Jun-Hyuck Im

GPS/DR Error Estimation for Autonomous Vehicle Localization

Autonomous vehicles require highly reliable navigation capabilities. For example, a lane-following method cannot be applied in an intersection without lanes, and since typical lane detection is performed using a straight-line model, errors can occur when the lateral distance is estimated in curved sections due to a model mismatch. Therefore, this paper proposes a localization method that uses GPS/DR error estimation based on a lane detection method with curved lane models, stop line detection, and curve matching in order to improve the performance during waypoint following procedures. The advantage of using the proposed method is that position information can be provided for autonomous driving through intersections, in sections with sharp curves, and in curved sections following a straight section. The proposed method was applied in autonomous vehicles at an experimental site to evaluate its performance, and the results indicate that the positioning achieved accuracy at the sub-meter level.

Vertical Corner Feature Based Precise Vehicle Localization Using 3D LIDAR in Urban Area

Tall buildings are concentrated in urban areas. The outer walls of buildings are vertically erected to the ground and almost flat. Therefore, the vertical corners that meet the vertical planes are present everywhere in urban areas. These corners act as convenient landmarks, which can be extracted by using the light detection and ranging (LIDAR) sensor. A vertical corner feature based precise vehicle localization method is proposed in this paper and implemented using 3D LIDAR (Velodyne HDL-32E). The vehicle motion is predicted by accumulating the pose increment output from the iterative closest point (ICP) algorithm based on the geometric relations between the scan data of the 3D LIDAR. The vertical corner is extracted using the proposed corner extraction method. The vehicle position is then corrected by matching the prebuilt corner map with the extracted corner. The experiment was carried out in the Gangnam area of Seoul, South Korea. In the experimental results, the maximum horizontal position error is about 0.46 m and the 2D Root Mean Square (RMS) horizontal error is about 0.138 m.

A Path Generation Algorithm for Obstacle Avoidance in Waypoint Navigation of Unmanned Ground Vehicle

In this paper, an effective path generation algorithm for obstacle avoidance producing small amount of steering action as possible is proposed. The proposed path generation algorithm can reduce unnecessary steering because of the small lateral changes in generated waypoints when UGV (Unmanned Ground Vehicle) encounters obstacles during its waypoint navigation. To verify this, the proposed algorithm and algorithm are analyzed through the simulation. The proposed algorithm shows good performance in terms of lateral changes in the generated waypoint, steering changes of the vehicle while driving and execution speed of the algorithm. Especially, due to the fast execution speed of the algorithm, the obstacles that encounter suddenly in front of the vehicle within short range can be avoided. This algorithm consider the waypoint navigation only. Therefore, in certain situations, the algorithm may generate the wrong path. In this case, a general path generation algorithm like is used instead. However, these special cases happen very rare during the vehicle waypoint navigation, so the proposed algorithm can be applied to most of the waypoint navigation for the unmanned ground vehicle.

Pedestrian Safety Road Marking Detection Using LRF Range and Reflectivity

In this paper, a detection method of a pedestrian safety road marking was proposed. The proposed algorithm uses laser range and reflectivity of a range finder (LRF). For a detection of crosswalk marking and stop line, the DFT (Discrete Fourier Transform) of reflectivity and cross-correlation method between the reference replica and the measured reflectivity are used. A speed bump is detected through measuring an altitude difference of two LRFs which have the different tilted angle. Furthermore, we proposed a velocity constrained a detection method of a speed bump. Finally, the proposed methods are tested in on-line, on the pavement of a road. The considered road markings are wholly detected. The localization errors of both road markings are smaller than 0.4 meter.

