Wheel Speed Identification and Longitudinal Vehicle Speed Estimation for Four-Wheel Drive Vehicles

To address the issue of vehicle speed estimation, particularly at low speeds, applicable to various environmental conditions, especially in the absence of driven wheels for vehicles equipped with four-wheel drive systems, a rule-based wheel speed identification and vehicle speed estimation method is proposed. This technique leverages wheel speed sensors and longitudinal acceleration sensors for vehicles to enhance the precision of wheel speed identification at low speeds and enhance the adaptability of vehicle speed estimation under different conditions. Firstly, to tackle the problem of no wheel speed pulse signal passing within one sampling period at low speeds, estimation is achieved through time accumulation, enabling effective wheel speed identification for both high and low-speed. Furthermore, utilizing the Kalman filtering algorithm and Butterworth filter, the wheel speed and wheel acceleration are filtered separately to reduce signal fluctuations caused by noise interference, thereby providing reliable input signals for estimating vehicle speed. Building upon this, A method for estimating vehicle speed based on the gradient constraint approach is proposed, taking into account the sequential relationships of each wheel speed and the longitudinal acceleration signal. Finally, real vehicle tests under different conditions demonstrate that the proposed vehicle speed estimation technique achieves high estimation accuracy, meeting the vehicle's speed estimation requirements under various environmental factors.