Sensor Fusion System for Localization of Autonomous Car

In the realm of advancing self-driving vehicle technology, numerous pivotal elements contribute significantly to its progress, with localization standing out as a critical aspect. A robust localization system holds immense significance, allowing a vehicle to precisely ascertain its position on the global map and navigate the complexities of the road network efficiently. Diverse methodologies, including the Global Navigation Satellite System (GNSS), Wheel Odometry, and Inertia Measurement, can be employed to develop this intricate system. Each method brings its own set of advantages and drawbacks, contributing to the nuanced landscape of self-driving vehicle localization. To enhance the accuracy and quality of data, this paper proposes a methodology that integrates information from various sources, such as GNSS, Wheel Odometry, and Inertia Measurement, using the Extended Kalman Filter. This approach serves as a linchpin for obtaining precise and reliable data, aiming to synergistically leverage the strengths of each method while mitigating their weaknesses. The overarching objective is to fortify the overall reliability of the generated data, emphasizing a holistic and integrated perspective. This methodology represents a significant stride toward refining data quality within self-driving vehicle localization systems, contributing to the realization of more robust and dependable autonomous driving technology.