Lazy clause generation is a hybrid SAT and finite domain propagation solver that tries to combine the advantages of both: succinct modelling using finite domains and powerful nogoods and backjumping search using SAT technology. It has been shown that it can solve hard scheduling problems significantly faster than SAT or standard finite domain propagation alone. This new hybrid opens up many choices in modelling problems because of its dual representation of problems as both finite domain and SAT variables. In this paper we investigate some of those choices. Arising out of the modelling choices comes a novel combination of bounds representation and domain propagation which creates a form of propagation of disjunctions. We show this novel modelling approach can outperform more standard approaches on some problems.
Olga Ohrimenko, Peter J. Stuckey