An Approach to Improve the Performance of Simulated Annealing Algorithm Utilizing the Variable Universe Adaptive Fuzzy Logic System

Abstract

This paper focuses on improving the standard form of the classical simulated annealing algorithm (CSAA). A novel method of improving the performance of CSAA by the variable universe adaptive fuzzy logic system (VUAFLS) is studied. We develop the VUAFLS to adjust the annealing temperature, which is a very important parameter governing the performance of CSAA, and this algorithm is named VUAFLS-CSAA. The main innovations of VUAFLS-CSAA lie in the application of VUAFLS containing the fast cooling mechanism and reheating mechanism relative to the characteristic of the sustained temperature fall of CSAA. Compared with the conventional method for controlling annealing temperature, VUAFLS-CSAA can control the annealing temperature more effectively, leading to the high efficiency of CSAA. The performance of the proposed method is evaluated and compared with CSAA through two examples. One is the image restoration problem, and the other is the traveling salesman problem (TSP). The experimental result indicates that the new method proposed in this paper can improve the efficiency of CSAA by tremendously shortening the iteration optimization process. And at the same time, the successful application of the new method for tackling two different problems demonstrates the generality of this method. In addition, techniques that can further improve the performance of CSAA are discussed.