• Proceedings of the Korean Society of Computer Information Conference
• /
• 2022.07a
• /
• pp.49-52
• /
• 2022

건물 외벽에 발생하는 균열은 시설물 구조 안전에 영향을 미치며 그 크기에 따라 위험도가 달라진다. 이에 따라 전문검사관의 현장 점검을 통해 발생 균열 두께를 정밀하게 측정할 필요가 있고 최근에는 이러한 현장 안전점검에 인공지능을 도입하려는 추세다. 그러나 기존의 균열 데이터셋은 주로 콘크리트에만 한정되어 다양한 외벽에 강인한 모델을 구축하기 어렵고 균열 두께를 측정하기 위해 정확한 마스크(Mask) 정보가 필요하나 이를 만족하는 데이터셋이 부재하다. 본 논문에서는 다양한 외벽에 강인한 균열 구획화 모델을 목적으로 2,744장의 이미지를 촬영하고 매직 완드 기법으로 라벨링을 진행해 데이터셋을 구축 후, 이를 바탕으로 딥러닝 기반 균열 구획화 모델을 개발했다. UNet-ResNet50을 최종모델로 선정 및 개발 결과, 테스트 데이터셋에 대해 81.22%의 class IoU 성능을 보였다. 본 연구의 기술을 바탕으로 균열 두께를 측정하여 건축물 안전점검에 활용될 수 있기를 기대한다.

• Proceedings of the KIPE Conference
• /
• 1998.07a
• /
• pp.309-313
• /
• 1998

고성능 서보 전동기 시스템을 구축하기 위해서는 저속영역과 고속영역을 포함하는 넓은 속도영역에서의 정확한 속도검출을 통한 정밀한 속도제어기 필수적이며, 관성모멘트와 같은 전동기의 파라메터 변동에 대해 강인한 속도제어와 외란 억제능력도 중요한 요소로서 고려되어야 한다. 변동하는 부하의 관성모멘트을 식별하여 PI 속도제어기를 실시간으로 적응 동정하고, 플랜트 잡음과 측정잡음을 고려하는 상태 관측기인 칼만필터의 부하관성에 대한 민감성을 제거하기 위해 이를 적응 동정하여 적응 상태 추정기를 구현함으로써 우수한 속도 추정 성능을 얻었다. 또한 외란과 불확실한 모델링은 등가 외란으로 추정되어 전향적으로 보상된다. 본 논문에서는 이러한 특징을 이용하여 전동기의 고성능 속도제어를 구현하고 유도전동기를 이용한 실험을 통하여 연구결과의 유효성을 확인한다.

• Journal of the Korean Society for Precision Engineering
• /
• v.11 no.2
• /
• pp.110-119
• /
• 1994

The Robust Controller for scan-loop is designed using LQG/LTR Methodology. The design and analysis of spiral, rosette and conical scan patterns are discussed. The perfermance and robustness of the LQG/LTR controller are analyzed through experiments and cpmpared with those of the P-controller. Especially to improve the scan performance at large look angle, the cage coil output is linearized using a binomial equation. It is demonstrated that the scan-loop system by the LQG/LTR control is very robust to phase uncertainties.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.10 no.10
• /
• pp.1835-1841
• /
• 2006

Demand of enhanced bilge separation sensor system has been recently increased due to the severe regulation reinforcement of MEPC(Marine Environment Production Committee). Up to date bilge separation sensor has to be extremely accurate and highly reliable. To design and build such a bilge separator. a precise oily water separation level sensor that distinguishes oil from water is critical. Three dimensional simulations have been carried out to figure out the characteristics of capacitive level sensors, which grounds the finding of the parameters required to design the sensors. The parasitic capacitance problem which is inherent to capacitive level sensors has been taken care of. This paper concludes with the future research direction that can be pursued with the newly defined parameters of the capacitive level sensors.

