A Method of Data Fusion Based on the Robust Minimum Variance Filtering

In the data fusion, Kalman filter is widely used to process the synchronous and asynchronous sensor data. However, the filter will not present a good performance when the model´s parameters and the noise´s characteristic are not assured. This paper gives a method of data fusion based on the robust filtering. The method can guarantee the stable filtering as long as the system´s parameters are in a certain range. This method can process both the synchronous data and the asynchronous data. The result of this experiment verified that this method can deal with the data fusion in the situation that the system parameters are not exact.