The aggregation function
plays an important role
in how we measure the distance between belief states. When we
compute more than one state distance measure, either exhaustively or
by sampling a subset (as previously mentioned), we must combine the
measures by some means, denoted
. There is a
range of options for taking the state distances and aggregating them
into a belief state distance. We discuss several assumptions
associated with potential measures:
.
The belief state distances are
and
. In this case we
prefer
to
. If each
state distance is admissible and we do not sample from belief states, then assuming positive interaction
is also admissible.
.
In our example,
, and
. In this case we have no
preference over
and
.
We notice that using the cardinality of a belief state
to measure
is
a special case of assuming state independence, where
.
If we use cardinality to measure distance in our example, then we have
,
and
. With cardinality we prefer
over
because we have better knowledge in
.
.
Depending on the type of search, we define
differently. We assume that sequences used in progression
search start at the same time and those used in regression end at the same
time. Thus, in progression all sequences are aligned at the
first step before we union steps, and in regression all
sequences are aligned at the last step before the union.
For example, in progression
because we align the
sequences at their first steps, then union each step. Notice that this resulting sequence
has seven actions, giving
,
whereas defining
as maximum gave a distance
of five and as summation gave a distance of eight. Compared with
overlap, positive interaction tends to under estimate distance,
and independence tends to over estimate distance. As we
will see during our empirical evaluation (in Section 6.5),
accounting for overlap provides more accurate distance measures
for many conformant planning domains.
There are two ways negative interactions play a role in belief state
distances. Negative interactions can allow us to prove it is
impossible for a belief state
to reach a belief state
, meaning
, or they can potentially
increase the distance by a finite amount. We use only the
first, more extreme, notion of negative interaction by computing ``cross-world'' mutexes [30]
to prune belief states from the search. If we cannot prune a belief state, then
we use one of the aforementioned techniques to aggregate state
distances. As such, we do not provide a concrete definition for
to measure negative interaction.
While we do not explore ways to adjust the distance measure for
negative interactions, we mention some possibilities. Like work
in classical planning [23], we can penalize the distance measure
to
reflect additional cost associated with serializing conflicting actions.
Additionally in conditional planning, conflicting actions can be conditioned on
observations so that they do not execute in the same plan
branch. A distance measure that uses observations would reflect the added cost of
obtaining observations, as well as the change in cost associated with introducing
plan branches (e.g., measuring average branch cost).
The above techniques for belief state distance estimation in terms
of state distances provide the basis for our use of multiple
planning graphs. We will show in the empirical evaluation that
these measures affect planner performance very differently across
standard conformant and conditional planning domains. While it can
be quite costly to compute several state distance measures,
understanding how to aggregate state distances sets the foundation
for techniques we develop in the
. As we have already
mentioned, the
conveniently allows us to implicitly aggregate
state distances to directly measure belief state distance.