NONLINEAR INTEGER PROGRAMMING BY DARWIN AND BOLTZMANN MIXED STRATEGY

Non-linear integer programming (NIP) is a NP-complete problem with ext ensive theoretical and practical backgrounds. Based on our proposed Da rwin and Boltzmann mixed strategy, this paper presents a general stoch astic iterative algorithm for the NIP problems. The algorithm synthesi zes the advantages of the Darwin strategy and the Boltzmann annealing strategy. It converges asymptotically to the global optimums and has s hown to be polynomial in complexity. The experimental evaluations also show that the proposed algorithm is more efficient than the simulated annealing algorithm. (C) 1998 Elsevier Science B.V.