Development of Monitoring, Modelling and Control Systems for Human Physiological Assessment with Wearable Devices
- Publication Type:
- Thesis
- Issue Date:
- 2020
Open Access
Copyright Clearance Process
- Recently Added
- In Progress
- Open Access
This item is open access.
Physiological signals play vital roles in studying the mechanism of human body reaction during exercise and human kinetics assessment. This thesis develops a wearable exercise monitoring system to monitor and regulate human cardiorespiratory responses to moderate exercise. To describe the relationship between the body’s physiological reactions and the exercise, the modelling approach has been extensively explored in a range of applications. In this thesis, the cardiorespiratory signal responses to the exercise phase are comprehensively analysed through the means of different modelling approaches. A non-parametric kernel based modelling approach has been proposed to address the complexity of the model dynamics. This thesis also develops a novel Inclination based Calibration method to address the static nonlinear modelling problem for the calibration of the sensors in an Inertial Measurement Units.
The non-parametric model is the preferable method when the system structure information is insufficient, or the system is too complex to be described by a simple parametric model. Hence, the non-parametric modelling method with kernel-based regularisation is developed to estimate the physiological signal response to the exercise phase during different types of exercise. The kernel selection and regularisation strategies are discussed, and a series of simulations are performed to compare the fitness, sensitivity and stability of different kernels.
For detecting the exercise phase, the innovative in-field calibration method for the portable tri-axial sensor is developed to calibrate the Inertial Measurement Units data. Based on the fact that the angle between the local gravity and magnetic field is invariant, this thesis proposed a new in-field calibration approach, called Inclination Based Calibration, which can reliably estimate the model parameters of the sensor with a simple linear Least Square estimator. Based on optimal experimental design, a 12-observation Icosahedron experimental scheme has been performed for micro Inertial Measurement Units. Both the calibrated results and the simulation comparison demonstrate the effectiveness of the proposed method.
This monitoring and control system could comprehensively study human kinetics and cardiorespiratory mechanism and help to make assessments. Some general approaches for physiological signal processing and modelling, parameters estimation, sensor calibration and experiment protocol control are proposed in this work. The effectiveness and benefits of different modelling approaches are demonstrated by a range of means. This system could be applied in strategic exercise design, athletic assessment, exercise enhancement and health monitoring.
Please use this identifier to cite or link to this item: