1
American Institute of Aeronautics and Astronautics
R EDUCED O RDER G UIDANCE M ETHODS A ND P ROBABILISTIC T ECHNIQUES I N
A DDRESSING M ISSION U NCERTAINTY
Mr. Daniel DeLaurentis
NASA Multidisciplinary Analysis Fellow
Dr. Dimitri N. Mavris
Assistant Professor & Associate Director ASDL
Dr. Anthony J. Calise
Professor, School of Aerospace
Dr. Daniel P. Schrage
Professor & Co-Director ASDL
Aerospace Systems Design Laboratory (ASDL)
*
School of Aerospace Engineering
Georgia Institute of Technology
Atlanta, GA 30332-0150
*
Copyright © 1996 by AIAA. Presented at the 6th
AIAA/NASA/ISSMO Symposium on Multidisciplinary
Analysis and Optimization, Bellevue, WA, Sept. 4-6, 1996.
Abstract
Recognizing that vehicle synthesis fulfills the
role of integrator of the mutually interacting
disciplines, difficulties persist in intelligently
implementing disciplinary analysis into this
synthesis process. This paper develops and describes
analytical and statistical approximation techniques
used to create design-oriented analyses which are
implementable in the process. Specifically,
techniques related to the vehicle guidance discipline
are examined. The ultimate goal is to investigate the
economic viability of an aerospace system in the face
of uncertainty at the system and discipline design
levels. The notion of a “design mission” as a
requirement is replaced by a modeling of mission
variability, since future aircraft will likely fly a
variety of missions. Aircraft guidance laws are key
components in the mission analysis portion of an
aircraft sizing code, and thus they must be included in
the investigation. Through the use of statistical
modeling techniques, a link between mission
uncertainty, optimal guidance, wing planform, and
economic objectives is obtained. This linkage allows
for the investigation of guidance and mission effects
on such quantities as gross weight and ticket price (on
a per mile basis). Further, the resulting solutions are
robust since they are obtained by choosing control
parameters which maximize the probability of
meeting a target while simultaneously assuring that
appropriate constraints (which are also probabilistic)
are met.
Introduction
The consideration of economic uncertainty on the
aircraft synthesis process is the focus of recent
research as a means to generate robust solutions.
1,2
Building on the insights of this work, this paper
expands the types of uncertainty considered to include
variability related to the vehicle guidance discipline as
well as to mission requirements. In general, defining
requirements is a key first step to a successful design
process. In fact, requirements definition is a key
element of the Integrated Product and Process
Development (IPPD)/Concurrent Engineering (CE)
concept which has emerged recently as a viable means
of implementing concurrent engineering practices in
aerospace systems design.
3
Designing aircraft in an
IPPD/CE framework is viewed as designing with a
focus on affordability, which implies an
understanding of how various discipline, mission,
design, and economic variables affect the feasibility
(“can it be built”) and viability (“should it be built”)
of an aircraft. It is in evaluating feasibility and
viability that one quickly realizes the non-
deterministic nature of the design problem. This is
because parameters traditionally treated as fixed
assumptions are in reality distributions around a most
likely value. For example, deterministic methods
proceed “assuming a fuel price of $1/gallon” whereas
probabilistic methods proceed “assuming a range of
possible fuel prices represented by a probability
distribution”. Probabilistic design methods have,
until recently, been discipline specific. In recent
years, research in the controls area has had its
emphasis on robustness, focusing on the conflicting