Modeling and Control of an Active Stabilizing Assistant System for a Bicycle.

Chih-Keng Chen,Trung-Dung Chu,Xiao-Dong Zhang

doi:10.3390/s19020248

Chih-Keng Chen, Trung-Dung Chu + Show 1 more

Open Access

https://doi.org/10.3390/s19020248

Copy DOI

Abstract

This study designs and controls an active stabilizing assistant system (ASAS) for a bicycle. Using the gyroscopic effect of two spinning flywheels, the ASAS generates torques that assist the rider to stabilize the bicycle in various riding modes. Riding performance and the rider’s safety are improved. To simulate the system dynamic behavior, a model of a bicycle–rider system with the ASAS on the rear seat is developed. This model has 14 degrees of freedom and is derived using Lagrange equations. In order to evaluate the efficacy of the ASAS in interacting with the rider’s control actions, simulations of the bicycle–rider system with the ASAS are conducted. The results for the same rider for the bicycle with an ASAS and on a traditional bicycle are compared for various riding conditions. In three cases of simulation for different riding conditions, the bicycle with the proposed ASAS handles better, with fewer control actions being required than for a traditional bicycle.

Highlights

The modeling of bicycle rider control is a common research topic
The rider control model (MPC1) used both steering and leaning torque to stabilize the bicycle from an initial angle
InResults the simulations, the rider control model (MPC1) used both steering and leaning torque to and Discussions stabilize the bicycle from an initial angle

Summary

Introduction

The modeling of bicycle rider control is a common research topic. Van Lunteren et al [1] used classical control methods to model rider control in order to study the effect of drugs, alcohol, and other factors on the ability of the rider to stabilize a bicycle. In a control-theoretical study, Hess et al [2]. Developed a bicycle rider model that is similar to an aircraft pilot model. This model includes gain values, a second-order filter, and a preview time. Another approach to modeling rider control is optimal control. Schwab [3] used an LQR

Objectives

Results

Conclusion