Modularized and hierarchical state estimation strategy of 4WIDEV based on global-local dynamics model decoupling

Area of Science:

• Automotive Engineering
• Control Systems
• Robotics

Background:

• Accurate state estimation is crucial for electric vehicle control.
• Traditional methods often rely on empirical tire models, limiting their applicability.
• Decoupling vehicle dynamics is essential for high-fidelity estimation.

Purpose of the Study:

• To develop a hierarchical estimation strategy for four-wheel independent drive electric vehicles (4WIDEV).
• To achieve high-fidelity state estimation without empirical tire models.
• To improve the accuracy of vehicle state estimation.

Main Methods:

• Global-local dynamic model decoupling.
• Electric drive wheel model (EDWM) and proportional integral observer (PIO) for longitudinal force decoupling.
• Unknown input observer (UIO) for lateral force estimation with stability proofs.
• Hierarchical strategy using strong tracking extended Kalman filter (STEKF).

Main Results:

• Successful decoupling of longitudinal tire force and model state variables.
• Accurate estimation of lateral tire forces on front and rear axles.
• Demonstrated improvement in vehicle state estimation accuracy through simulations and experiments.

Conclusions:

• The proposed hierarchical strategy effectively enhances state estimation for 4WIDEVs.
• The method avoids reliance on empirical tire models, offering broader applicability.
• This approach contributes to more robust and precise vehicle control systems.