As the necessity of wireless charging to support the popularization of electric vehicles (EVs) emerges, the development of a wireless power transfer (WPT) system for EV wireless charging is rapidly progressing. The WPT system requires alignment between the transmitter coils installed on the parking lot floor and the receiver coils in the vehicle. To automatically align the two sets of coils, the WPT system needs a localization technology that can precisely estimate the vehicle’s pose in real time. This paper proposes a novel short-range precise localization method based on ultrawideband (UWB) modules for application to WPT systems. The UWB module is widely used as a localization sensor because it has a high accuracy while using low power. In this paper, the minimum number of UWB modules consisting of two UWB anchors and two UWB tags that can determine the vehicle’s pose is derived through mathematical analysis. The proposed localization algorithm determines the vehicle’s initial pose by globally optimizing the collected UWB distance measurements and estimates the vehicle’s pose by fusing the vehicle’s wheel odometry data and the UWB distance measurements. To verify the performance of the proposed UWB-based localization method, we perform various simulations and real vehicle-based experiments.
The global electric vehicle (EV) market is growing rapidly due to the strengthening of international environmental regulations on vehicle emissions. The technical limitation that should be overcome to accelerate the popularization of EVs is the poor mileage. To this end, the capacity of the battery should be increased, but the current technology does not reach the mileage of internal combustion engine vehicles with a single charge. In addition, the charging time is too long. To compensate for this problem, a wireless power transfer (WPT) system that can easily charge EVs in a parking lot space has been proposed (
Vehicle localization technologies have been developed with different sensors and different methods for indoor or outdoor environments. Localization in outdoor environments uses the Global Positioning System (GPS) and vision sensors with high definition (HD) maps. Localization in indoor environments uses a vision sensor or lidar with prebuilt feature maps or grid maps, as GPS signals are unavailable in these environments. It is thus difficult to apply these conventional localization methods to the WPT system because a map cannot be constructed for all indoor environments.
Recently, many studies have been conducted to utilize an ultrawideband (UWB) distance sensor for vehicle localization technology (
The UWB sensor can also be used for the localization of various objects, such as mobile robots (
This paper proposes a novel short-range precise localization method based on a dual-anchor and dual-tag (DADT) UWB system that can be applied to WPT systems. The proposed DADT UWB-based localization method uses two UWB anchors placed in the parking area and two UWB tags mounted on the vehicle. When the vehicle approaches the parking slot where the WPT is located, the UWB anchors start to communicate with the tags, and the vehicle’s pose, i.e., position and heading angle, is initialized by processing the UWB distance measurement data. Then, the wheel odometry information and UWB distance measurements are fused based on a particle filter framework to continuously estimate the pose from the initial vehicle pose. The goal of this paper is to make a precise pose estimation so that the final parking position of the vehicle has an error of less than 0.1 m, which is required for alignment between the power transmitter and receiver coils of the WPT. To verify the performance of the proposed DADT UWB-based localization method, we perform various simulations and experiments with an actual vehicle.
The preliminary results of this paper were presented in
The rest of this paper is organized as follows: we introduce the WPT system for EVs and describe the proposed DADT UWB localization method. To verify the performance of the proposed method, simulation results with various scenarios and experimental results with a real vehicle are presented. Finally, a conclusion is presented.
The basic working principle of WPT for EVs is as follows (
When the power transmitter and receiver coils are kept close to each other, electric energy can be used to charge the battery. The alignment of the two coils is significant for the high performance and efficiency of the WPT.
Two UWB anchors are placed on both corners of the parking slot and two UWB tags are mounted on the vehicle. The observability of the vehicle’s pose estimated by the proposed DADT UWB system is analyzed based on the Fisher information matrix.
Localization technologies that can be applied in parking lot environments have been developed based on mono cameras (
The use of UWB sensors can overcome the limitations of cameras and lidar sensors. The proposed DADT localization method only needs to know the positions of two anchors and the transmitting coil installed in the parking slot. Thus, it does not require an inconvenient process of building a high-precision map with cameras or lidar sensors. As the UWB sensor is based on RF signals, it is unaffected by changes in lighting and robust against dynamic obstacles such as vehicles and pedestrians. In addition, if a pair of UWB sensors has a clear line of sight, the distance between them can be precisely measured with an error of approximately 0.05 m- 0.1 m. Due to the economics of UWB sensors, many automotive makers have plans to use UWB sensors in vehicles soon. Therefore, it is possible to implement a precise localization applicable to WPT for EVs with economical cost.
This section describes a novel dual-anchor and dual-tag (DADT) UWB-based localization method that can precisely estimate the pose of a vehicle near a parking area. The key idea of the proposed DADT UWB-based localization is shown in
To show the effectiveness of the proposed DADT UWB sensor system, the condition in which the vehicle’s pose can be uniquely determined by the DADT UWB sensor system is analytically derived. Then, observability analysis based on FIM is performed on the proposed DADT UWB sensor system.
