Review of State Estimation Methods for Autonomous Ground Vehicles: Perspectives on Estimation Objects, Vehicle...

Area of Science:

• Automotive Engineering
• Control Systems
• Machine Learning

Background:

• Vehicle driving state estimation is crucial for advanced driver-assistance systems (ADAS) and autonomous driving.
• Accurate estimation of vehicle dynamics and environmental conditions is essential for safe and efficient operation.

Purpose of the Study:

• To provide a comprehensive review of vehicle driving state estimation research.
• To analyze current techniques from multiple perspectives: estimation objects, vehicle characteristics, and algorithms.
• To identify future research trends in vehicle state estimation.

Main Methods:

• Review of existing literature on vehicle state estimation.
• Categorization of techniques based on estimation objects (attitude, dynamics, environment).
• Analysis of methods considering vehicle characteristics (coupling, multi-source data, state transitions).
• Examination of key algorithms: Kalman filtering, machine learning, and hybrid approaches.

Main Results:

• Identified key research areas within vehicle state estimation.
• Highlighted the strengths and weaknesses of different algorithmic approaches.
• Showcased the importance of considering vehicle characteristics for robust estimation.

Conclusions:

• Vehicle state estimation is a rapidly evolving field with significant implications for automotive safety and autonomy.
• Hybrid approaches combining model-based and data-driven methods show promise.
• Further research is needed to address complex vehicle dynamics and environmental uncertainties.