A Fusion Algorithm for Estimating Time-Independent/-Dependent Parameters and States

Area of Science:

• Automotive Engineering
• Control Systems
• Dynamic Analysis

Background:

• Accurate vehicle parameter estimation is crucial for dynamic analysis and control.
• Current methods often require specialized equipment (IPMD) and only estimate partial parameters.
• Key parameters like suspension and tire stiffnesses are frequently assumed known.

Purpose of the Study:

• To propose a fusion algorithm for comprehensive vehicle parameter identification without an IPMD.
• To differentiate and estimate time-independent parameters (TIPs) and time-dependent parameters (TDPs).
• To validate the algorithm's accuracy and convergence using experimental and simulation data.

Main Methods:

• Categorization of vehicle parameters into TIPs and TDPs.
• Utilizing a hybrid-mass state-variable (HMSV) approach for TIP identification.
• Employing a dual unscented Kalman filter (DUKF) for updating TDPs and online states.

Main Results:

• Experimental validation on a two-axle vehicle demonstrated high accuracy in estimating TIPs and updating TDPs.
• Numerical simulations confirmed robust performance under variations in sprung mass, model linearization errors, and diverse road conditions.
• The algorithm achieved quick convergence without requiring prior road information.

Conclusions:

• The proposed fusion algorithm effectively estimates comprehensive vehicle parameters in real-time.
• It eliminates the need for an inertial parameter measurement device (IPMD).
• The method offers high accuracy, rapid convergence, and adaptability to various operating conditions.