Date of Award

2022

Degree Type

Thesis

Degree Name

Doctor of Philosophy

School

School of Engineering

First Advisor

Iftekhar Ahmad

Second Advisor

Daryoush Habibi

Abstract

The success of electronic technologies has sparked interest in sports amongst coaches, athletes and sports scientists. Consequently, the sporting world has witnessed a rapid rise in sports technologies in recent years. In sports, technologies can assist in a variety of ways, ranging from winner selection to athlete injury prevention. In recent years, wearable technologies have become popular in sports. Wearable microtechnology devices, commonly known as electronic performance tracking system (EPTS) in sports, can be used to monitor and improve an athlete's performance. EPTSs can also play an important role in injury prevention. This thesis investigates the performance of wearable EPTSs and presents solutions to improve the accuracy of EPTSs in sports.

For an EPTS to capture and accurately quantify important athlete-related parameters, such as speed, distance covered, number and intensity of sprints, change of direction (COD), and positioning and movement, data is of vital importance. Whilst satellite-based positioning systems require a relatively short setup time, their accuracy and reliability are often low. Wearable inertial sensors may be able to track an athlete's movements, but they can also be noisy and their error accumulates over time. Consequently, designing an accurate and wearable EPTS is a complex task and requires critical investigation. This featured as the key motivation for this research, whose findings are presented in this thesis.

In this thesis, a new ultra-wide band (UWB)-based positioning system is proposed to increase the accuracy of an EPTS in indoor sports. A close analysis of real-world experiments is presented, and limitations associated with the UWB-based positioning systems are highlighted. In forming an economical hardware solution, geometric and machine learning algorithms are introduced. The accuracy of the developed solution is compared against conventional positioning systems. To mitigate errors in UWB-based indoor positioning systems, a range error reduction technique is introduced. The optimal hardware setup and complexity for implementing the proposed positioning solution are also discussed.

In order to track an athlete's movements using real-world data, a multi-sensor data fusion approach is introduced for measuring/tracking athletes' change of direction (COD) detection and quantification. This COD data is then used to reduce divergence errors incurred in state estimation-based particle filter algorithms. The developed solution achieves significant improvements as compared to existing techniques. Overall, the proposed EPTS solution offers higher accuracy in tracking an athlete's position and movement data. Further, the proposed solution enables the use of EPTS in indoor environments.

Access Note

Access to chapter 2, and section 3.2.5 of this thesis is not available.

Some images and figures are not available in this version of the thesis due to copyright considerations.

Included in

Engineering Commons

Share

 
COinS