How to deal with low performance IMUs in an integrated navigation system: Step by step

Abstract

The Project of Research and Technologies on Unmanned Air Vehicles (PITVANT) undertaken and led by the Portuguese Air Force Academy (PAFA) had its beginning in January, 2009. The PITVANT project main focus goes to the development of new technologies and operation concepts for unmanned aerial vehicles (UAV) of small dimensions, keeping in mind the necessary field tests for the transition to bigger vehicles. Within the project objectives are the development, construction and testing of small and medium platforms; cooperative control of various vehicles; enhanced systems' interoperability; development of advanced vision systems; fusion of multiple sensor data and development of navigation systems. Fulfilling the previous goals allows for the usage of UAV's in multiple mission scenarios such as Intelligence, Surveillance and Reconnaissance (ISR), combat missions performed by cooperative teams of unnamed vehicles, perimeter surveillance missions, etc. Among the wide range of objectives of such a project, our focus is the PITVANT UAV navigation system. The main focus is on the development and validation of a low performance IMU/GNSS navigation system, which can, in the meantime, replace the commercial autopilot existent onboard the UAVs. However, the task of selecting the correct Inertial Measurement Unit (IMU) to use in a specific integrated navigation system is not as easy as it could seem. As a matter of fact, there is an enormous diversity of technologies applied to inertial navigation sensors making them different from one another; the large quality range of the inertial sensor market leads to the need and importance of some type of inertial sensor classification. Although this may be true, it should be stated, that up to now there is no universally agreed definition of high, medium, and, low grade IMUs. In fact, frequently, the lack of information on the reliability of the IMUs (MEMS-based) leads to its incorrect application. Equally important is the fact that, when faced with such a variety of inertial sensors, it is also quite hard to know which technical data one should analyze and compare before making a decision. Sometimes it can be very difficult, especially for inexpert users of such technology, to make the right decision about which one will meet the application's requirements the best. Important to realize is also the fact that sometimes many different sensor technologies offer a different range of advantages and disadvantages while offering a similar performance. Therefore, this paper presents a methodology that has been developed to compare two different low performance IMUs in order to select one of them for our application. This methodology is also developed to evaluate if the procedures applied to one sensor are adaptable to others in order to achieve a unique methodology when working with such a category of inertial sensors.