MULTI-SENSOR PERSONAL NAVIGATOR SUPPORTED BY HUMAN MOTION DYNAMICS MODEL

This paper presents preliminary results of a prototype design and implementation of a multi-sensor personal navigator suitable for navigation in open areas and confined environments. This work, supported by the National Geospatial-Intelligence Agency (NGA), is focused on forming theoretical foundations for such a system by developing the algorithmic concept of a basic GPS-based, Micro-electro-mechanical inertial measurement unit (MEMS IMU)-augmented personal navigator system with an open-ended architecture, which would be able to incorporate additional navigation and imaging sensor data, extending the system's operations to indoor environments. The accuracy requirement is considered at 3-5 m CEP (circular error probable). In the current system design and implementation, the following sensors are integrated in the tightly coupled Extended Kalman Filter: GPS pseudoranges, Crossbow IMU400C, PTB220A barometer and Azimuth 1000 digital compass. In order to bridge GPS signal gaps in impeded environments, the dynamic model of human locomotion is currently included in the system architecture. The system is trained under the open sky conditions, where GPS signals are available, and is subsequently used to support navigation when GPS signals are obstructed. The calibrated model of stride length and stride interval extracted from the test data provided by GPS/IMU, and heading information from compass and IMU offer dead reckoning navigation, facilitating bridging of GPS gaps.