• 한국농업기계학회:학술대회논문집
• /
• 한국농업기계학회 2017년도 춘계공동학술대회
• /
• pp.135-135
• /
• 2017

Greenhouse have been developed to provide the plants with good environmental conditions for cultivation crop, two major factors of which are the inside air temperature and humidity. The inside temperature are influenced by the heating systems, ventilators and for systems among others, which in turn are geverned by some type of controller. Likewise, humidity environment is the result of complex mass exchanges between the inside air and the several elements of the greenhouse and the outside boundaries. Most of the existing models are based on the energy balance method and heat balance equation for modelling the heat and mass fluxes and generating dynamic elements. However, greenhouse are classified as complex system, and need to make a sophisticated modeling. Furthermore, there is a difficulty in using classical control methods for complex process system due to the process are non linear and multi-output(MIMO) systems. In order to predict the time evolution of conditions in certain greenhouse as a function, we present here to use of recurrent neural networks(RNN) which has been used to implement the direct dynamics of the inside temperature and inside humidity of greenhouse. For the training, we used algorithm of a backpropagation Through Time (BPTT). Because the environmental parameters are shared by all time steps in the network, the gradient at each output depends not only on the calculations of the current time step, but also the previous time steps. The training data was emulated to 13 input variables during March 1 to 7, and the model was tested with database file of March 8. The RMSE of results of the temperature modeling was $0.976^{\circ}C$, and the RMSE of humidity simulation was 4.11%, which will be given to prove the performance of RNN in prediction of the greenhouse environment.

• 한국컴퓨터정보학회논문지
• /
• 제4권1호
• /
• pp.62-67
• /
• 1999

신경회로망은 입력층. 출력층, 하나 이상의 은닉층으로 구성된 네드워크이다. 학습능력과 근사화 능력으로 말미암아 신경회로망은 패턴인식 및 시스템제어분야에서 많이 사용되고 있다. Elman 신경회로망은 J. Elman에 의해 제안되었으며, recurrent network의 형태로 구성되어 있다. Elman 신경회로망은 기존의 신경회로망에 context층을 새로 추가하여, 은닉층의 출력을 context층의 입력으로 피드백 하는 구조로 되어 있다. 본 논문에서는 Elman 신경회로망을 변형한 형태로, 은닉층 뿐 만 아니라 출력층의 출력도 context층으로 피드백 하는 새로운 형태의 Elman 신경회로망을 제안한다. 제안한 방식의 유용성을 확인하기 위해 X-Y cartesian에 적용하여 시뮬레이션한 결과는 기존의 신경회로망 및 Elman 신경회로망 보다 우수한 방식임을 보여 주고 있다.

• 한국통신학회논문지
• /
• 제31권2A호
• /
• pp.201-206
• /
• 2006

다중 빔 위성중계기 부품중의 하나인 MSM(Microwave Switch Matrix)에 필요한 SPST 스위치 MMIC를 설계 및 제작하였다. 설계된 스위치 MMIC는 3GHz 대역에서 동작하며, 새로운 구조를 채택하여 기존의 FET 스위치보다 전력 특성과 격리도를 개선하였으며, 스위치의 On/Off 상태에서의 입출력 반사손실 특성이 우수하다. MMIC는 0.15um GaAs pHEMT 공정으로 제작되었으며, 3$\∼$4GHz 대역에서 2.0dB 이하의 삽입손실과, 63dB 이상의 격리도 성능을 가지는 것으로 측정되었다. 또한 사용된 단위 pHEMT 소자가 0.2mm Gate Width 임에도 불구하고 320dBm 이상의 OIP3 특성을 가지고 있는 것으로 측정되었으며, 이 결과는 기존의 발표된 FET 스위치에 비해 높은 전력 특성이다.

• 한국통신학회논문지
• /
• 제29권8B호
• /
• pp.721-729
• /
• 2004

본 논문에서는 광 버스트 스위칭 네트워크의 코어 노드에 그룹 스케줄링 알고리즘을 적용하여, 전산 모의실험을 통해 그 성능을 측정하였다. 성능 평가를 위해, 다채널 입출력 포트를 갖는 코어 노드에 대하여, 즉시 스케줄링 방식과 비교하였다. 그룹 스케줄링은 노드에 먼저 도착하는 버스트 헤더 패킷의 정보를 이용하여 일정한 시간 창에 스케줄링 될 버스트들을 스케줄링하기 때문에, 전산 모의실험 결과 그룹 스케줄링 방식이 즉시 스케줄링 방식보다 버스트 손실 확률과 채널 이용률이 모두 개선되었으며 부하의 증가에 따라 차이가 더욱 커졌다. 또한, 출력포트에 파장 변환기를 사용한 경우에 대해서도 성능을 측정하였다. 이 경우에는, 그룹 스케줄링 방식과 즉시 스케줄링 방식의 버스트 손실 확률과 채널 이용률 모두 부하 범위 0.1-0.9에서 서로 비슷하게 나타났으나. 파장 변환기의 사용 빈도는 즉시 스케줄링이 그룹 스케줄링보다 약 7배 이상으로 높아, 그룹 스케줄링 방식을 사용하면 보다 경제적인 노드 구조를 구현할 수 있음을 알 수 있었다.

