BDD-Guided Clause Generation

  title={BDD-Guided Clause Generation},
  author={B. Kell and Ashish Sabharwal and W. V. Hoeve},
Nogood learning is a critical component of Boolean satisfiability (SAT) solvers, and increasingly popular in the context of integer programming and constraint programming. We present a generic method to learn valid clauses from exact or approximate binary decision diagrams (BDDs) and resolution in the context of SAT solving. We show that any clause learned from SAT conflict analysis can also be generated using our method, while, in addition, we can generate stronger clauses that cannot be… Expand
Decision Diagrams for Combinatorial Optimization and Satisfaction
This thesis presents a generic method for deducing valid clauses from a SAT instance using BDDs, with the aim of finding clauses that are bottlenecks for SAT solvers using conflict-directed clause learning. Expand
Decision Diagrams for Discrete Optimization
In this tutorial, I will present an overview of the recent successful applications of multivalued decision diagrams (MDDs) to discrete optimization problems, with applications to constraintExpand
The Size of BDDs and Other Data Structures in Temporal Logics Model Checking
It is formally proved that no kind of data structure of polynomial size can represent the set of valid initial states for all models and all formulae, and this result holds for all data structures where a state can be checked inPolynomial time. Expand
Integrating Relaxations for Combinatorial Optimization
A new method for finding lower bounds based on a Lagrangian relaxation is proposed, and a subgradient optimization scheme is applied to find good bounds quickly, and it is shown experimentally that this method can find bounds that are very close to the optimal LP bound in a fraction of the time that is needed to compute the LP bound. Expand
Domain reduction techniques for global NLP and MINLP optimization
This work surveys the various techniques used for domain reduction of NLP and MINLP optimization problems and presents a computational analysis of the impact of these techniques on the performance of various widely available global solvers on a collection of 1740 test problems. Expand


Lazy Clause Generation: Combining the Power of SAT and CP (and MIP?) Solving
This invited talk explains lazy clause generation, and explores some of the many design choices in building such a hybrid system, and discusses how to further incorporate mixed integer programming (MIP) solving to see if it can also inherit its advantages in combinatorial optimization. Expand
On the power of clause-learning SAT solvers as resolution engines
It is proved that conflict-driven clause-learning SAT solvers yield proof systems that indeed p-simulate general resolution without the need for any additional techniques, and this result can be generalized to certain other practical variations of the solvers, which are based on different learning schemes and restart policies. Expand
Nogood processing in csps
The constraint satisfaction problem is an NP-complete problem that provides a convenient framework for expressing many computationally hard problems. In addition, domain knowledge can be efficientlyExpand
Towards Understanding and Harnessing the Potential of Clause Learning
The first precise characterization of clause learning as a proof system (CL) is presented, and it is shown that with a new learning scheme, CL can provide exponentially shorter proofs than many proper refinements of general resolution satisfying a natural property. Expand
Conflict analysis in mixed integer programming
This paper presents heuristics for branch-and-cut solvers to generate valid inequalities from the current infeasible subproblem and the associated branching information and generalizes SAT infeasibility analysis to mixed integer programming. Expand
Explaining Flow-Based Propagation
This paper presents two new generic flow-based propagators with several novel features, and most importantly, the addition of explanation capability, and shows that the generic propagators can efficiently replace specialized versions, in particular for gcc and sequence constraints. Expand
Clause-Learning Algorithms with Many Restarts and Bounded-Width Resolution
It is proved that, although the solver is not explicitly designed for it, with high probability it ends up behaving as width-k resolution after no more than O(n2k+2) conflicts and restarts, where n is the number of variables. Expand
On Threshold BDDs and the Optimal Variable Ordering Problem
A 0/1 IP formulation provides the basis for the computation of the variable ordering spectrum of a threshold function, which is an NP-hard problem. Expand
Propagation via lazy clause generation
This paper describes how to mimic a finite domain propagation engine, by mapping propagators into clauses in a SAT solver, and shows that the resulting system solves many finite domain problems significantly faster than other techniques. Expand
Optimization Bounds from Binary Decision Diagrams
This work shows how to construct a BDD that represents a relaxation of an optimization problem with binary variables, and how to obtain a bound for any separable objective function by solving a shortest (or longest) path problem in the BDD. Expand