• 전기학회논문지P
• /
• 제64권4호
• /
• pp.267-272
• /
• 2015

Active noise control (ANC) is a method of cancelling a noise signal in an acoustic cavity by generating an appropriate anti-noise signal via canceling loudspeakers. The continuous progress of ANC involves the development of improved adaptive signal processing algorithms, transducers, and DSP hardware. In this paper, the convergence behavior and the stability of the FxLMS algorithm in ANC systems with real-time implementation is proposed. Specially, The advanced DSP H/W with dual core(DSP+ARM) and API(application programming interface) S/W programming was developed to improve the real-time implementation performance under the FxLMS algorithms of input noise such as road noise environment. The experimental results are found to be in good agreement with the theoretical predictions.

• 한국항해항만학회:학술대회논문집
• /
• 한국항해항만학회 2005년도 추계학술대회 논문집
• /
• pp.17-22
• /
• 2005

This paper presents a new adaptive autopilot for ships based on the Adaptive Neural Networks. The proposed adaptive autopilot is designed with some modifications and improvements from the previous studies on Adaptive Neural Networks by Adaptive Interaction (ANNAI) theory to perform course-keeping, turning and track-keeping control. A strategy for automatic selection c! the neural network controller parameters is introduced to improve the adaptation ability and the robustness of new ANNAI autopilot. In Part II of the paper, to show the effectiveness and feasibility of the proposed ANNAI autopilot, computer simulations of course-keeping and track-keeping tasks with and without the effects of measurement noise and external disturbances are presented.

• 한국항해항만학회지
• /
• 제29권9호
• /
• pp.771-776
• /
• 2005

This paper presents a new adaptive autopilot for ships based on the Adaptive Neural Networks. The proposed adaptive autopilot is designed with some modifications and improvements from the previous studies on Adaptive Neural Networks by Adaptive Interaction (ANNAI) theory to perform course-keeping, turning and track-keeping control. A strategy for automatic selection of the neural network controller parameters is introduced to improve the adaptation ability and the robustness of new ANNAI autopilot. In Part II of the paper, to show the effectiveness and feasibility of the proposed ANNAI autopilot, computer simulations of course-keeping and track-keeping tasks with and without the effects of measurement noise and external disturbances will be presented.

• 제어로봇시스템학회논문지
• /
• 제15권1호
• /
• pp.31-36
• /
• 2009

This paper presents a novel scalar adaptive filter, which is reformulated by additional acceleration term. The filter continuously estimates three different kinds of covariance such as the measurement noise covariance, the velocity error covariance and the acceleration error covariance. For estimating three covariances, we use the innovation method for the measurement noise covariance and the least square method for other covariances. In order to verify the proposed filter performance compared with the conventional scalar adaptive filter, we make indoor experimental environment similar to outdoor test using the ultrasonic sensors instead of GPS. Experimental results show that the proposed filter has better position accuracy than the traditional scalar adaptive filter.

• International Journal of Control, Automation, and Systems
• /
• 제3권1호
• /
• pp.137-142
• /
• 2005

Biomedical signals such as ECG, EMG, and EEG are extremely important in the diagnosis of patients. It is difficult to filter noise from these signals, and errors resulting from filtering can distort a biomedical signal. Existing systems have shown poor performance when complicated noise appears. Adaptive filtering is selected to contend with these defects. Existing adaptive filters can adjust the filter coefficient with the given filter order, but they can produce an unsuitable order in different environments. In order to solve this problem, an optimal adaptive filter with a dynamic structure was designed. Positive experimental results were obtained.

• 소음진동
• /
• 제8권5호
• /
• pp.832-840
• /
• 1998

Experimental studies on the adaptive vibration control of composite beams have been performed using a piezoelectric actuator and the neuro-controller. The variations in natural frequencies of the specimen and the actuation characteristics of the piezoelectric actuator according to the delamination in the bonding layer have been studied. In addition, the simulation of adaptive vibration control has been performed for the composite specimens with delaminated piezoelectric actuator using neuro-controller. The hardware for the adaptive vibration control experiment was prepared. A DSP(digital signal processor) has been used as a digital controller. Using neuro-controller, the adaptive vibration control experiment has been performed. The vibration control results using the neuro-controller show that the present neuro-controller has good performance and robustness with the system parameter variations.

