Reachability for linear systems using vrep

[mforets]

https://nbviewer.jupyter.org/github/mforets/escritoire/blob/master/sets/DiscreteSequence.ipynb

[schillic]

If it is not too difficult, it would be good to have a "generic" algorithm that we can then also instantiate with H-rep or other set representations for comparison.

Should this only be applicable to 2D systems or is the idea to have a decomposition as in BFFPSV18?

[mforets]

If it is not too difficult, it would be good to have a "generic" algorithm that we can then also instantiate with H-rep or other set representations for comparison.

Ok, i'll have that in mind. Assuming that the "general" set can handle linear maps and M-sums (either concrete or lazy), the additional requirement is always an overapproximate methods which is closed under the given set representation.

Should this only be applicable to 2D systems or is the idea to have a decomposition as in BFFPSV18?

Interesting question. Yes, i think that the method can be used as a drop-in replacement. Different blocks need to use the same Whatk though (it can be computed by each process or shared).