Inertial Pocket Navigation System: Unaided 3D Positioning.

Area of Science:

• Engineering
• Computer Science
• Biomedical Engineering

Background:

• Inertial navigation systems (INS) commonly use dead reckoning for pedestrian positioning.
• Existing step-and-heading approaches in INS often omit vertical displacement estimation.
• The strapdown algorithm, while accounting for vertical displacement, involves complex double integration of accelerometer data.

Purpose of the Study:

• To develop the first step-and-heading approach for 3D pedestrian positioning using only unaided inertial data.
• To introduce a novel step detector capable of identifying both upward and downward steps.
• To create a new vertical displacement estimator for enhanced pedestrian tracking.

Main Methods:

• Utilizing inertial data from a sensor placed in the front trouser pocket, simulating smartphone placement.
• Developing algorithms based on the leg's opening angle and the pitch angle for motion analysis.
• Implementing and evaluating a new step detector and a vertical displacement estimator against state-of-the-art methods.

Main Results:

• The proposed algorithms demonstrate superior performance compared to existing literature on step-and-heading inertial navigation.
• The system successfully achieves 3D pedestrian positioning using solely unaided inertial data.
• Analysis of pitch angle data enabled the differentiation of five distinct activities: standing, sitting, walking, walking up stairs, and walking down stairs.

Conclusions:

• The novel step-and-heading approach effectively solves 3D pedestrian positioning with unaided inertial sensors.
• The system's ability to estimate vertical displacement and activity type offers valuable data for applications like elderly care monitoring.
• The proposed algorithms represent a significant advancement in pedestrian dead-reckoning techniques.