Abstract:
A parallel technology that can be used to solve various problems of discrete
optimization is proposed. This technology is based on a number of efficient
procedures of reducing combinatorial optimization problems to SAT problems.
A process of solving an optimization problem is implemented as an iterative
scheme such that a SAT problem is solved at each step using various techniques
of parallelization. In order to take into account the information accumulated
during previous iterations, a special technique called “Incremental SAT” is
used. This technique is usually applied to solve the verification problems for
discrete systems. The technology proposed was tested using various combinatorial
problems, in particular, the quadratic assignment problem.
Keywords:inversion of discrete functions; satisfiability problem (SAT-problem); Boolean equations; problems of combinatorial optimization.