A Novel Methodology for Inertial Parameter Identification of Lightweight Electric Vehicle via Adaptive Dual Unscented Kalman Filter

Lightweight electric vehicles (LEVs) possess great advantages in the viewpoint of fuel consumption, environment protection, and traffic mobility. However, due to the drastic reduction of vehicle weights and body size, the effects of inertial parameter variation in LEV control system become much more pronounced and have to be systematically estimated. This paper presents a dual adaptive unscented Kalman filter (AUKF) where two Kalman filters run in parallel to synchronously estimate vehicle inertial parameters and additional dynamic states such as vehicle mass, vehicle yaw moment of inertia, the distance from front axle to centre of gravity and vehicle sideslip angle. The proposed estimation only integrates and utilizes real-time measurements of in-wheel-motor information and other standard in-vehicle sensors in LEV. The LEV dynamics estimation model including vehicle payload parameter analysis, Pacejka model, wheel and motor dynamics model is developed, the observability of the observer is analysed and derived via Lie derivative and differential geometry theory. To address nonlinearities and undesirable noise oscillation in estimation system, the dual noise adaptive unscented Kalman filter (DNAUKF) and dual unscented Kalman filter (DUKF)are also investigated and compared. Simulation with various manoeuvres are carried out to verify the effectiveness of the proposed method using MATLAB/Simulink-Carsim^®. The simulation results show that the proposed DNAUKF method can effectively estimate vehicle inertial parameters and dynamic states despite the existence of payload variations.