Sequence comparison techniques for multisensor data fusion andtarget recognition

A new class of techniques for multisensor fusion and target recognition is proposed using sequence comparison by dynamic programming and multiple model estimation. The objective is to fuse information on the kinematic state and “nonkinematic” signature of unclassified targets, assessing the joint likelihood of all observed events for recognition. Relationships are shown to previous efforts in pattern recognition and state estimation. This research applies “classical” speech processing-related and other sequence comparison methods to moving target recognition, extends the efforts of previous researchers through improved fusion with kinematic information, relates the proposed techniques to Bayesian theory, and applies parameter identification methods to target recognition for improved understanding of the subject in general. The proposed techniques are evaluated and compared with existing approaches using the method of generalized ambiguity functions, which lends to a form of Cramer-Rao lower bound for target recognition     

Reduced-order multiple model adaptive controller for flexiblespacestructure

The robustness of a moving-bank multiple model adaptive estimator/controller to order reduction in the controller design model is examined. It is shown that the adaptive mechanism and bank-moving logic are not confounded by the effects of unmodeled higher order modes of a large flexible spacestructure. Control characteristics are achieved that are essentially equivalent to those of an artificially informed benchmark controller     

MMAE-Based Control with Space-Time Point Process Observations

A multiple model adaptive estimator (MMAE) has been formulated to estimate the state of a dynamic system modeled by a linear stochastic differential equation, from which measurements, described as a noise-corrupted space-time point process functionally related to that state, are extracted. Assumed certainty equivalence is used to combine such an estimator with the LQ full-state feedback controller to synthesize a practical, implementable controller. Performance of the estimator and resultant controller characteristics are investigated via simulation as a function of approximation method used to limit the full-scale estimator to finite dimensionality and also as a function of important parameters defining the dynamics and observation processes.     

Multipath-adaptive GPS/INS receiver

Multipath interference is one of the contributing sources of errors in precise global positioning system (GPS) position determination. This paper identifies key parameters of a multipath signal, focusing on estimating them accurately in order to mitigate multipath effects. Multiple model adaptive estimation (MMAE) techniques are applied to an inertial navigation system (INS)-coupled GPS receiver, based on a federated (distributed) Kalman filter design, to estimate the desired multipath parameters. The system configuration is one in which a GPS receiver and an INS are integrated together at the level of the in-phase and quadrature phase (I and Q) signals, rather than at the level of pseudo-range signals or navigation solutions. The system model of the MMAE is presented and the elemental Kalman filter design is examined. Different parameter search spaces are examined for accurate multipath parameter identification. The resulting GPS/INS receiver designs are validated through computer simulation of a user receiving signals from GPS satellites with multipath signal interference present The designed adaptive receiver provides pseudo-range estimates that are corrected for the effects of multipath interference, resulting in an integrated system that performs well with or without multipath interference present.     

Characterization of Kalman filter residuals in the presence ofmismodeling

The mean and covariance of a Kalman filter residual are computed for specific cases in which the Kalman filter model differs from a linear model that accurately represents the true system (the truth model). Multiple model adaptive estimation (MMAE) uses a bank of Kalman filters, each with a different internal model, and a hypothesis testing algorithm that uses the residuals from this bank of Kalman filters to estimate the true system model. At most, only one Kalman filter model will exactly match the truth model and will produce a residual whose mean and standard deviation have already been analyzed. All of the other filters use internal models that mismodel the true system. We compute the effects of a mismodeled input matrix, output matrix, and state transition matrix on these residuals. The computed mean and covariance are compared with simulation results of flight control failures that correspond to mismodeled input matrices and output matrices     

An alternate algorithm for discrete-time filtering

The discrete-time Kalman filter is an optimal estimator for the states of a linear, stochastic system. It assumes that measurements are linear combinations of the states, and all disturbances are Gaussian. The influence diagram, a decision analysis tool that provides an algorithm for discrete-time filtering equivalent to the Kalman filter when the influence diagram represents Gaussian random variables, is discussed. The influence diagram algorithm is a factored form of the Kalman filter, similar to other factored forms such as the U-D filter. Compared with the Kalman filter, it offers improved numerical properties. Compared with other factored forms, it offers a reduced computational load     

Detection of interference/jamming and spoofing in a DGPS-aidedinertial system

Previous research at the Air Force Institute of Technology (AFIT) has resulted in the design of a differential Global Positioning System (DGPS) aided INS-based (inertial navigation system) precision landing system (PLS) capable of meeting the FAA precision requirements for instrument landings. The susceptibility of DGPS transmissions to both intentional and nonintentional interference/jamming and spoofing must be addressed before DGPS may be safely used as a major component of such a critical navigational device. This research applies multiple model adaptive estimation (MMAE) techniques to the problem of detecting and identifying interference/jamming and spoofing in the DGPS signal. Such an MMAE is composed of a bank of parallel filters, each hypothesizing a different failure status, along with an evaluation of the current probability of each hypothesis being correct, to form a probability-weighted average state estimate as an output. For interference/jamming degradation represented as increased measurement noise variance, simulation results show that, because of the good failure detection and isolation (FDI) performance using MMAE, the blended navigation performance is essentially that of a single extended Kalman filter (EKF) artificially informed of the actual interference noise variance. However, a standard MMAE is completely unable to detect spoofing failures (modeled as a bias or ramp offset signal directly added to the measurement). This work describes a moving-bank pseudoresidual MMAE (PRMMAE) to detect and identify such spoofing. Using the PRMMAE algorithm, spoofing is very effectively detected and isolated; the resulting navigation performance is equivalent to that of an EKF operating in an environment without spoofing     

Multiple model adaptive estimation with filter spawning

Multiple model adaptive estimation (MMAE) with filter spawning is used to detect and estimate partial actuator failures on the VISTA F-16. The truth model is a full six-degree-of-freedom simulation provided by Calspan and General Dynamics. The design models are chosen as 13-state linearized models, including first order actuator models. Actuator failures are incorporated into the truth model and design model assuming a "failure to free stream." Filter spawning is used to include additional filters with partial actuator failure hypotheses into the MMAE bank. The spawned filters are based on varying degrees of partial failures (in terms of effectiveness) associated with the complete-actuaton-failure hypothesis with the highest conditional probability of correctness at the current time. Thus, a blended estimate of the failure effectiveness is found using the filters' estimates based upon a no-failure hypothesis, a complete actuator failure hypothesis, and the spawned filters' partial-failure hypotheses. This yields substantial precision in effectiveness estimation, compared with what is possible without spawning additional filters, making partial failure adaptation a viable methodology.     

Adaptive Tracker Field-of-View Variation Via Multiple Model Filtering

Adaptive estimation using multiple model filtering is investigated as a means of changing the field of view as well as the bandwidth of an infrared image tracker when target acceleration can vary over a wide range. The multiple models are created by tuning filters for best performance at differing conditions of exhibited target behavior and differing physical size of their respective fields of view. Probabilistically weighted averaging provides the adaptation mechanism. Each filter involves online identification of the target shape function, so that this algorithm can be used against ill-defined and/or multiple-hot-spot targets. When each individual filter has the form of an enhanced correlator/linear Kalman filter, computational loading is very low. In contrast, an extended Kalman filter processing the raw infrared data directly and assuming a nonlinear constant turn-rate dynamics model provides superior tracking capability, especially for harsh maneuvers, at the cost of a larger computational burden.