Parallel Random Search Algorithm for Large-Scale Constrained Pseudo-Boolean Optimization Problems

Random search methods are successfully implemented for variety of discrete optimization NP-hard problems when any exact solution approaches cannot be implemented due to large computational demands. Initially designed for unconstrained optimization, the probability changing method gives an approximate solution for various linear and non-linear pseudo-Boolean optimization problems with constraints. Although, in case of large-scale problems, the computational demands are also significant and the precision of the result depends on the spent time. For constrained optimization, the search of any feasible solution may take significant computational resources. In this paper, we consider an approach to developing parallel versions of the algorithms based on the modified probability changing method for constrained pseudoBoolean optimization. Optimization algorithms are adapted for the systems with shared memory (OpenMP) and cluster systems (MPI). The parallel efficiency is estimated for the large-scale non-linear pseudo-Boolean optimization problems with linear constraints and travelling salesman problem.