Target tracking using infrared measurements and laser illumination

A missile target tracker is designed using a filter/correlator (with adaptive target shape identification) based on forward-looking infrared (FLIR) sensor measurements to track the center-of-intensity of the hardbody/plume combination, and another filter using Doppler and/or speckle information in the return from a low-power laser illuminator to estimate the offset between the intensity centroid and the hardbody center-of-mass. The Doppler information is shown to yield smaller bias and error variance from the tracker than the speckle information. Performance of trackers based on just Doppler or both Doppler and speckle information from the laser return is portrayed as a function of important parameters in the tracking environment     

Flexible spacestructure control via moving-bank multiple modelalgorithms

The development and performance of moving-bank multiple model adaptive control (MMAC) algorithms for quelling vibrations induced in the SPICE 2 space structure are presented. The structure consists of a large platform and a smaller platform connected by three legs in a tripod fashion. Deviations of the line-of-sight (LOS) vector from the center of the large platform to the center of the smaller platform are used for LQG controller performance evaluation. The parameter estimator implements the maximum entropy with identity covariance (ME/I) algorithm; the moving-bank logic employs parameter position monitoring; the controller uses the modified MMAC method. Whereas parameter variations of two percent caused instabilities in the single filter/controller design, the MMAC algorithm provides an excellent method to estimate a wide range of parameter variations and to quell oscillations in the structure     

Reducing lag in virtual displays using multiple model adaptiveestimation

A multiple model adaptive predictor is applied to a virtual environment flight simulator to remove the effect of computational and scene-rendering delay time. Angular orientation of the user's head is predicted a period of time into the future, so that the scene can be rendered appropriately by the time the user actually looks in that direction. Single nonadaptive predictors cannot adequately cover the dynamic range of head motion. By using three dissimilar models of head motion upon which to base the individual elemental filters within the multiple model adaptive estimator (MMAE) algorithm, an MMAE is designed which outperforms the nonadaptive Kalman predictor proposed by Liang (1991)     

Interrelationship of single-filter and multiple-model adaptivealgorithms

An equivalent filter bank structure for multiple model adaptive estimation (MMAE) is developed that uses the residual and state estimates from a single Kalman filter and linear transforms to produce equivalent residuals of a complete Kalman filter bank. The linear transforms, which are a function of the differences between the system models used by the various Kalman filters, are developed for modeling differences in the system input matrix, the output matrix, and the state transition matrix. The computational cost of this new structure is compared with the cost of the standard Kalman filter bank (SKFB) for each of these modeling differences. This structure is quite similar to the generalized likelihood ratio (GLR) structure, where the linear transforms can be used to compute the matched filters used in the GLR approach. This approach produces the best matched filters in the sense that they truly represent the time history of the residuals caused by a physically motivated failure model     

Enhancements to a multiple model adaptive estimator/image-tracker

Three major enhancements to a previously devised multiple model adaptive estimator (MMAE) for target image tracking are developed and analyzed. These are: allowing some of the elemental filters to have rectangular fields of view and to be tuned for target dynamics that are harsher in one direction than others; considering both Gauss-Markov acceleration models and constant turn-rate models for target dynamics; and devising an initial target acquisition algorithm to remove important biases in the estimated target template to be used in a correlator within the tracker. Particularly good adaptation responsiveness is demonstrated in the multiple model algorithm's ability to handle harsh maneuver onset, yielding performance essentially equivalent to that of the best artificially informed tracking algorithm     

Multiple-model adaptive estimation using a residual correlationKalman filter bank

We propose a modified multiple model adaptive estimation (MMAE) algorithm that uses the time correlation of the Kalman filter residuals, in place of their scaled magnitude, to assign conditional probabilities for each of the modeled hypotheses. This modified algorithm, denoted the residual correlation Kalman filter bank (RCKFB), uses the magnitude of an estimate of the correlation of the residual with a slightly modified version of the usual MMAE hypothesis testing algorithm to assign the conditional probabilities to the various hypotheses that are modeled in the Kalman filter bank within the MMAE. This concept is used to detect flight control actuator failures, where the existence of a single frequency sinusoid (which is highly time correlated) in the residual of an elemental filter within an MMAE is indicative of that filter having the wrong actuator failure status hypothesis. This technique results in a delay in detecting the flight control actuator failure because several samples of the residual must be collected before the residual correlation can be estimated. However, it allows a significant reduction of the amplitude of the required system inputs for exciting the various system modes to enhance identifiability, to the point where they may possibly be subliminal, so as not to be objectionable to the pilot and passengers     

Performance enhancement of a multiple model adaptive estimator

We describe performance improvement techniques for a multiple model adaptive estimator (MMAE) used to detect and identify control surface and sensor failures on an unmanned flight vehicle. Initially failure identification was accomplished within 4 s of onset, but by removing the “β dominance” effects, bounding the hypothesis conditional probabilities, retuning the Kalman filters, increasing the penalty for measurement residuals, decreasing the probability smoothing, and increasing residual propagation, the identification time was reduced to 2 s