Research of Optimized Adaptive Kalman Filtering

Standard Kalman Filtering leads to divergence because of inaccurate system model and noise statistic. Researchers have taken relative studies about Kalman filtering optimization method. But now most studies are based on applications, such as integrated navigation system, so most of these methods are lack of general applicability. This paper starts from innovation-based adaptive estimation (IAE) filtering and memory attenuated (MA) filtering. These two optimized filtering methods have respective advantages and disadvantages. We combined them to create a new optimal filtering, namely, Optimized Adaptive Kalman Filtering (OAKF). New method gained the attenuation factor values in real time by innovation analysis to control the effect of filtering. Software simulation shows the new optimized adaptive Kalman filtering´s good effect in different conditions. After compared with the standard Kalman filtering, IAE filtering and MA filtering, OAKF methods has better filtering effect than other methods.