Generalized Time-invariant Noise Algorithm and Related Bifurcation Problem
Wei He · Zhongyang Minzu Daxue xuebao · 2003
In this paper, at first, we introduce Simulated Annealing(SA), give its explanation of physic and mathematics. Secondly, a strict theoretical proof about the theory of the generalized Time\|Invariant Noise Algorithm(TINA), applying to global optimization in [5] is discussed. This is the first proof we have seen about the mentioned problem.\;We prove that the algorithm converges to a global extreme value point in probability for parameter α05; but the stochastic dynamical system may have bifurcation phenomena for parameter α=05.