• 동아시아식생활학회지
• /
• 제20권5호
• /
• pp.823-835
• /
• 2010

The research aimed to measure the efficiency of using multi stores in a foodservice company using by DEA (data envelopment analysis) which is a new management science technique. The study also attempted to identify relevant variables affecting DEA efficiency in order to suggest methods for improving efficiency. The data were collected from 148 contract foodservice operations, which were operated in similar fashion in October 2009. The DEA efficiency was calculated as an output-oriented BCC Model. Sales, and CSI (customer satisfaction index) were used as output variables whereas food cost, labor cost, and management expense were used as input variables to calculate the DEA efficiency. Operation process variables of the unit consisted of the were consist of ratio of regular employee, ratio of housekeeper, meal counts, meal price, food cost per meal, contract period, number of menu items, forecasting accuracy, order accuracy, inventory turnover, use of processed food, deviation of food cost, number of new menus, and number of events. According to the BCC score and profitability, units were classified into four groups: High efficiency-high profitability (HEHP), High efficiency-low profitability (HELP), Low efficiency-high profitability (LEHP), and Low efficiency-low profitability (LELP). The HEHP group contained 54 units, which mostly contracted management fee type and had a high meal price. The units were also very large and, served three meals. Twenty of the units were operated with high labor cost: most of these were factories and hospitals. The LEHP group contained 20 units, that were mainly office stores of large scale and medium price. Fifty-four LELP group had a low meal price. A high performance group must have high efficiency, profitability, and satisfaction. The BCC score was over 0.969, the meal price was over 4,116 won, the food cost was over 2,077 won, and meal counts per month were over 10,212 meals.

• 한국전자통신학회논문지
• /
• 제15권4호
• /
• pp.759-764
• /
• 2020

4차 산업혁명에 맞춰 제조·생산의 패러다임 또한 변경되고 있는 상황에서 인더스트리 4.0에 대한 사회적 요구 및 방향은 이미 되돌릴 수 없으며, 그로 인하여 3D프린팅 기술의 확장성과 범용성이 주목받고 있다. 3D 프린터는 제품 개발 비용 감소를 위한 목적으로 개발된 기술로, 최근 3D 프린터 기술특허가 만료되면서 관련 기술이 공개되었고 이를 적용한 다양한 기술이 연구 개발되고 있으며 이로 인해 다양한 특이점이 발견 및 보완되고 있다. 이에 본 연구에서는 3D 프린터 사용에 있어 오프라인에서 기계를 직접 조정하고, 모델링 데이터의 직접 입력을 통해 3D프린팅 시작해야 하는 현재 출력방식에 대한 불편함을 보완하기 위해 다중 3D 프린터에 대한 실시간 온라인 출력요청 및 정상적인 출력물을 기대할 수 있는 모니터링 기능 그리고 다중 3D 프린터에 대한 온라인 실시간 원격관리 기능을 수행하기 위해 IoT 시스템에 대한 설계를 제안한다.

• 한국항해항만학회:학술대회논문집
• /
• 한국항해항만학회 2006년도 International Symposium on GPS/GNSS Vol.1
• /
• pp.277-282
• /
• 2006

It is well known that Kalman filtering is an optimal real-time data fusion method for GPS/INS integration. However, it has some limitations in terms of stability, adaptability and observability. A Kalman filter can perform optimally only when its dynamic model is correctly defined and the noise statistics for the measurement and process are completely known. It is found that estimated Kalman filter states could be influenced by several factors, including vehicle dynamic variations, filter tuning results, and environment changes, etc., which are difficult to model. Neural networks can map input-output relationships without apriori knowledge about them; hence a proper designed neural network is capable of learning and extracting these complex relationships with enough training. This paper presents a GPS/INS integrated system that combines Kalman filtering and neural network algorithms to improve navigation solutions during GPS outages. An Extended Kalman filter estimates INS measurement errors, plus position, velocity and attitude errors etc. Kalman filter states, and gives precise navigation solutions while GPS signals are available. At the same time, a multi-layer neural network is trained to map the vehicle dynamics with corresponding Kalman filter states, at the same rate of measurement update. After the output of the neural network meets a similarity threshold, it can be used to correct INS measurements when no GPS measurements are available. Selecting suitable inputs and outputs of the neural network is critical for this hybrid method. Detailed analysis unveils that some Kalman filter states are highly correlated with vehicle dynamic variations. The filter states that heavily impact system navigation solutions are selected as the neural network outputs. The principle of this hybrid method and the neural network design are presented. Field test data are processed to evaluate the performance of the proposed method.

