In this paper, simulated annealing is analyzed and improved. Aiming at multipackphenomena which generally exist in variable optimization of process system, asimulated annealing approach is introduced to search the global optimal solution. Aproducing function of adjacent state is proposed, and effects of initial temperature,terminal temperature as well as annealing schedule to the optimal calculation areanalyzed, and their matched scopes are given. The simulated annealing and traditionaloptional methods are compared and analyzed in examples, these results show simulatedannealing algorithm can efficiently solve the global optional problem which isdifficulty dealt with by the traditional deterministic optional methods. |