• Transactions of the Korean Society for Noise and Vibration Engineering
• /
• v.17 no.1 s.118
• /
• pp.72-79
• /
• 2007

As environmental vibration requirements on precision equipment become more stringent, use of pneumatic isolators has become more popular and their performance is subsequently required to be further improved. Dynamic performance of passive pneumatic isolators is related to various design parameters in a complicated manner and, hence, is very limited especially in low frequency range by volume of chambers. In this study, an active control technique, so called time delay control which is considered to be adequate for a low frequency or nonlinear system, is applied to a single chamber pneumatic isolator. The procedure of applying the tine delay control law to the pneumatic isolator is presented and its effectiveness in enhancement of transmissibility performance is shown based on simulation and experiment. Comparison between passive and active pneumatic isolators is also presented.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.21 no.6
• /
• pp.755-760
• /
• 2011

This paper presents a method for improving the positioning accuracy of the laser navigation. As a wireless navigation system, the laser navigation which is more flexible than a wired guidance system is used for the localization and control of an AGV(automatic guided vehicle). However, the laser navigation causes the large positioning error while the AGV turns or moves fast. To solve the problem, we propose the method for improving the positioning accuracy of the laser navigation using particle filter which has robust and reliable performance in non-linear/non-gaussian systems. For the experiment, we use the actual fork-type AGV. The AGV has a gyro, two encoders and a laser navigation. To verify the performance, the proposed method is compared with the laser navigation which is a product. In the experimental result, we verified that the proposed method could improve the positioning accuracy by approximately 66.5%.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.7 no.6
• /
• pp.1413-1422
• /
• 2012

Adaptive back stepping control technique may provide robust control characteristics under parameter perturbation caused by changing external condition. In order to synthesize a high-precision velocity controller for IPMSM(Interior Permanent Magnet Synchronous Motor) using this method, the period of control loop should be very small. However, because of the resolution of the encoder for speed measurement, control cycle is limited, which makes it difficult to improve the performance of the controller. This paper proposes a velocity controller design method based on nonlinear adaptive back-stepping method to accomplish fast and accurate performance. Here, an EKF(Extended Kalman Filter) method is incorporated for the estimation of the motor speed into the design of a speed controller using adapted back-stepping control technique. The performance of the proposed controller is demonstrated through simulation using PSIM.

• Journal of Aerospace System Engineering
• /
• v.16 no.4
• /
• pp.77-87
• /
• 2022

The Precision Aerial Delivery System is an instrument designed to improve the poor landing accuracy of aerial delivery system with conventional circular parachutes, and is equipped with an Airborne Guidance Unit to safely transport supplies to the desired destination. Currently, the landing accuracy of the PADS product is reported as CEP₅₀ 100m and also differs significantly, depending on the actual topography and weather environment. In this study, HILS was constructed based on the 6DOF nonlinear modeling of PADS to analyze the maneuver characteristics of Ram Air Parachute under wind environments. By using the new algorithm a precision soft landing algorithm including Energy Management and Final Approach is designed. HILS results show that it is possible to achieve a precise soft landing within CEP₅₀ 40m, and it can be exploited to develop an actual PADS drop test.

• Journal of the Korean Society for Aeronautical & Space Sciences
• /
• v.47 no.2
• /
• pp.107-113
• /
• 2019

In this paper, a robust filter based wind velocity estimation algorithm without an air velocity sensor in an air vehicle is presented. The wind velocity is useful information for the air vehicle to perform precise guidance and control. In general, the wind velocity can be obtained by subtracting an air velocity which is obtained by an air velocity sensor such as a pitot-tube, and a ground velocity which is obtained by a navigation equipment. However, in order to simplify the configuration of the air vehicle, the wind estimation algorithm is necessary because the wind velocity can not be directly obtained if the air velocity measurement sensor is not used. At this time, the aerodynamic coefficient of the air vehicle changes due to the turbulence, which causes the uncertainty of the system model of the filter, and the wind estimation performance deteriorates. Therefore, in this study, we propose a wind estimation method using $H{\infty}$ filter to ensure robustness against aerodynamic coefficient uncertainty, and we confirmed through simulation that the proposed method improves the performance in the uncertainty of aerodynamic coefficient.

• Journal of the Institute of Convergence Signal Processing
• /
• v.14 no.2
• /
• pp.117-123
• /
• 2013

Since the nonlinear factors such as friction in a mechanical servo system can't be easily measured nor estimated accurately. Therefore, it is difficult to compensate friction correctly. Friction makes a significant error in a 2-axis stabilized gimbal system and finally fails to reach the ultimate control performance goals. To solve these problems, lots of studies on the control methods applying observer have been performed. However, these methods can be used in specific conditions and are limited to apply them to the accurate 2-axis stabilized gimbal system in military sector. This paper deals with the PI-LEAD algorithm which is modified with a general and robust PID algorithm, proves the effect of the algorithm through modeling and simulation, and verifies the performance by applying the algorithm to the real 2-axis stabilized system. It is verified through the performance test that the PI-LEAD algorithm minimizes the error caused by friction and meets requirements of the accurate servo system.