Iterative cooperative CAD for avionics navigation systems

The paper is devoted to the extension to avionics of the main ideas developed in concurrent engineering for creating design tools and procedures that enable an engineer to devise inertial navigation systems (INS) which will outperform comparable systems for accuracy and reliability thus affording a quantum leap forward in flight safety. The main goal of the paper is to see how the total INS error can be divided up so that contributions from the separate error sources can be obtained. The tools for reaching this goal are the following. First, we formulate a mathematically tractable extended INS error model that would include all the separate error sources contributing to the resulting INS readouts. Second, we construct a real data model to replace the real system for simulated tests. Third, we formulate a simplified INS error model on which the extended numerically stable Kalman filter is to be based. Fourth, we determine the best way to make this filter insensitive to uncertainty inherent to the INS error model. Fifth, the software (named Adaptive System Parameter Identification) implementing the above solutions, is developed and verified through the execution of a wide range of computational experiments. Finally, we address the efficiency issue of the ASPID in concurrent design.