Strong tracking filtering (STF) is a popular adaptiveestimation method to effectively deal with state estimationfor linear and nonlinear dynamic systems with inaccuratemodels or sudden change of state. The key of the STF is to usea time-variant fading factor, which can be evaluated based onthe current measurement innovation in real time, to forcefullycorrect one step state prediction error covariance. The strongtracking filtering technology has been extensively applied inmany practical systems, but the theoretical analysis is highlylacking. In an effort to better understand STF, a novel analysisframework is developed for the strong tracking filtering andsome new problems are discussed for the first time. For this, wepropose a new perspective that correcting the state predictionerror covariance by using the fading factor can be thought ofdirectly modifying the state model by correcting the covarianceof the process noise. Based on this proposed point of view,the conditions for the STF function to be effective are deeplyanalyzed in a certain linear dynamic system. Meanwhile, issuesof false alarm and alarm failure are also briefly discussed for thestrong tracking filtering function. Some numerical simulationexamples are demonstrated to validate the results.
展开▼
