Browsing by Subject "Laplace"
Now showing 1 - 2 of 2
Results Per Page
Sort Options
Item A Comprehensive Comparison Between Angles-Only Initial Orbit Determination Techniques(2012-02-14) Schaeperkoetter, Andrew VernonDuring the last two centuries many methods have been proposed to solve the angles-only initial orbit determination problem. As this problem continues to be relevant as an initial estimate is needed before high accuracy orbit determination is accomplished, it is important to perform direct comparisons among the popular methods with the aim of determining which methods are the most suitable (accuracy, robustness) for the most important orbit determination scenarios. The methods tested in this analysis were the Laplace method, the Gauss method (suing the Gibbs and Herrick-Gibbs methods to supplement), the Double R method, and the Gooding method. These were tested on a variety of scenarios and popular orbits. A number of methods for quantifying the error have been proposed previously. Unfortunately, many of these methods can overwhelm the analyst with data. A new method is used here that has been shown in previous research by the author. The orbit error is here quantified by two new general orbit error parameters identifying the capability to capture the orbit shape and the orbit orientation. The study concludes that for nearly all but a few cases, the Gooding method best estimates the orbit, except in the case for the polar orbit for which it depends on the observation interval whether one uses the Gooding method or the Double R method. All the methods were found to be robust with respect to noise and the initial guess (if required by the method). All the methods other than the Laplace method suffered no adverse effects when additional observation sites were used and when the observation intervals were unequal. Lastly for the case when the observer is in space, it was found that typically the Gooding method performed the best if a good estimate is known for the range, otherwise the Laplace method is generally best.Item Bayesian model selection using exact and approximated posterior probabilities with applications to Star Data(Texas A&M University, 2004-11-15) Pokta, SurianiThis research consists of two parts. The first part examines the posterior probability integrals for a family of linear models which arises from the work of Hart, Koen and Lombard (2003). Applying Laplace's method to these integrals is not entirely straightforward. One of the requirements is to analyze the asymptotic behavior of the information matrices as the sample size tends to infinity. This requires a number of analytic tricks, including viewing our covariance matrices as tending to differential operators. The use of differential operators and their Green's functions can provide a convenient and systematic method to asymptotically invert the covariance matrices. Once we have found the asymptotic behavior of the information matrices, we will see that in most cases BIC provides a reasonable approximation to the log of the posterior probability and Laplace's method gives more terms in the expansion and hence provides a slightly better approximation. In other cases, a number of pathologies will arise. We will see that in one case, BIC does not provide an asymptotically consistent estimate of the posterior probability; however, the more general Laplace's method will provide such an estimate. In another case, we will see that a naive application of Laplace's method will give a misleading answer and Laplace's method must be adapted to give the correct answer. The second part uses numerical methods to compute the "exact" posterior probabilities and compare them to the approximations arising from BIC and Laplace's method.