An improved method of using multiple point crash sensing and multiple sensor
occupant
position sensing for classifying a crash event and determining which restraints
should be deployed. A central controller collects crash data from multiple crash
sensors and combines severity characterization data from each of the multiple sensors
to construct a characterization table or matrix for the entire system. Each possible
crash event classification is represented by a characterization value mask, and
the various masks are sequentially applied to the system characterization table
until a match is found, with a match identifying the appropriate crash event classification.
The classification decision, in turn, is used to determine which, if any, of the
restraint devices should be deployed based upon the crash severity. Similarly,
the controller collects data from various occupant position sensors to construct
a characterization table or matrix for the occupant position detection system.
Each possible occupant position sensor classification is represented by a characterization
value mask, and various masks are sequentially applied to the table until a match
is found, with a match identifying the appropriate occupant position status. The
occupant position status, in turn, is used to determine which, if any, of the restraints
may be deployed. The system also includes a centrally located crash sensor, and
the controller constructs an intrusion table based on differences between the remote
and central sensors. The intrusion classification is determined and combined with
the crash classification and occupant position status to determine which restraints
should ultimately be deployed.
