and the effective surface generation as a result, especially to the design engineer.

• The KIPS Transactions:PartC
• /
• v.8C no.5
• /
• pp.621-628
• /
• 2001

SET is one of the useful protocol for credit payment in the Electronic market. Since, the delivery problem is conformed to delivery cooperation not payment problem, the classic SET protocol didn\`t consider about the certification of delivery. But the environment of electronic market be changed to manage the sold, delivery and payment etc.. In this paper, based on this consider, we propose a new SET protocol which has an function to verify the delivery based on SKP and verify it after analyzed by Petri nets. Specially, we consider SKP between Customer, Merchant, and Acquirer for improve the verify function.

• Journal of the Korean Society for Precision Engineering
• /
• v.16 no.7
• /
• pp.132-143
• /
• 1999

In this study, an intermediate model is presented as a new method for development of a parametric simulation program. This model enables us to analyze effectively the static and dynamic structure of a real production system. The static structure of the real system can be modelled in an entity-relationship diagram and the dynamic structure of the real system in a Petri net. The intermediate model consists of an entity-relationship diagram and a Petri net. Using this intermediate model man can not only reduce the time and cost for simulation program development, but also increase the modelling reliability of the developed simulation program. To show the usefulness of this intermediate model. the intermediate models for two production subsystems, manufacturing sub-system and transport subsystem, are set up.

• Management Science and Financial Engineering
• /
• v.3 no.1
• /
• pp.15-38
• /
• 1997

In a flexible manufacturing system, a cell controller is able to identify and evaluate a number of alternative decisions to meet the objectives set by the factory level controller. In this paper, a Petri net-based cell controller is presented to accomplish this task. A static model is developed by using the Integrated Computer Aided Definition(IDEF0) method to represent clear functional relationships among the objects of the system. Based on the static model, two Petri net models are developed for the physical part flow and for the information flow. Multiple decision alternatives are generated from the physical part flow model and are synchronized with the information flow model for execution of the selected alternative.