• 한국항행학회논문지
• /
• 제15권4호
• /
• pp.556-561
• /
• 2011

일반적으로 비선형 시스템에서 particle filter가 Kalman Filter보다 표적추적 성능이 뛰어나다고 알려져 있다. 그러나 particle filter는 많은 연산량을 요구하는 단점이 있다. 본 논문에서는 particle filter 보다 적은 particle의 수, 즉 적은 연산량으로 동일한 성능을 가지는 Rao-Blackwellized particle filter의 모델의 민감성을 줄인 Rao-Blackwellized Multiple Model Particle Filter(RBMMPF)의 알고리즘을 소개하고 이에 다중센서 정보를 융합하는 자료융합 기법을 적용하였다. 시뮬레이션을 통해 단일센서 정보를 이용한 RBMMPF 표적추적 성능과 다중센서정보를 융합한 RBMMPF의 표적추적 성능을 비교, 분석하였다.

• 로봇학회논문지
• /
• 제4권2호
• /
• pp.97-104
• /
• 2009

Recently, simultaneous localization and mapping (SLAM) approaches employing Rao-Blackwellized particle filter (RBPF) have shown good results. However, due to the usage of the accurate sensors, distinct particles which compensate one another are attenuated as the RBPF-SLAM continues. To avoid this particle depletion, we propose the strategic games to assign the particle's payoff which replaces the importance weight in the current RBPF-SLAM framework. From simulation works, we show that RBPF-SLAM with the strategic games is inconsistent in the pessimistic way, which is different from the existing optimistic RBPF-SLAM. In addition, first, the estimation errors with applying the strategic games are much less than those of the standard RBPF-SLAM, and second, the particle depletion is alleviated.

• 로봇학회논문지
• /
• 제3권4호
• /
• pp.308-314
• /
• 2008

Recently, simultaneous localization and mapping (SLAM) approaches employing Rao-Blackwellized particle filter (RBPF) have shown good results. However, no research is conducted to analyze the result representation of SLAM using RBPF (RBPF-SLAM) when particle diversity is preserved. After finishing the particle filtering, the results such as a map and a path are stored in the separate particles. Thus, we propose several result representations and provide the analysis of the representations. For the analysis, estimation errors and their variances, and consistency of RBPF-SLAM are dealt in this study. According to the simulation results, combining data of each particle provides the better result with high probability than using just data of a particle such as the highest weighted particle representation.

• 한국정보과학회:학술대회논문집
• /
• 한국정보과학회 2006년도 한국컴퓨터종합학술대회 논문집 Vol.33 No.1 (B)
• /
• pp.151-153
• /
• 2006

we present a method of sequential speech enhancement, where we infer clean speech signal using a Rao-Blackwellized particle filter (RBPF), given a noise-contaminated observed signal. In contrast to Kalman filtering-based methods, we consider a non-Gaussian speech generative model that is based on the generalized auto-regressive (GAR) model. Model parameters are learned by a sequential Newton-Raphson expectation maximization (SNEM), incorporating the RBPF. Empirical comparison to Kalman filter, confirms the high performance of the proposed method.

• 제어로봇시스템학회논문지
• /
• 제19권8호
• /
• pp.682-687
• /
• 2013

Navigation is a crucial capability for all types of manned or unmanned vehicles. However, vehicle navigation in underwater environments still remains a challenging problem since GPS signals for position fixes are not available in the water. Terrain-referenced underwater navigation is an alternative navigation technique that utilizes geometric information of the subsea terrain to correct drift errors due to dead-reckoning or inertial navigation. Terrain-referenced navigation requires the description of an undulating terrain surface as a mathematical function or table, which often leads to a highly nonlinear estimation problem. Recently, PFs (Particle Filters), which do not require any restrictive assumptions about the system dynamics and uncertainty distributions, have been widely used for nonlinear filtering applications. However, PF has considerable computational requirements which used to limit its applicability to problems of relatively low state dimensions. This study proposes the use of a Rao-Blackwellized particle filter that is computationally more efficient than the standard PF for terrain-referenced underwater navigation involving a moderate number of states, and its performance is compared with that of the extended Kalman filter algorithm. The validity and feasibility of the proposed algorithm is demonstrated through numerical simulations.

• 로봇학회논문지
• /
• 제4권4호
• /
• pp.305-311
• /
• 2009

FastSLAM is a factored solution to SLAM problem using a Rao-Blackwellized particle filter. In this paper, we propose a practical FastSLAM implementation method using an infrared camera for indoor environments. The infrared camera is equipped on a Pioneer3 robot and looks upward direction to the ceiling which has infrared tags with the same height. The infrared tags are detected with theinfrared camera as measurements, and the Nearest Neighbor method is used to solve the unknown data association problem. The global map is successfully built and the robot pose is predicted in real time by the FastSLAM2.0 algorithm. The experiment result shows the accuracy and robustness of the proposed method in practical indoor environment.

• 제어로봇시스템학회논문지
• /
• 제16권1호
• /
• pp.40-47
• /
• 2010

This paper presents a simultaneous localization and mapping (SLAM) of an indoor hallway environment using Rao-Blackwellized particle filter (RBPF) along with a line segment as a landmark. Based on the fact that fluent line features can be extracted around the ceiling and side walls of hallway using vision sensor, a horizontal line segment is extracted from an edge image using Hough transform and is also tracked continuously by an optical flow method. A successive observation of a line segment gives initial state of the line in 3D space. For data association, registered feature and observed feature are matched in image space through a degree of overlap, an orientation of line, and a distance between two lines. Experiments show that a compact environmental map can be constructed with small number of horizontal line features in real-time.

• 제어로봇시스템학회논문지
• /
• 제15권6호
• /
• pp.619-625
• /
• 2009

FastSLAM which uses the Rao-Blackwellized particle filter is one of the famous solutions to SLAM (Simultaneous Localization and Mapping) problem that estimates concurrently a robot's pose and surrounding environment. However, the particle depletion problem arises from the loss of the particle diversity in the resampling process of FastSLAM. Then, the performance of FastSLAM degenerates over the time. In this work, DIR (Density Information-based Resampling) technique is proposed to solve the particle depletion problem. First, the cluster is constructed based on the density of each particle, and the density of each cluster is computed. After that, the number of particles to be reserved in each cluster is determined using a linear method based on the distance between the highest density cluster and each cluster. Finally, the resampling process is performed by rejecting the particles which are not selected to be reserved in each cluster. The performance of the DIR proposed to solve the particle depletion problem in FastSLAM was verified in computer simulations, which significantly reduced both the RMS position error and the feature error.

• 제어로봇시스템학회논문지
• /
• 제19권10호
• /
• pp.907-914
• /
• 2013

A FastSLAM is an algorithm for SLAM (Simultaneous Localization and Mapping) using a Rao-Blackwellized particle filter and its performance is known to degenerate over time due to the loss of particle diversity, mainly caused by the particle depletion problem in the resampling phase. In this paper, the GeSPIR (Geographically Stratified Particle Information-based Resampling) technique is proposed to solve the particle depletion problem. The proposed algorithm consists of the following four steps : the first step involves the grouping of particles divided into K regions, the second obtaining the normal weight of each region, the third specifying the protected areas, and the fourth resampling using regional equalization weight. Simulations show that the proposed algorithm obtains lower RMS errors in both robot and feature positions than the conventional FastSLAM algorithm.