• Structural Engineering and Mechanics
• /
• 제86권1호
• /
• pp.119-137
• /
• 2023

The purpose of this research is to assess the performance of CBN and ceramic tools during the dry turning of gray cast iron EN GJL-350. During the turning operation, the variable machining parameters are cutting speed, feed rate, depth of cut and type of the cutting material. This contribution consists of two sections, the first one deals with the performance evaluation of four materials in terms of evolution of flank wear, surface roughness (2D and 3D) and cutting forces. The focus of the second section is on statistical analysis, followed by modeling and optimization. The experiments are conducted according to the Taguchi design L₃₂ and based on ANOVA approach to quantify the impact of input factors on the output parameters, namely, the surface roughness (Ra), the cutting force (Fz), the cutting power (Pc), specific cutting energy (Ecs). The RSM method was used to create prediction models of several technical factors (Ra, Fz, Pc, Ecs and MRR). Subsequently, the desirability function approach was used to achieve a multi-objective optimization that encompasses the output parameters simultaneously. The aim is to obtain optimal cutting regimes, following several cases of optimization often encountered in industry. The results found show that the CBN tool is the most efficient cutting material compared to the three ceramics. The optimal combination for the first case where the importance is the same for the different outputs is Vc=660 m/min, f=0.116 mm/rev, ap=0.232 mm and the material CBN. The optimization results have been verified by carrying out confirmation tests.

• Nuclear Engineering and Technology
• /
• 제56권5호
• /
• pp.1584-1602
• /
• 2024

It is vital to output the required electrical power following various task requirements when the space reactor power supply is operating in orbit. The dynamic performance of the closed Brayton cycle thermoelectric conversion system is initially studied and analyzed. Based on this, a load tracking power regulation method is developed for the liquid metal cooled space reactor power system, which takes into account the inlet temperature of the lithium on the hot side of the intermediate heat exchanger, the filling quantity of helium and xenon, and the input amount of the heat pipe radiator module. After comparing several methods, a power regulation method with fast response speed and strong system stability is obtained. Under various changes in power output, the dynamic response characteristics of the ultra-small liquid metal lithium-cooled space reactor concept scheme are analyzed. The transient operation process of 70 % load power shows that core power variation is within 30 % and core coolant temperature can operate at the set safety temperature. The second loop's helium-xenon working fluid has a 65K temperature change range and a 25 % filling quantity. The lithium at the radiator loop outlet changes by less than ±7 K, and the system's main key parameters change as expected, indicating safety. The core system uses less power during 30 % load power transient operation. According to the response characteristics of various system parameters, under low power operation conditions, the lithium working fluid temperature of the radiator circuit and the high-temperature heat pipe operation temperature are limiting conditions for low-power operation, and multiple system parameters must be coordinated to ensure that the radiator system does not condense the lithium working fluid and the heat pipe.

• 한국인터넷방송통신학회논문지
• /
• 제16권6호
• /
• pp.329-336
• /
• 2016

옥외광고 및 산업 현장에서는 실시간으로 다양한 의사를 표현하기 위하여 영상처리 기반 이미지 변환 출력 LED 전광판 시스템을 구현하려고 한다. 최근 각종 현장에서는 단순한 문장 표현이 아닌 이미지를 이용한 직관적인 의사소통의 중요성이 커지고 있다. 따라서 의사소통을 위해 단순히 입력된 정보를 출력하는 것이 아닌 실시간 정보를 출력할 수 있는 시스템이 요구되고 있다. 이를 위해 본 시스템은 다양한 임의의 이미지 출력이 불가능한 기존의 LED 전광판에 이미지를 맵핑하는 문제를 해결하는 것과 이미지 출력이 가능한 형태로 변환하여 저전력의 LED를 활용하여 한정된 자원 안에서 효율적으로 메시지와 이미지를 출력하도록 개발하였다. 따라서 본 논문에서는 LED 전광판을 무선네트워크로 관리하고 마이크로 컨트롤러 중 하나인 ATmega2560과 Wi-Fi 모듈 및 서버와 안드로이드 어플리케이션 클라이언트를 통해 텍스트만 출력하는 기존의 전광판과는 달리 이미지 출력과 다양한 출력을 가능하게 하여 출력할 문자와 이미지에 대한 처리를 서버에서 관리함으로서 이미지 출력에 필요한 여러 변환 과정에 생기는 부하를 줄이는데 초점을 맞춘 시스템을 제안한다.