• Title/Summary/Keyword: 함수 궤적

Search Result 125, Processing Time 0.021 seconds

Acceleration of Learning speed Neural Networks by Reducing Weight Oscillations (가중치 진동의 감소를 이용한 신경회로망의 학습속도 향상)

  • 임빈철;박동조
    • Proceedings of the Korean Institute of Intelligent Systems Conference
    • /
    • 1998.10a
    • /
    • pp.251-254
    • /
    • 1998
  • 본 논문에서는 신경회로망의 수렴속도를 높이기 위한 알고리즘을 제안한다. 전형적인 역전파 학습방식은 느린 수렴속도가 단점으로 제기되는데 이는 비용함수의 계곡부근에서 가중치의 궤적이 심한 진동현상을 보이기 때문이다. 이 문제를 해결하기 위해서 본 논문에서는 경사법에서 사용되는 갱신방향을 계곡의 진행방향을 이용하여 변경한다. 모의실험을 통하여 제안된 방법으로 가중치의 궤적에 나타나는 진동을 줄이고 수렴속도를 향상시킬 수 있음을 보인다.

  • PDF

On Singularities of SISO Nonlinear Systems (단일 입출력 비선형시스템의 특이점의 성질에 관한 연구)

  • Jin, Ju-Wha;Cho, Seong-Il;Jung, Soo-Yul;Seo, Joong-Eon;Shin, Dong-Ho
    • Proceedings of the KIEE Conference
    • /
    • 2001.07d
    • /
    • pp.1938-1939
    • /
    • 2001
  • 시스템을 입력과 출력간의 함수관계로 볼 때 출력에 대한 상대차수가 정의되지 않는 점을 특이점이라고 정의할 수 있다. 본 논문에서는 적절한 기하학적 조건을 만족하는 해석적인 시스템(analytic system)에 대하여 그러한 특이점의 성질을 살펴본다. 특이점을 지나는 궤적은 특이점의 특이도와 그에 관련된 함수 값의 부호에 의해서 특이점의 주변에서는 특이 매니폴드를 기준으로 한 영역에서 나머지 다른 영역으로 통과하거나 혹은 그를 기준으로 어느 한 쪽의 영역에만 머물러 있게 됨을 보임으로써 특이점과 이를 지나는 궤적의 상관관계를 명확히 규명하였다.

  • PDF

Optimal Trajectory Design of Descent/Ascent phase for a Lunar Lander With Considerable Sub-Phases (Sub-Phase를 고려한 달착륙선의 Descent/Ascent phase 최적 궤적 생성)

  • Jo, Sung-Jin;Min, Chan-Oh;Lee, Dae-Woo;Cho, Kyeum-Rae
    • Journal of the Korean Society for Aeronautical & Space Sciences
    • /
    • v.38 no.12
    • /
    • pp.1184-1194
    • /
    • 2010
  • The descent and ascent phases for a lunar lander are composed of several phases. Accordingly, the constraints and control values adequate for each phase are required in order to generate optimal lander's trajectory. The optimal trajectories for descent and ascent phases are generated by the cost function to minimize fuel consumption & attitude variation rates. In this paper, the optimal control problem to make trajectory uses Gauss pseudo-spectral method which is one of the direct approach method. This problem generates lander's reference trajectory, states and controls.

Trajectory tracking control system of unmanned ground vehicle (무인자동차 궤적 추적 제어 시스템에 관한 연구)

  • Han, Ya-Jun;Kang, Chin-Chul;Kim, Gwan-Hyung;Tac, Han-Ho
    • Journal of the Korea Institute of Information and Communication Engineering
    • /
    • v.21 no.10
    • /
    • pp.1879-1885
    • /
    • 2017
  • This paper discusses the trajectory tracking system of unmanned ground vehicles based on predictive control. Because the unmanned ground vehicles can not satisfactorily complete the path tracking task, highly efficient and stable trajectory control system is necessary for unmanned ground vehicle to be realized intelligent and practical. According to the characteristics of unmanned vehicle, this paper built the kinematics tracking models firstly. Then studied algorithm solution with the tools of the optimal stability analysis method and proposed a tracking control method based on the model predictive control. The controller used a kinematics-based prediction model to calculate the predictive error. This controller helps the unmanned vehicle drive along the target trajectory quickly and accurately. The designed control strategy has the true robustness, simplicity as well as generality for kinematics model of the unmanned vehicle. Furthermore, the computer Simulink/Carsim results verified the validity of the proposed control method.

Continuous Speech Recognition based on Parmetric Trajectory Segmental HMM (모수적 궤적 기반의 분절 HMM을 이용한 연속 음성 인식)

  • 윤영선;오영환
    • The Journal of the Acoustical Society of Korea
    • /
    • v.19 no.3
    • /
    • pp.35-44
    • /
    • 2000
  • In this paper, we propose a new trajectory model for characterizing segmental features and their interaction based upon a general framework of hidden Markov models. Each segment, a sequence of vectors, is represented by a trajectory of observed sequences. This trajectory is obtained by applying a new design matrix which includes transitional information on contiguous frames, and is characterized as a polynomial regression function. To apply the trajectory to the segmental HMM, the frame features are replaced with the trajectory of a given segment. We also propose the likelihood of a given segment and the estimation of trajectory parameters. The obervation probability of a given segment is represented as the relation between the segment likelihood and the estimation error of the trajectories. The estimation error of a trajectory is considered as the weight of the likelihood of a given segment in a state. This weight represents the probability of how well the corresponding trajectory characterize the segment. The proposed model can be regarded as a generalization of a conventional HMM and a parametric trajectory model. The experimental results are reported on the TIMIT corpus and performance is show to improve significantly over that of the conventional HMM.