An Analysis of Spoofing Effects on a GNSS Receiver Using Real-Time GNSS Spoofing Simulator

In this paper, spoofing effects on a GNSS receiver were analyzed. The spoofer (spoofing device) was classified to two categories. One is an active spoofer and the other is a passive spoofer. The active spoofer was considered for analysis. For the analysis of spoofing effects on a GNSS receiver, a real-time GNSS spoofing simulator was developed. The simulator was consisted with two parts which are a baseband signal generation part and a RF up-conversion part. The first GNSS baseband signal was generated according to spoofing parameters such as range, range rate, GNSS navigation data, spoofing to GNSS signal ratio, and etc. The generated baseband signal was up-converted to GNSS L1 band. Then the signal transmitted to a GNSS signal. For a perfect spoofing, a spoofer knew an accurate position and velocity of a spoofing target. But, in real world, that is not nearly possible. Although uncertainty of position and velocity of the target was existed, the spoofer was operated as an efficient jammer.

Obstacle Parameter Modeling for Model Predictive Control of the Unmanned Vehicle

The MPC (Model Predictive Control) is one of the techniques that can be used to control an unmanned vehicle. It predicts the future vehicle trajectory using the dynamic characteristic of the vehicle and generate the control value to track the reference path. If some obstacles are detected on the reference paths, the MPC can generate control value to avoid the obstacles imposing the inequality constraints on the MPC cost function. In this paper, we propose an obstacle modeling algorithm for MPC with inequality constraints for obstacle avoidance and a method to set selective constraint on the MPC for stable obstacle avoidance. Simulations with the field test data show successful obstacle avoidance and way point tracking performance.

Extended Line Map-Based Precise Vehicle Localization Using 3D LIDAR

An Extended Line Map (ELM)-based precise vehicle localization method is proposed in this paper, and is implemented using 3D Light Detection and Ranging (LIDAR). A binary occupancy grid map in which grids for road marking or vertical structures have a value of 1 and the rest have a value of 0 was created using the reflectivity and distance data of the 3D LIDAR. From the map, lines were detected using a Hough transform. After the detected lines were converted into the node and link forms, they were stored as a map. This map is called an extended line map, of which data size is extremely small (134 KB/km). The ELM-based localization is performed through correlation matching. The ELM is converted back into an occupancy grid map and matched to the map generated using the current 3D LIDAR. In this instance, a Fast Fourier Transform (FFT) was applied as the correlation matching method, and the matching time was approximately 78 ms (based on MATLAB). The experiment was carried out in the Gangnam area of Seoul, South Korea. The traveling distance was approximately 4.2 km, and the maximum traveling speed was approximately 80 km/h. As a result of localization, the root mean square (RMS) position errors for the lateral and longitudinal directions were 0.136 m and 0.223 m, respectively.

Position and attitude estimation of a vehicle using laser scanner in the obstacle environment

Laser scanner provides very accurate distance information up to around landmark from the sensor. It is possible to estimate the position and orientation relative to the drive from the surrounding landmarks that are recognized by utilizing these points of the laser scanner. In this paper, after setting arbitrarily the environment of the obstacle, we performed the simulation of the position and attitude estimation of the vehicle by generating the laser scanner data. At the simulation result, position error was within 3m, attitude error was revealed in 10 degrees or less.

Software-Based Loran-C Signal Processing

With GPS being the primary navigation system, Loran use is in steep decline. However, according to the final report of vulnerability assessment of the transportation infrastructure relying on the global positioning system prepared by the John A. Volpe National Transportation Systems Center, there are current attempts to enhance and re-popularize Loran as a GPS backup system through the characteristic of the ground based low frequency navigation system. To advance the Loran system such as Loran-C modernization and eLoran development, research is definitely needed in the field of Loran-C receiver signal processing as well as Loran-C signal design and the technology of a receiver. We have developed a set of Matlab tools, which implement a software Loran-C receiver that performs the receiver`s position determination through the following procedure. The procedure consists of receiving the Loran-C signal, cycle selection, calculation of the TDOA and range, and receiver`s position determination through the Least Square Method. We experiences the effect of an incorrect cycle selection and various error factors (ECD, ASF, sky wave, CRI, etc.) from the result of the Loran-C signal processing. It is apparent that researches which focus on the elimination and mitigation of various error factors need to be investigated on a software Loran-C receiver. These aspects will be explored in further work through the method such as PLL and Kalman filtering.