Hyun-Taek Choi

Real-time acoustic source localization for autonomous navigation of underwater vehicles

This paper presents a method of acoustic source localization for the autonomous navigation of underwater vehicles. The proposed method is based on the time delay of acoustic signals received by two hydrophones. A probabilistic estimation is used to determine an accurate time delay. The time delay is estimated by using a Bayesian update process, and the direction of the acoustic source is obtained from the estimated time delay. Acoustic source localization is accomplished by accumulating direction information from various vehicle locations. Using extended Kalman filter, the location of the acoustic source is estimated recursively. The proposed method can provide a reliable estimation of the direction and location of the acoustic source, even under for a noisy acoustic signal. Experimental results demonstrate the performance of the proposed acoustic source localization method in a real sea environment.

Navigation and control for a test bed AUV-SNUUV I

This paper describes the design of underwater navigation and control system useful for towing tank experiments of a test bed autonomous underwater vehicle (AUV) named by SNUUV I (Seoul National University Underwater Vehicle I). A 6 DOF dynamic model for the SNUUV I is derived with hydrodynamic coefficients, which are estimated with the help of the Extended Kalman Filter (EKF) and a potential code. Based on a mathematical model, a nonlinear sliding mode controller is designed for diving and steering maneuvers of SNUUV I. A navigation algorithm is developed using three ranging sonars keeping the towing tank condition in mind. It is demonstrated numerically that the navigation system together with the controller guides the vehicle to follow the desired depth and path with sufficient accuracy.

Experimental results on EKF-based underwater localization algorithm using artificial landmark and imaging sonar

This paper proposes an underwater localization algorithm using artificial landmarks with imaging sonar. It is organized as follows; 1) recognizing artificial objects 2) A look at EKF (Extended Kalman Filter) SLAM for lacalization. Acoustic images from imaging sonar are very unstable, for this reason we developed artificial landmarks that can easily be detected in noisy environments. We also designed a probability-based, recognition framework. In this way, the distance and bearing of the recognized artificial landmarks are acquired, allowing us to perform the localization for our underwater vehicle. And then, EKF-based localization algorithm is carried out the localization which produce a path of underwater robot and a location of landmarks. The proposed localization algorithm is verified by experiments in a basin.

Underwater precise navigation using multiple sensor fusion

This paper introduces the implementation of a precise underwater navigation solution using multi-sensor fusion technique, which is based on USBL, DGPS, DVL and AHRS measurements. To realize the accurate, precise and frequent update rate underwater navigation solution, three strategies are chosen. The first one is the heading angle calibration to enhance the performance of standalone dead-reckoning algorithm. The second one is introduction of effective outlier rejection algorithm. The third one is that absolute position is fused timely to prevent error accumulation of dead-reckoning, where the absolute position can be acquired from USBL or DGPS measurements considering predefined finite state machine. The performance of the developed algorithm is verified with experimental data of UUV(Unmanned underwater vehicle) at Sea with various operation scenario.

Bundle adjustment from sonar images and SLAM application for seafloor mapping

This paper reports on two-view bundle adjustment using sonar images, specifically focusing on feature detection and a sensor measurement model for imaging sonar. To overcome limited sensor information for underwater navigation, we use Dual frequency IDentification SONar (DIDSON) in the imaging mode to provide spatial constraints when a scene is revisited. Unlike terrestrial images, sonar images are usually low resolution with highly speckled noise. We found that exploiting features from nonlinear scale space improves feature detection. In this paper, we adopt KAZE features and use random sample consensus (RANSAC) to refine correspondences. Using these correspondences, we propose point-based relative pose estimation via bundle adjustment. The target application that this work focuses on is underwater seafloor mapping, and the proposed model assumes a fixed elevation. Through this work, we present (i) validation of nonlinear scale space features for sonar images and (ii) proposal of a sonar sensor measurement model for underwater simultaneous localization and mapping (SLAM). The proposed method will be validated through both synthetic data sets and a tank test for seafloor mapping.

