Experiences in Managing the Prometheus Project

Congress authorized NASA's Prometheus Project in February 2003, with the first Prometheus mission slated to explore the icy moons of Jupiter. The project had two major objectives: 1) to develop a nuclear reactor that would provide unprecedented levels of power and show that it could be processed safely and operated reliably in space for long-duration, deep-space exploration; and 2) to explore the three icy moons of Jupiter - Callisto, Ganymede, and Europa - and return science data that would meet the scientific goals as set forth in the Decadal Survey Report of the National Academy of Sciences. Early in project planning, it was determined that the development of the Prometheus nuclear-powered spaceship would be complex and require the intellectual knowledge residing at numerous organizations across the country. In addition, because of the complex nature of the project and the multiple partners, approaches beyond those successfully used to manage a typical JPL project would be needed. This describes the key experiences in managing Prometheus, which should prove useful for future projects of similar scope and magnitude.     

Optimum Realizable Continuous Range Estimation

The ability to detect the presence or absence of a target is no longer the fundamental design criterion when the vehicle to be tracked is cooperative. In spacecraft tracking or navigation systems, for example, emphasis is placed on post-acquisition performance. Therefore, classical radar theory and design techniques are not specifically applicable. On the other hand, there are optimization techniques for extracting the tracking data from noise that are more to the point. In particular, optimum demodulation theory is directed specifically to the problem of continuously extracting data from a nonlinear modulation process. In this paper, the tracking properties of a multitone PM ranging signal are reviewed and are shown to be nearly optimum for cooperative vehicles. An optimum, but nonrealizable, maximum a posteriori (MAP) continuous estimator of range is derived for this signal. The linearized model of this receiver is the optimum nonrealizable Wiener filter for the data. Interpretation of this optimum nonrealizable estimator leads to a receiver design that is both practical and intuitively satisfying. With the aid of post-detection processing in the Wiener-Hopf sense, almost optimum performance is obtained from the resulting receiver, above threshold.     

Nonlinear target enhancement for the hostile nuclear environment

A necessary requirement of a strategic defence system is the detection of incoming nuclear warheads in an environment that may include nuclear detonations of undetected or missed target warheads. A computer model which simulates incoming warheads as distant endoatmospheric targets is described. A model of the electromagnetic noise expected in the nuclear environment is developed. Predicted atmospheric effects are also included. The ability of nonlinear image enhancement algorithms to their ability to suppress the noise and atmospheric effects of the nuclear environment is examined. These algorithms are then tested, using the combined target and noise models, and evaluated in terms of noise removal and their ability to resolve closely spaced targets     

Implementation and convergence considerations of a linearlyconstrained adaptive array

O.L. Frost (1972) introduced a linearly constrained optimization algorithm that allows certain main beam properties to be preserved while good cancellation is attained. An open-loop implementation of this algorithm is developed. This implementation is shown to be equivalent to the technique developed by C.W. Jim (1977), L.J. Griffiths and C.W. Jin (1982), and K.M. Buckley and L.J. Griffiths (1982) whereby the constrained problem is reduced to an unconstrained problem. Analytical results are presented for the convergence rate when the sampled matrix inversion (SMI) or Gram-Schmidt (GS) algorithm are employed. It has been previously shown that the steady-state solution for the optimal weights is identical for both constrained and reduced unconstrained problems. It is shown that if the SMI or GS algorithm is employed, then the transient weighting vector solution for the constrained problem is identical to the equivalent transient weight vector solution for the reduced unconstrained implementation     

Comments, with reply, on `Lyapunov stability designs for switchingregulator/low-pass filter system' by M.L. Brooks

Erroneous conclusions reached in the original paper (see ibid., vol.25, no.5, p.602-10, Sept. 1989) concerning the applicability of the Rough criterion in studying the stability of fourth-order linear systems and its relationship with Lyapunov's stability tests are corrected. A brief review of relevant theory is presented, and faults in the analysis that led to these conclusions are identified. The relationship between the Routh criterion and the second method of Lyapunov is reviewed, and a correct application of the Lyapunov method is presented. It is shown that contrary to the conclusions of the paper discussed, these two methods are equivalent for any linear, time-invariant system, including the fourth-order system in question     

An interactive system for global ship monitoring

The US Navy has more than 300 ships equipped with over 500 PHALANX weapon systems. Ordalts (ordnance alterations) and casreps (casualty reports) are received frequently by Naval Ordnance. Ordalts are done when parts need replacing or a design is exchanged for an old. Casreps are an urgent request for parts or service. Information pertaining to ordalts and casreps is put into several databases, which are continually updated and used in reports among the PHALANX community. Ship locations are used for sending parts and/or assistance to the ship. The ordalt and casrep requests have been combined with the ship locations report to produce a more efficient manner in which to monitor ordalt requests, casrep requests and ship locations, using a sample database containing ordalt and casrep data and ship locations. The database is interfaced to a computer graphics program that monitors ship movement throughout the world and supplies information from the database whenever requested. The hardware and software used to implement the project are described