• 한국소음진동공학회:학술대회논문집
• /
• 한국소음진동공학회 2000년도 추계학술대회논문집
• /
• pp.431-436
• /
• 2000

Adaptive feed forward control algorithms based largely upon LMS approach have developed in recent two decades, and they have been widely applied to practical sound and vibration control problems in the case of the reference signal is available. Feedforward control can be applied only when reference signals can be measured or regenerated, while feedback controllers are used to reduce; sound and vibration when reference signals are not available. In recent years, hybrid control schemes in which adaptive feed forward controllers are combined with feedback ones have been studied based on simulations and experiments. The results have shown that the hybrid control may have better control performances in convergence speed and steady state error than the single control schemes. Hybrid control has the advantages of improving stability and performance as well as the disturbance rejection property. However, little effort has been made to the analysis or interpretation of hybrid control systems. In this study, we discussed the feedback controller effects on the stability of feed forward control algorithm in the presence of uncertain error path and a simple example showed that a stable feedback controller could make the feedforward controller unstable. A design criterion of feedback controllers is proposed in order to guarantee the stability of feedforward algorithms in the presence of error paths with uncertainties.

• 대한기계학회논문집A
• /
• 제21권11호
• /
• pp.1837-1843
• /
• 1997

This study experimentally demonstrates the use of active muffler attached to the automotive exhaust system to reduce exhaust noise. For improving the signal to noise ratio in the process of estimation of secondary path transfer functions, the on-line algorithm that conventional inverse modeling is combined with adaptive line enhancer is used as the control algorithm. Active muffler is designed that the primary noise and the control sound are propagated as a plane wave in the outlet. Therefore, the error microphone could be placed out of the tail pipe center of a high temperature and the radiation noise to the outside could be reduced in the whole area around the outlet. The control experiment for reducing exhaust noise with active muffler is implemented during run-up at no load. From the experimental results presented, compared with the conventional off-line method, the proposed on-line method is capable to acquire a reduction of exhaust noise above 5 dB in overall sound power level.

• 한국음향학회지
• /
• 제13권6호
• /
• pp.15-23
• /
• 1994

본 논문은 소음원이 여러개 존재할 때 다중입력 다중제어의 능동적응제어에 관한 연구이다. 다중채널 제어의 ANC(Active Noise Control)에 관한 현재까지의 연구는 소음원이 여러개 존재하는 경우, 적응필터의 입력신호사이에 상관이 없는 것으로 가정하고 연구되었다. 그러나 실제는 소음신호의 수신마이크로폰에서 제거하고자 하는 소음신호들 사이에 상관이 존재하며, 다수개의 소음원 신호를 독립적으로 검출하는 것이 곤란한 경우가 많고 여기에 관한 연구가 불충분하다. 본 연구에서는 적응필터의 입력신호간에 소음원의 상관을 제거하는 상관제거필터를 제안하였다. 제안한 상관제거 필터를 삽입하여 소음제거 시뮬레이션 결과, 혼합비율을 $b_{12}=0.95$일 때 상관을 제거한 경우가 약 2배 정도 빠른 수렴속도를 나타냈다.

• 한국정보통신학회:학술대회논문집
• /
• 한국정보통신학회 2015년도 추계학술대회
• /
• pp.903-905
• /
• 2015

디지털 시대를 맞이하여 디지털 장치들의 필요성이 증가되고 있다. 일반적으로 영상은 압축, 인식, 처리하는 과정에서 여러 가지 원인에 의해 잡음이 발생한다. Salt and pepper 잡음을 제거하기 위한 대표적인 필터는 SMF, CWMF, AWMF 등이 있으며, 잡음 밀도가 높은 영역에서는 잡음 제거 특성이 다소 미흡해진다. 본 논문은 salt and pepper 잡음을 효과적으로 제거하기 위해, 비잡음 화소는 그대로 보존하고 잡음 화소는 변형된 적응 메디안 필터 알고리즘을 제안하였다.