Now showing 1 - 2 of 2
No Thumbnail Available
Publication

A hybrid simulated annealing and threshold accepting for satisfiability problems using dynamically cooling schemes

2007 , Martínez Ríos, Félix Orlando , Frausto-Solís, Juan

For Satisfiability (SAT) Problem there is not a deterministic algorithm able to solve it in a polynomial time. Simulated Annealing (SA) and similar algorithms like Threshold Accepting (TA) are able to find very good solutions of SAT instances only if their control parameters are correctly tuned. Classical TA usually uses the same Markov chain length for each temperature cycle but they spend a lot of time. In this paper a method based on the neighborhood structure to get the Markov chain length in a dynamical way for each temperature cycle is proposed. Three cooling schemes are also presented in the paper. The experimentation presented in the paper shows that the proposed method is more efficient than the classical one.

No Thumbnail Available
Publication

Simulated Annealing for SAT Problems Using Dynamic Markov Chains with Linear Regression Equilibrium

2008 , Martínez Ríos, Félix Orlando , Frausto-Solís, Juan

Since the appearance of Simulated Annealing (SA) algorithm it has shown to be an efficient method to solve combinatorial optimization problems. This algorithm is based on two cycles: the external or temperature cycle and the internal or Metropolis Cycle. In this paper a new SA method named LRSA is presented. LRSA dynamically finds the equilibrium in the Metropolis cycle by using Linear Regression. Experimentation shows that the proposed method is more efficient than the classical one, since it obtains the same quality in the final solution with less processing time.