The design and implementation of advanced driver assistance system (ADAS) data acquisition engine based on heterogeneous computation platform
<p>ADAS supports drivers with the required tools and augments to properly decide while</p><p>driving the car. It can safely control the car either by providing relevant information</p><p>around the car to the driver, or by taking...
Saved in:
| Main Author: | |
|---|---|
| Format: | thesis |
| Language: | eng |
| Published: |
2021
|
| Subjects: | |
| Online Access: | https://ir.upsi.edu.my/detailsg.php?det=7343 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | <p>ADAS supports drivers with the required tools and augments to properly decide while</p><p>driving the car. It can safely control the car either by providing relevant information</p><p>around the car to the driver, or by taking control of the vehicle movement, partially or</p><p>completely. The purpose of this study is to develop an ADAS Data Acquisition Engine</p><p>based on heterogenous platform that utilises SoC and FPGA platforms. Furthermore,</p><p>the design methodology of heterogenous SoC-FPGA platform is developed to reduce</p><p>design complexity of heterogenous design flow. Furthermore, it unifies the hardware</p><p>and software design flows to reduce design cycle time required for development. The</p><p>proposed system was verified indoor to confirm system functionality. After that, the</p><p>proposed system was implemented on car for real-time system validation and testing.</p><p>The system was able to interact with LiDAR sensor, four ultrasonic sensors, and Inertial</p><p>Movement Unit (IMU) sensors. The LiDAR and ultrasonic were used for long distance,</p><p>and short distance measurements, respectively. The proposed system implemented on</p><p>FPGA consumed 15% of logic resources, and 76% of internal memory. The proposed</p><p>test plan has been derived based on case study reliability tests, system functionality</p><p>tests, data validation tests, and ADAS application functionality tests. Each test was</p><p>executed four times to ensure system reliability. The proposed system was able to detect</p><p>objects in short-range perspective through the ultrasonic sensor from 20 centimetres to</p><p>450 centimetres. Furthermore, the system was able to detect long-range distance</p><p>through the LiDAR from 4 meters, up to 70 meters. The car steering wheel orientation</p><p>was measured through the IMU sensor ranged from 58.8 angle (clockwise steering</p><p>movement) to -61.4 angle (anti-clockwise steering movement). The collected data was</p><p>postprocessed through Rapidminer studio software tool and was presented for further</p><p>Artificial Intelligence (AI) future applications.</p> |
|---|