Highly accurate motion control system for omni-directional underwater robot

Since an underwater robot can be used as various purposes, fast and accurate motion becomes a key issue because basically overall efficiency of given tasks and quality of jobs highly depend on the low-level control performance. For this, an underwater robot is typically designed to be able to generate omni-directional motion. An omni-directional thruster configuration brings lots of advantages for fast and accurate motion. But, controlling of an omni-directional underwater robot is not easy because 3 or 4 horizontal thrusters have slightly different characteristics and each thruster has nonlinearity with dead-zone. Additionally, navigation sensors such as IMU(inertia measurement unit), DVL(doppler velocity log) provide accurate information only when their geometrical relationships are appropriately considered. In this paper, we briefly explain an underwater robot, yShark and its thruster configuration matrix (TCM). Then, methods to achieve fast and accurate motion for depth, heading, and lateral motion control are proposed. Also, this paper presents some experimental works to show incompleteness of thruster configuration, which shows a gap between theory and practice. The proposed schemes are experimentally confirmed by yShark which has been developed by Korea Institute of Ocean Science and Technology (KIOST).

Underwater navigation system with velocity measurement by a horizon Kalman filter

The problem of constructing a receding horizon filter for inertial navigation systems affected by external disturbance has been described. Noises are assumed to be bounded, additive and contained in both state and measurement equations. The estimator is designed according to the sliding-window strategy, so that it minimizes the receding horizon estimation cost function. The derived filter is applied to a velocity aided inertial navigation system. It is clearly demonstrated that the derived filter is more accurate than the standard Kalman filter for underwater navigation systems under the action of temporary unknown disturbances based on simulations.

Underwater robot exploration and identification using dual imaging sonar : Basin test

For the underwater search task, imaging sonar is very useful equipment to obtain the visual data of the underwater environment. Generally, the search range of imaging sonar depends on the frequency of the sound. High-frequency imaging sonar is suitable to search near area within 5m and low-frequency is able to search the area between 5m to 100m. In this paper, we presents experimental results of landmarks identification using dual imaging sonar for the autonomous underwater robot exploration. The underwater robot explores autonomously and estimates the robot location using inertial sensor data. Artificial landmarks are designed and detected by dual imaging sonar. If the landmark is recognized, it is mapped to the estimated robot map. To verify the suitability of the proposed method, we perform an experiment in basin using the underwater robot, ‘yShark’.

Nontemporal relative pose estimation for Opti-Acoustic bundle adjustment

This paper presents the bundle adjustment (BA) between a sonar and optical image for an Opti-Acoustic sensor model. An Opti-Acoustic sensor model allows heterogeneous sensor fusion of an optical camera and acoustic sonar that have a mutual supplementary relation. There are two challenges to the Opti-Acoustic method. First, it is difficult to guarantee a common view from the two significantly different sensor field of view. Second, it is tough to figure out the correspondences between the camera and sonar image, since those sensors have heterogeneous sensor geometries. This paper focuses on the first issue by proposing asynchronously applicable relative motion estimation via BA. For the second issue, we propose using a Dense Adaptive Self-Correlation Descriptor (DASC) to establish correspondences between an optical camera and a sonar image. The proposed Opti-Acoustic BA is validated through autonomous underwater vehicle (AUV) simultaneous localization and mapping (SLAM) in both simulated and real tank tests.

A preliminary study on development of haptic interface for underwater vehicles

This paper presents a method of force feedback for the haptic interface of underwater vehicles. The force feedback is implemented by estimating repulsive force based on the external sensor fusion. The repulsive force of the haptic system is composed of two parts: 1) the applied force to the underwater vehicle, and 2) a virtual force from the underwater obstacles. The applied force to the underwater vehicle is acquired by estimating the status of the vehicle using external sensors. Moreover, the virtual force is acquired from the distance between underwater obstacles and the vehicle. Using both sensorial data, the proposed method generates the repulsive force for the force feedback of haptic interface. The proposed method was verified by simple experiment in a basin.