Autonomous Vehicle State Estimation and Mapping Using Takagi-Sugeno Modeling Approach.

Area of Science:

• Robotics
• Control Systems
• Estimation Theory

Background:

• Accurate state estimation is crucial for autonomous vehicle navigation.
• Nonlinear system dynamics pose challenges for traditional estimation methods.
• Takagi-Sugeno (TS) fuzzy models offer a way to represent complex nonlinear systems.

Purpose of the Study:

• To develop an optimal state estimation approach using a Takagi-Sugeno (TS) Kalman filter.
• To integrate this estimation with a nonlinear model-predictive control (NMPC) system for vehicle control.
• To achieve stable and real-time performance without requiring system linearization.

Main Methods:

• Utilized a Takagi-Sugeno (TS) Kalman filter for state estimation.
• Employed linear matrix inequality (LMI) optimization based on Lyapunov stability and dual linear quadratic regulator (LQR) for TS Kalman gain.
• Leveraged TS fuzzy representation for real-time Kalman gain computation, avoiding repeated LMI optimization.
• Integrated the estimation schema with a nonlinear model-predictive control (NMPC) for vehicle control.

Main Results:

• The proposed TS Kalman filter provides an optimal approach for state estimation.
• Real-time Kalman gain computation was achieved by exploiting the TS fuzzy representation.
• The integrated system demonstrated effective vehicle control through NMPC.
• The approach was validated through simulations and experiments on a small-scale autonomous car.

Conclusions:

• The TS Kalman filter offers an effective method for state estimation in nonlinear systems.
• Real-time performance is achievable, crucial for autonomous systems.
• The integration with NMPC enables robust vehicle control.