|
Nonlinear Pedestrian Tracking by Unscented Particle Filter for Autonomous Vehicle under Existence of Non-Gaussian Distributed Uncertainty |
|
Jin Ho Yang, Ju Won Seo, Chung Choo Chung(Hanyang University, Korea) |
| We propose an Unscented Particle Filter(UPF) based pedestrian tracking system with non-Gaussian uncertainty. Accurate recognition and state estimation of pedestrians are required for successful autonomous driving. We performed modeling to estimate the nonlinear movement of a pedestrian. To confirm the performance and effectiveness of UPF, multiple simulations were conducted under various uncertainty scenarios. In particular, the performance was compared under the non-Gaussian distribution characteristic that the object tracking model and recognition by sensors can have. As a result, the proposed UPF-based tracking method for all results has the slightest error compared to other baselines. |
|
|
A-BERF: Action-weighted Ensemble by Bootstrapping Extremely Randomized Forest for Pre-Crash Moral Decision in Autonomous Driving |
|
Jin Ho Yang, Chung Choo Chung(Hanyang University, Korea) |
| This study proposes a novel and high-precision decision-making methodology of an Action-weighted ensemble by Bootstrapping Extremely Randomized Forest (A-BERF) for the moral collision dilemma during urban autonomous driving. By simulating the pedestrian-crossing situation, the decision result from the experiment participants and the features were combined into the dataset. The performance between the tree or forest-based ensemble baselines and A-BERF was compared. A-BERF had the highest classification accuracy and the lowest feature bias compared to other ensemble methods using various types of datasets in validation. In addition, we confirmed that the operation time was improved. |
|
|
Dynamic Obstacle Avoidance for Autonomous Motorcycles on Racing Circuits by Nonlinear Model Predictive Control |
|
Michikuni Eguchi, Hisashi Date(University of Tsukuba, Japan) |
| This paper presents a method for dynamic obstacle avoidance of autonomous motorcycles. In motorcycle racing on circuits, it is necessary to balance the motorcycle's posture, minimize travel time, and avoid collisions with other competing motorcycles. This study proposes a control system using Model Predictive Control (MPC) to achieve these objectives. The constraints of the MPC are given based on the motorcycle's lean angle, the width of the course, the obstacle's trajectory, and the obstacle's size. These allow the motorcycle to balance posture and avoid moving obstacles in various directions on the course effectively. The proposed control system was tested on a physical simulator. |
|
|
Improving Indoor Positioning Systems with UWB and Filtering Techniques: A Comparative Analysis |
|
Rahul Ranjan(Sunmoon University, Korea), Shin Donggyu, Younsik Jung(Sun Moon University, Korea), Sanghyun Kim(Kyung Hee University, Korea), Chang-Hyun Kim(Korea Institute of Machinery & Materials, Korea), Seungjae Lee, Joongeup Kye(Sun Moon University, Korea) |
| The study focuses on the significance of indoor positioning systems (IPS) in today's world, particularly for mobile robots in dynamic indoor environments. The complexity arises from multiple positioning techniques, obstacles, and material variations. The challenge lies in refining accuracy and reducing noise in indoor localization systems. The paper examines an Ultra-Wideband (UWB) based localization system, utilizing two methods for measurement enhancement: Low-Pass Filter (LPF) and Kalman Filter (KF), integrated into the Robot Operating System (ROS) platform.
|
|
|
Controller Design for Heave and Surge Motions for AUVs |
|
Sandhya Rani Gumpina(Konkuk university, Korea), Tae sam Kang(konkuk university, Korea) |
| In this study, we created a simulation block for an underwater vehicle and designed simple 2 DOF PID controllers for heave and surge maneuvers using linearly approximated models. Despite minor oscillations caused by nonlinear couplings, both controllers exhibited sufficient stability margins when applied to the original nonlinear models. These results indicate the potential usability of the proposed controllers in real-world applications.
|
|
|
A study on analysis of wheel characteristics using amphibious toys |
|
Jongho Bae(KIRO, Korea), SUNGHO JIN, HYUN SEOK SHIN, MAN GI LEE, SUNG HO PARK, GUN RAE CHO(KIRO(KOREA INSTITUTE OF ROBOTICS & TECHNOLOGY CONVERGENCE), Korea) |
| A study on analysis of wheel characteristics using amphibious toys |
|