  • PDF

Implementation of moving sound effect using modified HRTF (개선된 머리전달함수를 이용한 이동음 효과 구현)

  • Seo, Bo-Kug;Han, Eui-Hwan;Cha, Hyeong-Tae;Moon, Nam-Mee
    • Proceedings of the Korean Society of Broadcast Engineers Conference
    • /
    • 2012.11a
    • /
    • pp.254-255
    • /
    • 2012
  • 임의의 위치에 음상을 정위하는 기본적인 방법으로, 머리전달함수(HRTF : Head Related Transfer Function)을 원음에 convolution하는 기법이 사용된다. 또한 머리전달함수의 주파수 영역에서의 보간을 이용하여 이동음 효과를 생성할 수 있다. 하지만 더미헤드를 이용하여, 각각의 고도각과 방위각에서 측정된 머리전달함수는 사용자에 따라 정위감을 저하시킬 수 있으며, 클릭음 발생, 보간 특성의 변화등의 문제점이 발생하게 된다. 또한 좌표로 표현되는 영상과는 달리, 소리는 들려오는 방향을 정확한 좌표로 표현하기가 힘들다. 이에 본 논문에서는, 이동음 궤적상의 머리전달함수와 그 주변의 머리전달함수를 선형 보간법으로 그룹화하여 생성된 새로운 머리전달함수를 이용하여 이동음을 구현하는 방법을 제안한다. 또한, 생성된 머리전달함수를 이용하여 구현한 이동음의 청감평가를 수행하여 개선 효과를 확인하였다.

  • PDF

The Proximity Scheme of the Perceptual Space for Indexing The Trajectories of Tags (태그 궤적 색인을 위한 인식공간 근접성 기법)

  • Kim, Dong-Hyun;Ahn, Swng-Woo
    • Journal of the Korea Institute of Information and Communication Engineering
    • /
    • v.13 no.10
    • /
    • pp.2140-2146
    • /
    • 2009
  • Since tags do not have location informations, the identifiers of tags which are symbolic data are used as the location informations. Therefore, it is difficult to define the proxmity between two trajectories of tags and inefficient to process the user queries for tags. In this paper, we define the perceptual space to model the location of a tag and propose the proximity of the perceptual spaces. The proximity of the perceptual spaces is composed of the static proximity and dynamic proximity. Using the proximity of the perceptual spaces, it is possible to measure the proximity between two trajectories of tags and build the efficient indexes for tag trajectories. We evaluated the performance of the proposed proximity function for tag trajectories on the IR-tree and the $R^*$-tree.

Energy Minimization Model for Pattern Classification of the Movement Tracks (행동궤적의 패턴 분류를 위한 에너지 최소화 모델)

  • Kang, Jin-Sook;Kim, Jin-Sook;Cha, Eul-Young
    • The KIPS Transactions:PartB
    • /
    • v.11B no.3
    • /
    • pp.281-288
    • /
    • 2004
  • In order to extract and analyze complex features of the behavior of animals in response to external stimuli such as toxic chemicals, we implemented an adaptive computational method to characterize changes in the behavior of chironomids in response to treatment with the insecticide, diazinon. In this paper, we propose an energy minimization model to extract the features of response behavior of chironomids under toxic treatment, which is applied on the image of velocity vectors. It is based on the improved active contour model and the variations of the energy functional, which are produced by the evolving active contour. The movement tracks of individual chironomid larvae were continuously measured in 0.25 second intervals during the survey period of 4 days before and after the treatment. Velocity on each sample track at 0.25 second intervals was collected in 15-20 minute periods and was subsequently checked to effectively reveal behavioral states of the specimens tested. Active contour was formed around each collection of velocities to gradually evolve to find the optimal boundaries of velocity collections through processes of energy minimization. The active contour which is improved by T. Chan and L. Vese is used in this paper. The energy minimization model effectively revealed characteristic patterns of behavior for the treatment versus no treatment, and identified changes in behavioral states .is the time progressed.

Income Trajectories of Working Poor and Working Non-poor: A Latent Growth Model (근로빈곤층과 근로비빈곤층의 차별적 소득 궤적 - 잠재성장모형의 응용 -)

  • Lee, Sohyeon;Lim, Up
    • Journal of the Korean Regional Science Association
    • /
    • v.37 no.1
    • /
    • pp.15-27
    • /
    • 2021
  • This study investigates the difference in income trajectories of the working poor and the non-working poor and explains the effects of socio-demographic (marriage, education) and regional (living in large cities) factors on intergroup differences. We use Seoul Survey data collected between 2009-2018 and the latent growth modeling approach. It was found that the trajectory difference between groups was statistically significant. Since 2016, the income gap widened as the income of the working poor stagnated. The three variables included in this model better explained the income trajectory of the working poor compared to the working non-poor. In particular, the change in income growth rate was positively related to whether they live in large cities. This suggests the possibility that living in a large city would act as an economic premium for the working poor. It is necessary to conduct follow-up studies on urban premiums for the working poor.

A study on the C-shape Sharp Turn of fish robot according to biological mimic (생물학적 모방에 따른 물고기 로봇의 빠른 방향 전환 연구)

  • Park, Jin-Hyun;Lee, Tae-Hwan;Choi, Young-Kiu
    • Journal of the Korea Institute of Information and Communication Engineering
    • /
    • v.15 no.12
    • /
    • pp.2626-2631
    • /
    • 2011
  • CST(C-shape sharp turn) represented the motion whereby fish bend their tail quickly in a C-shape to achieve an emergent changing of its swimming direction on fish swimming. But there is not yet the general motion trajectory functions related to CST. In this paper, we proposed the very simple motion functions related to CST sequence recorded from a real fish by biologists. Through the computer simulations, we confirmed the usefulness of the proposed function.