• 전자공학회논문지B
• /
• 제31B권9호
• /
• pp.133-141
• /
• 1994

Conventional competitive learning algorithms compute the Euclidien distance to determine the winner neuron out of all predetermined output neurons. In such cases, there is a drawback that the performence of the learning algorithm depends on the initial reference(=weight) vectors. In this paper, we propose a new competitive learning algorithm that dynamically generates output neurons. The proposed method generates output neurons by dynamically changing the class thresholds for all output neurons. We compute the similarity between the input vector and the reference vector of each output neuron generated. If the two are similar, the reference vector is adjusted to make it still more like the input vector. Otherwise, the input vector is designated as the reference vector of a new outputneuron. Since the reference vectors of output neurons are dynamically assigned according to input pattern distribution, the proposed method gets around the phenomenon that learning is early determined due to redundant output neurons. Experiments using speech data have shown the proposed method to be superior to existint methods.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제16권9호
• /
• pp.2976-2990
• /
• 2022

The adaptive adjustment of reference or weight vectors in decomposition-based methods has been a hot research topic in the evolutionary community over the past few years. Although various methods have been proposed regarding this issue, most of them aim to diversify solutions in the objective space to cover the true Pareto fronts as much as possible. Different from them, this paper proposes a knee point-based reference vector adaptive adjustment strategy to concurrently balance the convergence and diversity. To be specific, the knee point-based reference vector adaptive adjustment strategy firstly utilizes knee points to construct the adaptive reference vectors. After that, a new fitness function is defined mathematically. Then, this paper further designs a many-objective evolutionary algorithm with knee point-based reference vector adaptive adjustment strategy, where the mating operation and environmental selection are designed accordingly. The proposed method is extensively tested on the WFG test suite with 8, 10 and 12 objectives and MPDMP with state-of-the-art optimizers. Extensive experimental results demonstrate the superiority of the proposed method over state-of-the-art optimizers and the practicability of the proposed method in tackling practical many-objective optimization problems.

• Journal of Power Electronics
• /
• 제18권4호
• /
• pp.1027-1036
• /
• 2018

In this paper a new generalized space vector pulse width modulation scheme is proposed based on the principle of reverse mapping to drive the switches of multilevel inverters. This projected scheme is developed based on the middle vector of the subhexagon which holds the tip of the reference vector, which plays a major role in mapping the reference vector. A new approach is offered to produce middle vector of the subhexagon which holds tip of the reference vector in the multilevel space vector plane. By using middle vector of the subhexagon, reference vector is linked towards the inner two level sub-hexagon. Then switching vectors, switching sequence and dwell times corresponding to a particular sector of a two-level inverter are determined. After that, by using the two level stage findings, the switching vectors related to exact position of the reference vector are directly generated based on principle of the reverse mapping approach and do not need to be found at n level stage. In the reverse mapping principle, the middle vector of subhexagon is added to the formerly found two level switching vectors. The proposed generalized algorithm is efficient and it can be applied to an inverter of any level. In this paper, the proposed scheme is explained for a five-level inverter and the performance is analyzed for five level and three level inverters through MATLAB. The simulation results are validated by implementing the propose scheme on a V/f controlled three-level inverter fed induction motor using dSPACE control desk.

• 전력전자학회논문지
• /
• 제3권2호
• /
• pp.118-124
• /
• 1998

This paper describes a relationship of a space vector PWM and a sinusoidal PWM and presents that the space vector PWM can produce linearly the output voltage to the unity MI(modulation index). At first, reference angles and holding angles are derived from expanding a Fourier series of the reference voltage waveform and then the angles are used for the inverter switching to linearize transfer characteristics of the inverter. In addition, the harmonic components of the output voltage are analyzed and on-line control is shown to be feasible by approximating in piecewise-linearization the reference and holding angles versus the MI. In V/f control of the induction motor, it is verified by the experiment that the motor current is changed smoothly for the variation of the inverter input voltage and the change of the reference voltage.

• 한국통신학회논문지
• /
• 제36권5A호
• /
• pp.457-463
• /
• 2011

본 논문에서는 다중 셀 환경에서의 부분공간 간섭 정렬에서 레퍼런스 벡터간의 직교성을 통해 여러 개의 직교 레퍼런스 벡터 조합을 제공함으로써 전체 생의 합 용량이 최대화되도록 레퍼런스 벡터를 선정하는 방법을 제안하였다. 제안된 방법은 전체 셀의 채널용량을 최대화시키는 하나의 벡터를 선정하고 나머지 셀에서는 선택된 레퍼런스 벡터와 직교하는 벡터를 생성하여 채널 용량이 가장 큰 벡터를 각 셀의 레퍼런스 벡터로 선정한다. 제안된 방법은 주어진 채널에 적합한 레퍼런스 벡터를 선정 시 선택의 자유도를 높임으로써 기존보다 전체 셀 전송 용량을 증가시킨다. 전산 모의 실험을 통해 기존의 방법에 비해 높은 셀의 합용량 증대 성능을 보였다.

