A Greedy Approach for Selection of Residual Generators
This paper considers the problem of selecting a set of
residual generators, fulfilling requirements in terms of
fault isolability and minimal cardinality, for inclusion
in a model-based FDI-system. Two novel algorithms
for solving the selection problem are proposed. The
first one provides an exact solution fulfilling both requirements
and is suitable for small problems. The
second one, which is the main contribution and suitable
for large problems, relaxes the minimal cardinality
requirement and provides an approximate solution
by means of a greedy heuristic. Both algorithms take
the realizability properties of the considered residual
generation method into account, but are general in
the sense that they support any computerized residual
generation method. In a case study the greedy algorithm
is applied to the problem of finding a suitable
set of residual generators for detection and isolation of
faults in a complex truck diesel engine system. In this
study a prior known sequential residual generation
method is considered.
Carl Svärd, Mattias Nyberg and Erik Frisk
22nd International Workshop on Principles of Diagnosis (DX-11),
2011
Informationsansvarig: webmaster
Senast uppdaterad: 2021-11-10
LiU Homepage