We denote the vehicle pose state vector at a time step
The uncertainty of the pose of the vehicle estimated by UWB distance measurements is determined by the geometric distribution of the anchors fixed on the parking lot and the tags mounted on the vehicle. To estimate the amount of uncertainty about the vehicle pose estimated by the proposed DADT UWB system, FIM-based observability analysis is performed as follows. The FIM can be defined as (
(A)
In the first step, the vehicle’s pose is initialized by globally optimizing the UWB distance measurements. In the second step, based on the initialized vehicle pose, the wheel odometry and UWB distance measurement collected as the vehicle moves are fused to estimate the pose of the vehicle in real time.
The determinant value of the FIM represents the amount of Fisher information that can be observed for vehicle pose state variables; i.e., as the determinant of the FIM increases, the vehicle pose can be estimated with higher accuracy.
The proposed DADT UWB localization algorithm consists of two major steps, as shown in
The purpose of this step is to quickly find the approximate initial pose of the EV using only UWB distance measurements under the assumption that no prior information about the EV’s pose is available. From the mathematical analysis of the proposed DADT UWB system, the EV pose can always be uniquely determined by the DADT UWB system in the area of
We propose a particle swarm optimization (PSO) (
The PSO-based vehicle pose initialization method is as follows. When a vehicle approaches a parking area where communication between UWB anchors and tags is possible, UWB distance measurements between each UWB tag and anchor pair are sampled. When a certain number of measurements is collected, the average value is estimated by removing outliers. The error function
To estimate the vehicle’s pose precisely as the vehicle moves, the UWB distance measurements and wheel odometry data are fused through a particle filter. The method of estimating the vehicle’s pose through the particle filter is shown in the right block of
In particle filter-based localization, a group of particles represents the probability distribution of vehicle states, with each particle
The motion model of the vehicle is
The vehicle pose
The vehicle starts from four different positions and moves along the path for front-end parking or back-in parking.
To verify the performance of the proposed DADT UWB localization method, we perform the following simulations: (1) Initialize the vehicle’s pose by globally optimizing the error function defined in
True pose | Levenberg–Marquardt | Proposed DADT | |||
---|---|---|---|---|---|
Estimated Pose | Estimated Pose | ||||
Test 1 | 2.05 | 1.42E–02 | |||
Test 2 | 2.37 | 2.24E–02 | |||
Test 3 | 0.48 | 5.83E–02 | |||
Test 4 | 0.76 | 2.24E–02 |
The vehicle starts from its initial pose
The vehicle starts from its initial pose
The vehicle starts from its initial pose
The vehicle starts from its initial pose
Odometry | Proposed DADT | |||||||
---|---|---|---|---|---|---|---|---|
Mean | Min | Max | Std | Mean | Min | Max | Std | |
Test 1 | 0.3396 | 0.0012 | 0.8322 | 0.2613 | 0.0691 | 0.0016 | 0.2831 | 0.0476 |
Test 2 | 0.3555 | 0.0140 | 0.6925 | 0.1414 | 0.0720 | 0.0129 | 0.2220 | 0.0397 |
Test 3 | 0.3930 | 0.0034 | 0.7266 | 0.2362 | 0.0714 | 0.0035 | 0.2075 | 0.0382 |
Test 4 | 0.3831 | 0.0032 | 0.5838 | 0.1525 | 0.0732 | 0.0030 | 0.2344 | 0.0401 |
The proposed DADT UWB-based localization method is tested with a real vehicle. The tests are performed with the UWB modules manufactured by Pozyx (
(A) Two UWB anchors are installed near both corners of the parkingslot. (B) two UWB tags and DGPS are mounted on the vehicle roof.
The vehicle is performing front-end parking.
The black dashed-dotted line shows the DGPS trajectory, the red dotted line shows the odometry trajectory, and the blue solid line shows the proposed DADT UWB-based localization results.
The black dashed-dotted line shows the DGPS trajectory, the red dotted line shows the odometry trajectory, and the blue solid line shows the proposed DADT UWB-based localization results.
Distance error | ||||
---|---|---|---|---|
0.4690 | 0.5826 | 0.7479 | ||
0.0493 | 0.0763 | 0.0908 | ||
0.9038 | 0.1604 | 0.9179 | ||
0.0328 | 0.0031 | 0.0329 |
Through the experiments, the average computation time required to update the vehicle’s pose at each time instant is estimated while increasing the number of particles from 20 to 100.
This paper proposed a novel short-range precise localization method using a DADT UWB sensor system for application to a WPT system. An observability analysis of the proposed DADT UWB sensor system consisting of two anchors and two tags was performed based on the FIM. The proposed localization algorithm determines the vehicle’s initial pose by globally optimizing the collected UWB distance measurements and estimates the vehicle’s pose by fusing the vehicle’s wheel odometry data and the UWB distance measurements. The effectiveness of the proposed method was confirmed through various simulations and real vehicle-based experiments.
The source code includes (i) DADT UWB-based localization algorithm, (ii) particle filter framework to estimate a vehicle’s pose.
The authors declare there are no competing interests.
The following information was supplied regarding data availability:
The source code is available in the