Online wheel speed filtering for periodic disturbance reduction: a strategy for an advanced bicycle application

TL;DR

This paper introduces an online filtering method for bicycle wheel speed measurements to reduce periodic noise, improving cadence estimation for advanced vehicle dynamics applications.

Contribution

The paper presents a novel online filtering strategy based on a sensor model, specifically designed for bicycles, outperforming traditional filters in cadence estimation accuracy.

Findings

Effective noise reduction demonstrated on experimental data

Improved cadence estimation accuracy over traditional filters

Reliable for real-time bicycle speed measurement applications

Abstract

Due to geometrical errors and possible misalignment of the sensors, wheel speed measurements provided by incremental encoders in road vehicles are usually affected by significant periodic noises. This paper presents an online wheel speed filtering procedure, based on a model of the sensor, aimed at processing the speed measurement to make it suitable for advanced vehicle dynamics applications. In particular, differently from low-pass and notch filtering, this strategy is reliable for the cycling cadence estimation from the wheel speed on bicycles. Experimental data are used to show the effectiveness of the proposed approach.