Slope estimation based on multi-source information fusion

Abstract This study presents a novel methodology for slope estimation, addressing limitations in current unidirectional estimation research. A comprehensive vehicle kinematics and dynamics model is developed using Inertial Measurement Unit (IMU) data, enabling a thorough assessment of vehicle slope. The proposal advocates employing the Interactive Multiple Model Unscented Kalman Filter algorithm to enhance algorithmic stability and prevent divergence from singular model errors. This approach ensures precise computation of dynamic slopes. The algorithm’s efficacy is demonstrated through collaborative simulations on the CarSim/Simulink platform. Introducing innovative concepts and methods for slope estimation, this research has significant practical and theoretical implications.