• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.23 no.1
• /
• pp.167-181
• /
• 2024

Utilization of a driving simulator in the development of autonomous driving technology allows us to perform various tests effectively in criticial environments, thereby reducing the development cost and efforts. However, there exists a serious drawback that the driving simulator has a big difference from the real environment, so a problem occurs when the autonomous driving algorithm developed using the driving simulator is applied directly to the real vehicle system. This is defined as so-called Sim2Real problem and can be classified into scenarios, sensor modeling, and vehicle dynamics. This Paper presensts on a method to solve the Sim2Real problem in autonomous driving simulator focusing on IMU sensor. In order to reduce the difference between emulated virtual IMU sensor real IMU sensor, IMU sensor emulation techniques through precision error modeling of IMU sensor are introduced. The error model of IMU sensors takes into account bias, scale factor, misalignmnet, and random walk by IMU sensor grades.

• Journal of the Korean Institute of Telematics and Electronics S
• /
• v.35S no.10
• /
• pp.64-76
• /
• 1998

In this paper, we propose the dynamic modeling, path planning and tracking algorithms of 4-wheeled 2-d.o.f.(degree of freedom) mobile robot(WMR). The gaussian functions are applied to design the smooth path of WMR. To calculate the WMR position in real time, we use three components of inertial measurement units(IMU). These units have initial error because of the rotation rate of earth, gravity acceleration and so on. Therefore we derive the initial error model of IMU, and compare the fitness diagnosis about probability characteristics of real data adn estimated data. The performance of IMU with error model and Kalman filter is compared to that without filter and error model. The simulation results show that the proposed dynamic model, path planning and tracking algorithms are more useful than the conventional control algorithm.

• Journal of Institute of Control, Robotics and Systems
• /
• v.19 no.4
• /
• pp.372-378
• /
• 2013

In this paper, we research the localization problem of the crawler-type inspection robot for underwater structure which travels an outer wall of underwater structure. Since various factors of the underwater environment affect an encoder odometer, it is hard to localize robot itself using only on-board sensors. So in this research we used a depth sensor and an IMU to compensate odometer which has extreme error in the underwater environment through using Extended Kalman Filter(EKF) which is normally used in mobile robotics. To acquire valid measurements, we implemented precision sensor modeling after assuming specific situation that robot travels underwater structure. The depth sensor acquires a vertical position of robot and compensates one of the robot pose, and IMU is used to compensate a bearing. But horizontal position of robot can't be compensated by using only on-board sensors. So we proposed a localization algorithm which makes horizontal direction error bounded by using kinematics relationship. Also we implemented computer simulations and experiments in underwater environment to verify the algorithm performance.

• Journal of the Korean Society for Aeronautical & Space Sciences
• /
• v.47 no.8
• /
• pp.566-573
• /
• 2019

In order to obtain high-quality images by motion compensation in the airborne synthetic aperture radar (SAR), accurate motion sensing in image acquisition section is necessary. Especially, reducing relative position error and discontinuity in motion sensing is important. To overcome the problem, we propose a pre-parametric error modeling (P-PEM) algorithm which is a real-time motion sensing algorithm for the airborne SAR in this paper. P-PEM is an extended version of parametric error modeling (PEM) method which is a motion sensing algorithm to mitigate the errors in the previous work. PEM estimates polynomial coefficients of INS error which can be assumed as a polynomial in the short term. Otherwise, P-PEM estimates polynomial coefficients in advance and uses at image acquisition section. Simulation results show that the P-PEM reduces relative position error and discontinuity effectively in real-time.

• Journal of Korea Society of Industrial Information Systems
• /
• v.28 no.1
• /
• pp.17-25
• /
• 2023

As the range of use of urban unmanned aerial vehicles (UAV) expands, it is necessary to operate UAVs efficiently because of its limited battery capacity. For this, it is required to find the optimal flight profile with various simulations. Therefore, it is important to predict the power and energy consumption of the UAV battery. In this paper, we analyzed the relationship between the speed and acceleration of the UAV and power consumption during the flight. Then, we derived a linear model, which is easily utilized. In addition, we also derived an accurate power consumption model based on deep neural network learning. To find the efficient model, we used learning data as 1) the GPS 3-axis velocity and acceleration data, 2) the IMU 3-axis velocity only, and 3) the IMU 3-axis velocity and acceleration data. The final model shows 5.86% error rate for power consumption and 1.50% error rate for the cumulative energy consumption.

• The Journal of Korea Robotics Society
• /
• v.12 no.3
• /
• pp.258-262
• /
• 2017

In this paper, we propose a new collision detection algorithm for human-robot collaboration. We use an IMU sensor located at the tip of the manipulator and the kinematic behavior of the manipulator to detect the unexpected collision between the robotic manipulator and environment. Unlike other method, the developed algorithm uses only the kinematic relationship between the manipulator joint and the end effector. Therefore, the collision estimation signal is not affected by the error of the dynamics model. The proposed collision detection algorithm detects the collision by comparing the estimated acceleration of the end effector derived from the position, velocity and acceleration trajectories of the robot joints with the actual acceleration measured by the sensor. In simulation, we compare the performance of our method with the conventional Residual Observer (ROB). Our method is less sensitive to the load variation because of the independency on the dynamic modeling of the manipulator.