Accession Number : ADA194622
Title : Tracking and Control of a Neutral Particle Beam Using Multiple Model Adaptive Meer Filter.
Descriptive Note : Master's thesis,
Corporate Author : AIR FORCE INST OF TECH WRIGHT-PATTERSON AFB OH SCHOOL OF ENGINEERING
Personal Author(s) : Harambasic, Louis J , Jr
PDF Url : ADA194622
Report Date : Dec 1987
Pagination or Media Count : 239
Abstract : The purpose is to point the centroid of a Neutral Particle Beam (NPB) at an intended target. A Multiple Model Adaptive Estimator using elemental Meer Filters is used to estimate the centroid of a NPB model as a one-dimensional first-order Gauss-Markov position process. The MMAE Meer Filter is also used to estimate the beam time constant. 'Merge Method' of filter pruning is used to limit the size of the elemental Meer filters. A bank of three Kalman filters are used to estimate the states of the target which has a variable dynamics driving noise strength. The target is modelled as a third-order Gauss-Markov position process. A Multiple Model Adaptive Controller is designed using LQG methods, and true states are replaced by their best estimates by invoking the principle of assumed certainty equivalence. MMAE Meer Filter performance analysis is performed for an uncontrolled beam and for a controlled beam. Controller baselines are established.
Descriptors : *PARTICLE BEAMS, *TRACKING, *ADAPTIVE FILTERS, *BEAM FORMING, ADAPTIVE CONTROL SYSTEMS, ADAPTIVE SYSTEMS, BASE LINES, ESTIMATES, FILTER ANALYSIS, GAUSSIAN NOISE, KALMAN FILTERING, MARKOV PROCESSES, MODELS, NEUTRAL, PERFORMANCE TESTS, POSITION(LOCATION), TIME, MAXIMUM LIKELIHOOD ESTIMATION, CONTROL THEORY, BAYES THEOREM, POISSON DENSITY FUNCTIONS, PSEUDO RANDOM SYSTEMS, COVARIANCE, MONTE CARLO METHOD, THESES
Subject Categories : Particle Accelerators
Distribution Statement : APPROVED FOR PUBLIC RELEASE