An Integrity and Quality Control Procedure for use in Multi Sensor Integration

Abstract

Real-time estimation of parameters in dynamic systems becomes increasingly important in thefield of high precision navigation. The real-time estimation inevitably requires real-time integrity monitoring of the models underlying the navigation system. This paper presents a real-time recursive detection, identification, and adaptation (DIA) procedure for use in integrated navigation systems. It is based on the concept of multi-sensor integration and makes use of the redundancy information stemming from both the measurement model and dynamic model. The tests proposed are optimal in the uniformly-most-powerful invariant sense. Their inverted power function is used to introduce the concept of minimal detectable biases (MDB). The MDB is a diagnostic tool for inferring the detectability of particular model errors. It can be used for the design of a navigation filter that allows for a sufficient control on the presence of bias.