A novel approach to integrated GPS/INS tracking

Modeling an integrated global positioning system / inertial navigation system (GPS/INS) in an optimal sense involves taking into account the non-linear relationships that exist in the physical real world model. Utilizing non-linear filtering techniques such as particle filters instead of the sub optimal Kalman filter and its variants allows one to more accurately model systems of this nature. Systems such as these can also have multiple models for differing scenarios; each model being better equipped to handle a particular set of conditions, and therefore encapsulates a more accurate and complete representation of the system. These systems are therefore suitable candidates for a hybrid system approach. Though these two techniques have been used to great success on their own, not much research has been devoted into the prospects of a merger, and to investigate the vast potential benefits to be gained. This paper present a method of modeling an integrated GPS/INS system utilizing both techniques. This model is based on the multiple model particle filter (MMPF). This paper also illustrates the comparative performance of this approach with other well-known approaches, and outlines the benefits to be gained.