A Stochastic Limit Approach to the SAT Problem