• 한국콘텐츠학회논문지
• /
• 제10권4호
• /
• pp.28-35
• /
• 2010

다중 참조 영상을 이용한 가변 블록 크기의 움직임 예측 및 보상 기법이 부호화 효율을 높이기 위해 H.264/AVC에 채택되었다. 하지만 움직임 예측 및 보상으로 인한 계산량은 다중 참조 영상과 가변 블록의 수에 비례하여 증가한다. 본 논문에서는 화질은 유지하며 계산량을 줄이기 위한 새로운 고속 참조 영상선택 방법을 제안한다. 먼저 다중 참조 영상들에 대해서 $4{\times}4$ 블록의 SAD값을 이용하여 움직임 벡터 참조 지도를 만든다. 다음으로 가변 블록 크기의 움직임 예측 및 보상이 움직임 벡터 참조 지도를 활용하여 실시된다. 제안하는 방법은 H.264/AVC 표준과 비교하여 BDPSNR은 평균적으로 0.01dB 나빠지고 BDBR은 약 0.27% 증가하지만 영상 부호화 속도를 약 38% 단축시킨다.

• 대한전기학회논문지:전기기기및에너지변환시스템부문B
• /
• 제54권4호
• /
• pp.201-207
• /
• 2005

This paper has investigated the sensorless vector control method of induction motor based on the stator flux oriented voltage equation and the digital low pass filter (LPF) with compensator of phase/gain. The Proposed vector control method is easy to decide the stator reference voltages and control of motor, since it is based on stator flux vector But this method has sensitive structure to excessive sensor noise and PWM pulsating components of stator currents because the measured stator currents are directly used to compensate the internal resistive voltage drop at the determination of stator reference voltages. To eliminate the noise sensitive of proposed vector drive, this paper propose the digital LPF with compensator of phase/gain base on orthogonal property of stator current vector in stationary $\alpha$, $\beta$ reference frame. The proposed methods have been simulated and implemented on a sensorless vector drive for 750W three-phase induction motor. The simulation and experimental results demonstrate effectiveness of the proposed methods.

• 대한전자공학회논문지TC
• /
• 제47권7호
• /
• pp.23-28
• /
• 2010

본 논문은 다중 셀 상향링크 시스템의 셀 간 간섭을 완화시키기 위한 부분공간 간섭 정렬을 적용했을 때 높은 채널 이득과 높은 합용량을 얻기 위해 간섭 신호공간 형성에 사용되는 레퍼런스 벡터 설정 방법을 제안하였다. 제안된 방법은 레퍼런스 벡터의 설정에 따라 채널과의 매칭이 달라지고 이로부터 셀의 합용량이 변화하는 레퍼런스 벡터의 다이버서티를 얻는 것이다. 각 셀의 처리량을 최대화시키기 위해 전송 전에 여러 개의 레퍼런스 벡터를 전송하고 이중에서 최적의 레퍼런스 벡터를 선정함으로써 기존의 간섭정렬방식보다 높은 채널 이득과 높은 합용량을 얻을 수 있다. 전산 모의실험을 통해 제안된 방법을 적용한 경우 임의로 설정된 레퍼런스 벡터를 설정했을 때의 셀 합용량에 비해 약 60% 향상된 값을 얻었다.

• 한국인터넷방송통신학회논문지
• /
• 제17권6호
• /
• pp.217-222
• /
• 2017

지구 중력장 내에 위치한 물체는 연직 아래 방향의 힘을 받고 있다. 중력장 내에서 물체의 운동을 기술하기 위하여 운동상태 방정식을 이용한다. 자유 낙하하는 물체를 해석할 때 기준 방향은 연직 하방을 +y으로, 연직위로 던져 올린 물체를 해석할 때 기준방향은 연직상방을 +y으로, 연직 아래로 던져 내린 물체를 해석할 때, 기준 방향은 연직 하방을 +y으로 선택하여 해석함이 일반적이다. 이 논문에서는 두 벡터의 스칼라 곱 (즉, 도트 곱)을 이용하여 연직 상방 또는 하방 두 경우를 방향으로 선택하여 구성한 벡터 운동 상태 방정식(vector kinematics equations)을 해석의 결과가 서로 일치함을 제시한다. 두 벡터의 스칼라 곱 (즉, 도트 곱)을 이용하여 물체의 상태 방정식를 해석한 예는 선행 연구에서 거의 찾아볼 수가 없다. 이 결과를 이용하면, 수평면의 방향 또는 빗각을 이루는 방향의 초속도로 던져 올리거나 던져 내린 물체의 운동 상태를 해석하기 위하여 연직 기준 방향을 상방 또는 하방으로 임의 선택할 수가 있다.

• 한국전기전자재료학회:학술대회논문집
• /
• 한국전기전자재료학회 2001년도 추계학술대회 논문집
• /
• pp.65-68
• /
• 2001

In the vector control methods of induction motor, the stator current is divided into the flux and torque component current. By controlling these components respectively, the methods control independently flux and torque as in the DC motor and improve the control effects. To apply the vector control methods, the position of the rotor current is identified. The indirect vector control use the parameters of the machine to identify the position of rotor flux. But due to the temperature rise during machine operation, the variation of rotor resistance degrades the vector control. To solve the problem, the q-axis is aligned to reference frame without phase difference by comparing the real flux component with the reference flux component. Then to compensate the slip, PI controller is used. The proposed method keeps a constant slip by compensating the gain of direct slip frequency when the rotor resistance of induction motor varies. To prove the validations of the proposed algorithm in the paper, computer simulations is executed.