Symbolic
This folder contains the data structures needed to encode the different abstractions.
Dionysos.Symbolic.SymbolicModel
— TypeSymbolicModel{N, M}
is the abtract type which defines a symbolic model.
Dionysos.Symbolic.SymbolicModelList
— TypeSymbolicModelList{N, M, S1 <: DO.DomainType{N}, S2 <: DO.DomainType{M}, A} <: SymbolicModel{N, M}
is one implementation of the SymbolicModel
type for classical abstraction-based methods, i.e. when the whole domain is partitioned/covered.
Dionysos.Symbolic.LazySymbolicModel
— TypeLazySymbolicModel{N, M, S1 <: DO.DomainType{N}, S2 <: DO.DomainType{M}, A} <: SymbolicModel{N, M}
is one implementation of the SymbolicModel
type for the lazy abstraction-based methods, i.e. when a subset of the domain is partitioned/covered.
Dionysos.Symbolic.compute_symmodel_from_hybridcontrolsystem!
— Functioncompute_symmodel_from_hybridcontrolsystem!(symmodel::SymbolicModel{N}, transitionCost::AbstractDict, transitionCont::AbstractDict,
hybridsys::AbstractHybridSystem, W, L, U, opt_sdp, opt_qp)
Builds an abstraction symmodel
where the transitions have costs given in transitionCost
and are parameterized by affine-feedback controllers in transitionCont
. The concrete system is hybridsys
and W
, L
and U
are defined as in _has_transition
. An SDP optimizer opt_sdp
and a QP optimizer opt_qp
must be provided as JuMP optimizers.