|
You are at: Wagner
Home > Technologies > Mission Planning > Search Optimization > TBM Search Optimization
Tactical
Ballistic Missile (TBM) Radar Search Optimization for Aegis
The U.S. Navy's Aegis system was designed as a
total weapon system, from detection to kill. The heart of the
system is an
advanced, automatic detect and track, multi-function phased-array
radar, the AN/SPY-1. This high powered (four megawatt) radar
can perform search, track and missile guidance functions simultaneously
with a track capacity of over 100 targets.
To examine how this system might be optimized to defend against
large TBM raids, Daniel H. Wagner Associates developed a prototype radar
search optimizer. This prototype software recommends an energy scheduling plan
for multiple phased-array radars, based on threat probability maps for
multiple incoming TBMs. A phased-array radar such
as the AN/SPY-1 is freed from the constraints of a mechanically
rotating antenna and can radiate in any direction at any time.
Therefore it can benefit significantly from an algorithm that
can optimally allocate the available radar energy among all of
the possible individual directions at all times. The radar's
effectiveness can be particularly enhanced if the optimal allocation
takes into account such factors as the locations and values of
target areas to be defended. The output of the optimizer is a
schedule of beam energy allocations over the time of interest.
The research described here was supported by the Navy AEGIS
program through the Naval Surface Warfare Center-Dahlgren Division.
Approach
The prototype SPY-1 optimizer models the flight path of each
incoming TBM through a collection of sub-targets representing
the TBM. These sub-targets are generated using a ballistic/3-D
Monte Carlo motion model. The ballistic/3-D motion model generates
tracks for these sub-targets by: (1) choosing initial points
for each sub-target based on the TBMs last reported position,
and (2) projecting these tracks forward in time, until impact,
using a ballistic motion model. Initially, each of the sub-targets
is assigned a likelihood weight of 1.0.
Using these tracks, the software then assigns each sub-target
a threat-based weight which will depend on where the sub-target
impacts relative to the area which is being protected from TBMs.
Only sub-targets having non-zero weights are used: others are
discarded. This allows the optimizer to produce plans which concentrate
on the flight paths that present the greatest threat to the surface
regions being protected.
|
All Monte Carlo sub-targets
for a single TBM are generated from the same initial DSP report.
Those with velocities that carry them outside any defended area
are discarded. The remaining sub-targets are weighted according
to the area values and used by the optimizer to produce radar
search plans. |
 |
Given these tracks for the sub-targets and their associated
threat based weights, the algorithm then optimizes the allocation
of SPY-1 radar energy over equally-spaced time intervalsusing
Brown's algorithm to maximize the probability of detecting the
TBM while it can still be intercepted. (Brown's algorithm is
an iterative optimization approach first proven to be optimal
for moving targets by Dr. Scott Brown of Wagner Associates.)
Multiple-TBM
Attack Example
In the full scenario example, there are 10 simultaneous inbound
TBMs. Defending radars are located at Wallops Island, Virginia and on
"Sideship" (onshore). They and "Picket" (onshore)
are possible shooters and are given a nominal radar energy budget
roughly equivalent to half the total available energy.
The main defended area is a 50 nm circle around Battle Group
(BG) with a relative weight of 10. There are two other circles,
one at Wallops and one at Sideship, each with relative weight
of 1. All the circles have a 10% (5 nm) border in which the weight
decreases linearly to zero. All weight outside these circles
is zero.
Range of all attacking TBMs is about 300 nm. The blue tracks
are at the maximum range angle. The red tracks are lofted to
an apogee about twice as high.
All TBMs enter radar range of Wallops at the same time, at LAUNCH+320
seconds. The optimizer maximizes kill probability by interceptors,
so it gives preference to early detection of sub-targets at which
two shots are possible. The optimized plan splits the search
effort between Wallops and Sideship, with Wallops taking the
northern half of the weight.
The moving image shows the top-view probability map for the 10
TBMs at 20 second intervals of simulated time. The desired result
is that all undetected TBM probability mass be reduced to a minimum
value before entering the areas being defended. Notice that the
southernmost TBMs are detected earlier because those TBMs pass
very close to Sideship. Note that the legend shows the cumulative
probability of containment only for undetected TBMs.
The probability scale is cumulative and not normalized. That
is, the numbers alongside the colors indicate the cumulative
distribution of undetected missiles (100% = 10 missiles). The CDP gives the Cumulative Detection Probability obtained up to
the time of the probability map.
For more information contact
Dr.
Peter McMorran at (757) 727-7700.
|
