A Cost-Effective and Highly Accurate Observer for the Real-Time Estimation of the Vehicle Velocity and the Road Inclination and Bank Angles

Abstract

This paper proposes a low cost and low-complexity observer that can effectively and accurately estimate in real-time the longitudinal and lateral vehicle velocities, and the road inclination and bank angles. The above are needed in several systems of a vehicle, such as the battery energy management of an electric vehicle, the active suspension and other driver assistance systems. The suggested observer utilizes the neural network technique and is realized by using information provided by sensors that measure the angular velocity of the wheels, the position of the accelerator and decelerator pedals, and the angle of the steering wheel, with which the modern vehicles are usually equipped. The only additional sensors that are required are four inertial measurement units mounted on each wheel-carrier. However, they are relatively low-cost sensors and therefore, the total cost of the vehicle does not considerably increase. The training of the neural network can be easily performed in a test road. The feasibility and effectiveness of the suggested observer are verified on simulation models with the CarMaker and Matlab/Simulink programs. Several simulation results are presented to validate the satisfactory operation of the observer, which are also compared with other techniques from the literature.